Date: 2/11/19 VARIABLE ANNUITIES CAPITAL AND RESERVE … · 2019. 2. 6. · Oliver Wyman’s recommendation was for smoothing to use a ratio to cash value, ... testing for actuarial

© 2019 National Association of Insurance Commissioners 1

Date: 2/11/19 VARIABLE ANNUITIES CAPITAL AND RESERVE (E/A) SUBGROUP

Conference Call Wednesday, February 6, 2019

3:00 p.m. ET / 2:00 p.m. CT / 1:00 p.m. MT / 12:00 p.m. PT

ROLL CALL

Peter Weber, Chair Ohio Philip Barlow District of Columbia Nicole Boyd Kansas Fred Andersen Minnesota William Leung Missouri Rhonda Ahrens Nebraska Seong-min Eom New Jersey William Carmello New York Mike Boerner Texas

AGENDA 1. Discuss Risk-Based Capital Instructions—Pete Weber (OH)

a. Outlineb. LR027 Instructionsc. Oliver Wyman Proposed Framework C-3 Recommendations

Attachment 1 Attachment 2 Attachment 3

2. Continue Review of Comments Received on VM-21, Sections 1-5—Pete Weber (OH)a. Organized Commentsb. ACLI Commentsc. Moody’s Commentsd. California’s Proposed Edits

Attachment 4 Attachment 5 Attachment 6 Attachment 7

3. Discuss Any Other Matters Brought Before the Subgroup—Pete Weber (OH)

4. Adjournment

W:\QA\RBC\LRBC\VACRSG\2019\Calls and Meetings\2_6_2019 Call\2-6-19 VACRSG Agenda.docx

1

This page intentionally left blank.

2

John Bruins (via email)

Based on the Joint WG call of Jan. 23:

1) LR027 instructions have been modified to reflect comments from the call, including:a. I modified Appendix 2 to develop an RBC rather than TAR, and adjusted paragraph C on page 17b. I modified the intro to the Line 37 instructions to provide context for the two sets – 2019 and 2020c. In Step 4, I defined TAR to be the Stochastic Reserve plus the RBC – note that both values are prior to

any Phase-in or smoothing, and do not include the additional standard projection amount in thereserve. Since this amount is used to determine if non-prescribed scenarios are appropriate, it seemsthat the value used should be the stochastic value.

d. I moved the comment about the alternative calculations (C3P1 survey) to follow the line 37 instructions,immediately ahead of the overall reserve requirement cross-check.

2) The following is a listing of the Oliver Wyman recommendations and how they were addressed in the drafta. The proposed formulas are accurately included in the instructions for Line (37) for 2020.b. The requirement to use the same distribution of Scenario Reserves is included – see page 17 item #1.c. The additional standard projection amount is included in the RBC – see formulas on top of page 18.d. The reference to reserve in the RBC is to reported reserve, which would include any voluntary reserves

and is therefore consistent with OW.e. The recommendation for regulator approval to use the Specific Tax Recognition is not included. The

draft includes a documentation requirement.f. The recommendation to disclose the macro tax amount even if doing the specific tax is not included and

can be discussed with the documentation requirements.g. Follows the recommendation to smooth RBC rather than TAR – see section on smoothing.h. Oliver Wyman’s recommendation was for smoothing to use a ratio to cash value, but their redline used

a ratio to reserve. We used reserve since we started with their redline. Should this go back to cashvalue?

i. Oliver Wyman changed the smoothing requirement to apply to companies with a CDHS rather than anycompany as LR027 ha always applied, The recommendation then added a required approval to continuesmoothing if there is a material modification to the CHDS – we removed reference to CDHS to beconsistent with current requirements and therefore the requirement for approval.

Attachment One

3


4

© 1993-2018 National Association of Insurance Commissioners 1 8/17/2018

INTEREST RATE RISK AND MARKET RISK LR027

The following instructions for the Interest Rate Risk and Market Risk will remain effective independent of the status of the sunset provision, Section 8, of Actuarial Guideline XLVIII (AG 48) in a particular state or jurisdiction. This instruction will be considered for change once the amendment referenced in AG 48, Section 8, regarding credit for reinsurance, is adopted by the NAIC.

Basis of Factors

The interest rate risk is the risk of losses due to changes in interest rate levels. The factors chosen represent the surplus necessary to provide for a lack of synchronization of asset and liability cash flows.

The impact of interest rate changes will be greatest on those products where the guarantees are most in favor of the policyholder and where the policyholder is most likely to be responsive to changes in interest rates. Therefore, risk categories vary by withdrawal provision. Factors for each risk category were developed based on the assumption of well- matched asset and liability durations. A loading of 50 percent was then added on to represent the extra risk of less well-matched portfolios. Companies must submit an unqualified actuarial opinion based on asset adequacy testing to be eligible for a credit of one-third of the RBC otherwise needed. The interrogatory on Line (1.1) should be answered Yes if the opinion is unqualified. It should also be answered Yes if the opinion is qualified but the only reason for qualification of the opinion is because of the direction provided in Actuarial Guideline XLVIIIAG 48.

Consideration is needed for products with credited rates tied to an index, as the risk of synchronization of asset and liability cash flows is tied not only to changes in interest rates but also to changes in the underlying index. In particular, equity-indexed products have recently grown in popularity with many new product variations evolving. The same C-3 factors are to be applied for equity-indexed products as for their non-indexed counterparts; i.e., based on guaranteed values ignoring those related to the index.

Cash Flow Modeling for C-3 RBC A company may be required or choose to perform cash flow modeling to determine its C-3 RBC requirement. Because of the widespread use of increasingly well-disciplined scenario testing for actuarial opinions based upon an asset adequacy analysis involving cash flow testing, it was determined that a practical method of measuring the degree of asset/liability mismatch existed. It involves further cash flow modeling. In addition, Ssome companies may choose to or be required to calculate part of the C-3 RBC requirement on Certain Annuities and Single Premium Life Insurance under a method using cash flow testing modeling techniques. Refer to LR049 Exemption Test: Cash Flow Testing for C-3 RBC for determination of exemption from this cash flow testing modeling requirement. Companies are required to calculate the C-3 RBC requirement on Variable Annuities and Similar Products as described in the instructions for line (37).

Factor-Based RBC for Reserves on contracts on Certain Annuities and Single Premium Life Insurance that weare Cash Flow Modeled for Interest Rate RiskTested for Asset Adequacy – Factor-Based RBCLines (2) though (16) include the reserves for contracts that were modeled for interest rate risk following the guidance ofSee Appendix 1 of the instructions for more details. ½ of this factor based amount is used in the floor determined in line (34)

The risk categories are: (a) Low-Risk Category

The basic risk-based capital developed for annuities and life insurance in the low-risk category was based on an assumed asset/liability duration mismatch of 0.125 (i.e., awell- matched portfolio). This durational gap was combined with a possible 4 percent one-year swing in interest rates (the maximum historical interest rate swing 95 percentof the time) to produce a pre-tax factor of 0.006377. For a less well matched portfolio, In addition to the 50 percent loading discussed above, the risk-based capital pre-taxfactor reflecting the 50 percent loading discussed above is 0.009115.

Attachment Two

5


(b) Medium and High-Risk Category The factors for the medium and high-risk categories were determined by measuring the value of the additional risk from the more discretionary withdrawal provisions based on assumptions of policyholder behavior and 1,000 random interest rate scenarios. Supplementary contracts not involving life contingencies and dividend accumulations are included in the medium-risk category due to the historical tendency of these policyholders to be relatively insensitive to interest rate changes.

Additional Component for Callable/Pre-Payable Assets Identify the amount of callable/pre-payable assets (including IOs and similar investments) supporting reserves classified in this section. The C-3 requirement after taxes is 50 percent of the excess, if any, of book/adjusted carrying value above current call price. The calculation is done on an asset-by-asset basis. NOTE: If a company is required to calculate part of the RBC based on cash flow testing for C-3 RBC, the factor based requirements for callable/pre-payable assets used in that testing is zero.

Factor Based RBC for All Other Reserves not included in Reserves that are Cash Flow Modeled for Interest Rate Risk This captures all reserves not included in Reserves on Certain Annuities and Single Premium Life Insurance that were Cash Flow Tested or products included under the “Recommended Approach for Setting Risk-Based Capital Requirements for Variable Annuities and Similar Products.”

The risk categories are:

(a) Low-Risk Category The basic risk-based capital developed for annuities and life insurance in the low-risk category was based on an assumed asset/liability duration mismatch of 0.125 (i.e., a well-matched portfolio). This durational gap was combined with a possible 4 percent one-year swing in interest rates (the maximum historical interest rate swing 95 percent of the time) to produce a pre-tax factor of 0.006377. For a less well-matched portfolio, In addition to the 50 percent loading discussed above, the risk-based capital pre-tax factor reflecting the 50 percent loading discussed above is 0.009115.

(b) Medium and High-Risk Category

The factors for the medium and high-risk categories were determined by measuring the value of the additional risk from the more discretionary withdrawal provisions based on assumptions of policyholder behavior and 1,000 random interest rate scenarios. Supplementary contracts not involving life contingencies and dividend accumulations are included in the medium-risk category due to the historical tendency of these policyholders to be relatively insensitive to interest rate changes.

Additional Component for Callable/Pre-Payable Assets Identify the amount of callable/pre-payable assets (including IOs and similar investments) not reported elsewhere in this schedule. This excludes callable/pre-payable assets supporting Reserves on Certain Annuities and Single Premium Life Insurance that were Cash Flow ModeledTested or supporting the Interest Rate Risk Component for products included under the “Recommended Approach for Setting Risk-Based Capital Requirements for Variable Annuities and Similar Products.” This includes callable/pre-payable assets supporting other reserves and capital and surplus. The C-3 requirement after taxes is 50 percent of the excess, if any, of book/adjusted carrying value above current call price. The calculation is done on an asset-by-asset basis and reported in aggregate.

Cash Flow Testing for C-3 RBC A company may be required or choose to perform cash flow testing to determine its RBC requirement. Because of the widespread use of increasingly well-disciplined scenario testing for actuarial opinions based upon an asset adequacy analysis involving cash flow testing, it was determined that a practical method of measuring the degree of asset/liability mismatch existed. It involves further cash flow testing. See Appendix 1 – Cash Flow Testing for C-3 RBC for details. Specific Instructions for Application of the Formula Lines (2) through (16) These lines deal with Certain Annuities and Single Premium Life Insurance for which reserves were cash flow tested for asset adequacymodeled for RBC. Guaranteed Indexed separate accounts following a Class 1 investment strategy are reported as low-risk Line (2)

Attachment Two

6


The fixed portion of equity-based variable products and should not be included. Guaranteed indexed separate accounts following a Class I investment strategy are reported as low-risk Line 2 and those following a Class II investment strategy are excluded. See Proposed new Risk-Based Capital Method for Separate Accounts that Guarantee an Index, June 2003. Company source records entered in Column (3) of Lines (13), (15) and (16) should be adjusted to a pre-tax basis. Line (17) Should equal the sum of Lines (6) + (11) + (14) + (15). Line (16) is not included in the Line (17) total. Instead, it is included in the Line (32) total. Lines (18) through (31) These lines cover:

(a) The remaining company business that was not cash flow tested for asset adequacymodeled for C-3 RBC (see Appendix 1 for details) excluding products included under the “Recommended Approach for SettingCash Flow Modeling for C-3 Risk-Based Capital Requirements for Variable Annuities and Similar Products” and

(b) Business in companies that did not cash flow test for asset adequacymodel for C-3 RBC. The calculation for risk-based capital should not include unitized separate accounts without guarantees even though they may be included in Item 32 of the Notes to Financial Statements. Separate accounts with guarantees should be included, except for those separate accounts that guarantee an index and follow a Class II investment strategy and certain other guaranteed separate accounts as defined below. Synthetic GICs net of certain credits should be included in this section. The provisions for these credits to C-3 requirements is provided in the Separate Accounts section of the risk-based capital instructions. Experience-rated pension contracts defined below should be excluded from “annuity reserves with fair value adjustment” and “annuity reserves not withdrawable.” All amounts should be reported net of reinsurance, net of policy loans and adjusted for assumed and ceded modified coinsurance. Experience-rated group and individual pension business that meets all of the following four conditions is excluded from C–3 factor-based risk:

(a) General account funded; (b) Reserve interest rate is carried at no greater than 4 percent and/or fund long-term interest guarantee (in excess of a year) does not exceed 4 percent; (c) Experience rating mechanism is immediate participation, retroactive credits, or other technique other than participating dividends; and (d) Either is not subject to discretionary withdrawal or is subject to fair value adjustment, but only if the contractually defined lump sum fair value adjustment reflects portfolio

experience as well as current interest rates and is expected to pass both credit risk and rate risk to the policyholder at withdrawal. (A lump sum settlement based only on changes in prevailing rates does not meet this test. Book value cash out options meet this test as long as the present value of payments using U.S. Treasury spot rates is less than or equal to the lump sum fair value on the valuation date and the policyholder does not have an option to change the payment period once payments begin.)

For companies not exempt from cash flow testing for C-3 RBC, such testing is to include those experience-rated products exempted from the formula factors, but for which cash flow testing is done as a part of the asset adequacy testing. Non-indexed separate account business with guarantees that satisfy both conditions (b) and (d) above is excluded from C–3 factor-based risk. Guaranteed indexed separate account business following a Class I investment strategy is reported on Line (18). Note that in the AAA Report “Proposed New Risk-Based Capital Method for Separate Accounts That Guarantee an Index” (adopted by the NAIC Life Risk-Based Capital Working Group in New York, NY, June 2003), there is a stress test applicable to Class I investment strategies for a company that is not subject to scenario testing requirements. Company source records entered in Column (3) of Lines (30) and (31) should be adjusted to a pre-tax basis. Line (33)

Attachment Two

7


Enter in Column (3) the pre-tax interest rate risk results of cash flow testing per the Appendix 1a methodology. Line (33) should be completed by all companies who do cash flow testing modeling of Certain Annuities and Single Premium Life Insurance for asset adequacy C-3 RBC (see Appendix 1) except those with less than $100 million in admitted assets at year-end, unless the answer to Line (14) or Line (22) of LR049 Exemption Test: Cash Flow Testing for C-3 RBC is “Yes” or if the company chooses to do C-3 RBC cash flow testing on a continuing basis. Once a company chooses to use the C-3 RBC cash flow testing method to calculate RBC it must continue to do so unless regulatory approval from the domiciliary jurisdiction is received to go back to the factor-based method. The interest rate risk component for Variable Annuities and Similar Products should be entered into Line (35). Line (34) If Line (33) is equal to zero, then Line (34) should equal Line (32). Otherwise, Line (34) should equal Line (32) plus Line (33) less Line (16) less Line (17) subject to a minimum of 0.5 times Line (32). Line (35) Enter the interest rate risk component from the Cash Flow Modeling for C-3 RBC Requirements Variable Annuities and Similar Products (see Line (37). The interest rate risk component should be entered on a pre-tax basis using the enacted maximum corporate income tax rate. Line (36) Total interest rate risk. Equals Line (34) plus Line (35). Line (37) Cash Flow Modeling for C-3 RBC Requirements for Variable Annuities and Similar Products: Instructions for 2019: 2019 is a transition year to a new modeling framework. A company must follow one of two options to develop the C-3 RBC amount:

A. If the company has elected to apply the requirements of VM-21 from the 2020 version of the NAIC valuation manual to determine reserves for the Variable Annuities for 12/31/19, the company shall follow the instructions beginning on page 16 labeled “Instructions for 2020 and Later” for determining the RBC on the Variable Annuities and similar contracts, but may not apply the phase in provisions of paragraph E on page 18. Otherwise,

A.B. The company shall follow the nine step process below through page 15. Overview (2019) The amount reported on Line (37) is calculated using a nine-step process. As in Step 3 of the Single Scenario C-3 Measurement Considerations section of Appendix 1a – Cash Flow Testing for C-3 RBC Methodology, existing AVR-related assets should not be included in the initial assets used in the C-3 modeling unless AVR has been excluded from TAC due to its use in the asset adequacy analysis supporting reserves. AVR-related assets may be included with C-3 testing to the extent that the AVR has been used in the cash flow testing and is therefore excluded from TAC, and that portion of the AVR-related assets relates to the business being tested. These assets are available for future credit loss deviations over and above expected credit losses. These deviations are covered by C-1 risk capital. Similarly, future AVR contributions should not be modeled. However, the expected credit losses should be in the C-3 modeling. (Deviations from expected are covered by both the AVR and C-1 risk capital and should not be modeled). IMR assets should be used for C-3 modeling. If negative cash flows are handled by selling assets, then appropriate modeling of contributions to and amortization of the IMR need to be reflected in the modeling.

Attachment Two

8


(1) The first step is determined by applying the methodology described in the report “Recommended Approach for Setting Risk-Based Capital Requirements for Variable Annuities and

Similar Products Presented by the American Academy of Actuaries’ Life Capital Adequacy Subcommittee to the National Association of Insurance Commissioners’ Capital Adequacy Task Force (June 2005)” to calculate the total asset requirement. Although Appendix 2 in the Report notes path dependent models under a different set of initialization parameters might produce scenarios that do not satisfy all the calibration points shown in Table 1, to be in compliance with the requirements in this first step, the actual scenarios used for diversified U.S. equity funds must meet the calibration criteria. The scenarios need not strictly satisfy all calibration points in Table 1 of Appendix 2, but the actuary should be satisfied that any differences do not materially reduce the resulting capital requirements. See the Preamble to the Accounting Practices and Procedures Manual for an explanation of materiality. Include the Tax Adjustment as described in the report using the enacted maximum federal corporate income tax rate. If using the Alternative Method for GMDB Risks, use 1 minus the enacted maximum federal corporate income tax rate in place of the 65% adjustment contained in paragraph 4 (page 55) and the enacted maximum federal corporate income tax rate in place of 35% Income Tax Rate shown in Table 8-9 (page 78). The discount rate in Table 8-9 should also be adjusted for the appropriate enacted maximum federal corporate income tax rate.

(2) The second step is to reduce the amount calculated in (1) above by the interest rate portion of the risk (i.e., only the separate account market risk is included in this step). (3) The third step is to calculate the Standard Scenario Amount. (4) Take the greater of the amounts from steps (2) and (3). (5) Apply the smoothing and transition rules (if applicable) to the amount in step (4). (6) Add the general account interest rate portion of the risk to the amount in step (5). (7) Subtract the reported statutory reserves for the business subject to the Report from the amount calculated in step (6). Floor this amount at $0. (8) Divide the result from step (7) by (1-enacted maximum federal corporate income tax rate) to arrive at a pre-tax amount. (9) Split the result from step (8) into an interest rate risk portion and a market risk portion. Note that the interest rate portion may not equal the interest rate portion of the risk used in

steps (2) and (6) above even after adjusting these to a pre-tax basis. The interest rate portion of the risk should be included in Line (35) and the market risk portion in Line (37). The lines on the alternative calculations page will not be required for 2018. Calculation of the Total Asset Requirement The method of calculating the Total Asset Requirement is explained in detail in the AAA’s June 2005 report, referenced above. In summary, it is as follows: A. Aggregate the results of running stochastic scenarios using prudent best estimate assumptions (the more reliable the underlying data is, the smaller the need for margins for

conservatism) and calibrated fund performance distribution functions. If utilizing prepackaged scenarios as outlined in the American Academy of Actuaries’ report, Construction and Use of Pre-Packaged Scenarios to Support the Determination of Regulatory Risk Based Capital Requirements for Variable Annuities and Similar Products, Jan. 13, 2006, the Enhanced C-3 Phase I Interest Rate Generator should be used in generating any interest rate scenarios or regenerating pre-packaged fund scenarios for funds that include the impact of bond yields. Details concerning the Enhanced C-3 Phase I Interest Rate Generator can be found on the American Academy of Actuaries webpage at the following address http://www.actuary.org/pdf/life/c3supp_jan06.pdf. The Enhanced C-3 Phase 1 Interest Rate Generator with its ability to use the yield curve as of the run date and to regenerate pre-packaged fund returns using interest rate scenarios based on the current yield curve replaces the usage of the March 2005 pre-packaged scenarios.

Attachment Two

9

http://www.actuary.org/pdf/life/c3supp_jan06.pdf


B. Calculate required capital for each scenario by calculating accumulated statutory surplus, including the effect of federal income taxes at the enacted maximum federal corporate income tax rate, for each calendar year-end and its present value. The negative of the lowest of these present values is the asset requirement for that scenario. These values are recorded for each scenario and the scenarios are then sorted on this measure. For this purpose, statutory surplus is modeled as if the statutory reserve were equal to the working reserve.

C. The Total Asset Requirement is set at the 90 Conditional Tail Expectation by taking the average of the worst 10 percent of all the scenarios’ asset requirements (capital plus starting

reserve). Risk-based capital is calculated as the excess of the Total Asset Requirement above the statutory reserves. For products with no guaranteed living benefit, or just a guaranteed death benefit, an alternative method is allowed, as described in the AAA report.

D. Risk-based capital is calculated as the excess of the Total Asset Requirement above the statutory reserves. Except for the effect of the Standard Scenario and the Smoothing and

Transition Rules (see below), this RBC is to be combined with the C-1cs component for covariance purposes. E. A provision for the interest rate risk of the guaranteed fixed fund option, if any, is to be calculated and combined with the current C-3 component of the formula. F. The way grouping (of funds and of contracts), sampling, number of scenarios, and simplification methods are handled is the responsibility of the actuary. However, all these

methods are subject to Actuarial Standards of Practice, supporting documentation and justification. G. Certification of the work done to set the RBC level will be required to be submitted with the RBC filing. Refer to Appendices 10 and 11 of the AAA LCAS C-3 Phase II RBC

Report (June 2005) for further details of the certification requirements. The certification should specify that the actuary is not opining on the adequacy of the company's surplus or its future financial condition. The actuary will also note any material change in the model or assumptions from that used previously and the impact of such changes (excluding changes due to a change in these NAIC instructions). Changes will require regulatory disclosure and may be subject to regulatory review and approval. Additionally, if hedging is reflected in the stochastic modeling, additional certifications are required from an actuary and financial officer of the company.

The certification(s) should be submitted by hard copy with any state requiring an RBC hard copy.

H. An actuarial memorandum should be constructed documenting the methodology and assumptions upon which the required capital is determined. The memorandum should also

include sensitivity tests that the actuary feels appropriate, given the composition of their block of business (i.e., identifying the key assumptions that, if changed, produce the largest changes in the RBC amount). This memorandum will be confidential and available to regulators upon request.

Application of the Tax Adjustment Tax Adjustment: Under the U.S. IRC, the tax reserve is defined. It can never exceed the statutory reserve nor be less than the cash surrender value. If tax reserves assumed in the projection are set equal to Working Reserves and if tax reserves actually exceed Working Reserves at the beginning of the projection, a tax adjustment is required. A tax adjustment is not required in the following situations:

• Tax reserves are projected directly; that is, it is not assumed that projected tax reserves are equal to Working Reserves, whether these are cash values or other approximations. • Tax reserves at the beginning of the projection period are equal to Working Reserves. • Tax reserves at the beginning of the projection period are lower than Working Reserves. This situation is only possible for contracts without cash surrender values and when

these contracts are significant enough to dominate other contracts where tax reserves exceed Working Reserves. In this case the modeled tax results are overstated each year for reserves in the projection, as well as the projected tax results reversed at the time of claim.

Attachment Two

10


If a tax adjustment is required, the Total Asset Requirement (TAR) must be increased on an approximate basis to correct for the understatement of modeled tax expense. The additional taxable income at the time of claim will be realized over the projection and will be measured approximately using the duration to worst, i.e., the duration producing the lowest present value for each scenario. The method of developing the approximate tax adjustment is described below. The increase to TAR may be approximated as the corporate tax rate (i.e., 35 percent) times f times the difference between tax reserves and Working Reserves at the start of the projections. For this calculation, f is calculated as follows: For the scenarios reflected in calculating 90 CTE, the lowest of these present values of accumulated statutory surplus is determined for each calendar year-end and its associated projection duration is tabulated. At each such duration, the ratio of the number of contracts in force (or covered lives for group contracts) to the number of contracts in force (or covered lives) at the start of the modeling projection is calculated. The average ratio is then calculated, over all 90 CTE scenarios, and f is one minus this average ratio. If instead, RBC is determined under the standard scenario method then f is based on the ratio at the worst duration under that scenario. If the Alternative Method is used, f is approximated as 0.5. Calculation of the Standard Scenario Amount Standard Scenario for C-3 Phase II Risk Based Capital (RBC) Determination

I) Overview

A) Application to Determine RBC.

A Standard Scenario Amount shall be determined for all of the contracts under the scope described in the June 2005 report, “Recommended Approach for Setting Risk-Based Capital Requirements for Variable Annuities and Similar Products”. If the Standard Scenario Amount is greater than the Total Asset Requirement less any amount included in the TAR but attributable to and allocated to C-3 (Interest Rate Risk) otherwise determined based on the Report, then the Total Asset Requirement before tax adjustment used to determine C-3 Phase 2 (Market Risk) RBC shall be the Standard Scenario Amount.

The Standard Scenario Amount shall be the sum of the following:

1. For contracts for which RBC is based on the Alternative Methodology applied without a model office using 100 percent of the MGDB mortality table, the Standard Scenario Amount shall be the sum of the total asset requirement before tax adjustment from the Alternative Methodology applied to such contracts.

2. For contracts without guaranteed death benefits for which RBC is based on the Alternative Methodology applied without a model office, the Standard Scenario Amount shall be the sum of the total asset requirements before tax adjustment from the Alternative Methodology applied to such contracts.

3. For contracts under the scope of the Report other than contracts for which paragraphs 1 and 2 apply, the Standard Scenario Amount is determined by use of The Standard Scenario Method described in Section III. The Standard Scenario Method requires a single projection of account values based on specified returns on the assets supporting the account values. On the valuation date an initial drop is applied to the account values based on the supporting assets. Subsequently, account values are projected at the rate earned on supporting assets less a margin. Additionally, the projection includes the cash flows for certain contract provisions, including any guaranteed living and death benefits using the assumptions in Section III. Thus the calculation of the Standard Scenario Amount will reflect the greatest present value of the accumulated projected cost of guaranteed benefits less the accumulated projected revenue produced by the margins in accordance with Subsection III (D).

B) The Standard Scenario Amount under the Standard Scenario Method.

The Standard Scenario Amount for all contracts subject to the Standard Scenario Method is determined as of the valuation date under the Standard Scenario Method described in Section III based on a rate, DR. DR is the annual effective equivalent of the 10-year constant maturity treasury rate reported by the Federal Reserve for the month of valuation plus 50 basis points. However, DR shall not be less than 3 percent or more than 9 percent. If the 10-year constant maturity treasury rate is no longer available, then a substitute rate determined by the National Association of Insurance Commissioners shall be used. The accumulation rate, AR, is the product of DR and one minus the tax rate defined in paragraph III(D)(10).

Attachment Two

11


No modification is allowed from the requirements in Section III unless the Domiciliary Commissioner approves such modification as necessary to produce a reasonable result.

C) Illustrative Application of the Standard Scenario Method to a Projection, Model Office and Contract by Contract.

To provide information on the significance of aggregation, a determination of the Standard Scenario Amount based on paragraphs III(B)(1) and III(B)(2) is required for each contract subject to paragraph I(A)(3). The sum of all such Standard Scenario Amounts is described as row B in Table A. In addition, if the Conditional Tail Expectation Amount in the Report is determined based on a projection of an inforce prior to the statement date and/or by the use of a model office, which is a grouping of contracts into representative cells, then additional determinations of the Standard Scenario Amount shall be performed on the prior inforce and/or model office. The calculations are for illustrative purposes to assist in validating the reasonableness of the projection and or the model office and to determine the significance of aggregation.

Table A identifies the Standard Scenario Amounts required by this section. The Standard Scenario Amounts required are based on how the Conditional Tail Expectation projection or Alternative Methodology is applied. For completeness, the table also includes the Standard Scenario Amount required by paragraph I(A)(3). The amounts in Table A should be included as part of the certifying actuary’s annual supporting memorandum specified in paragraph (H) of the “Calculation of the Total Asset Requirement” section of the RBC instructions.

• Standard Scenario Amounts in rows A and B in Table A are required of all companies subject to paragraph I(A)(3). No additional Standard Scenario Amounts are required if a company’s stochastic or alternative methodology result is calculated on the statement date using individual contracts (i.e., without a model office).

• A company that uses a model office as of the statement date to determine its stochastic or alternative methodology result must provide the Standard Scenario Amount for the model office. This is row C.

• A company that uses an aggregation by duration of contract by contract projection of a prior inforce to determine its stochastic or alternative methodology with result PS and then projects requirements to the statement date with result S must provide the Standard Scenario Amount for the prior inforce, row D.

• A company that uses a model office of a prior inforce to determine its stochastic or alternative methodology requirements with result PM and then projects requirements to the statement date with result S must provide the Standard Scenario Amount for the model office on the prior inforce date, row E.

Attachment Two

12


Table A

Standard Scenario Amounts

Guideline Variations

Validation Measures Model Office

Projection Projection of Prior Inforce

A. Aggregate valuation on the statement date on inforce contracts required in I(A)(3)

None None None

B. Seriatim valuation on the statement date on inforce contracts

None: Compare to A None None

C. Aggregate valuation on the statement date on the model office

If not material to model office validation

A/C compare to 1.00 None

D. Aggregate valuation on a prior inforce date on prior inforce contracts

If not material to projection validation None A/D - S/PS

Compare to 0

E. Aggregate valuation on a prior inforce date of a model office

If not material to model office or projection

validation.

(A/E – S/PM) compare to 0

Modification of the requirements in Section III when applied to a prior inforce or a model office is permitted if such modification facilitates validating the projection of inforce or the model office. All such modifications should be documented. No modification is allowed for row B as of the statement date unless the Domiciliary Commissioner approved such modification as necessary to produce a reasonable result under the corresponding amount in row A.

II) Basic Adjusted Reserve

For purposes of determining the Standard Scenario Amount for Risk-Based Capital, the Basic Adjusted Reserve for a contract shall be the Working Reserve, as described in the Report, as of the valuation date.

III) Standard Scenario Amount - Application of the Standard Scenario Method

A) General

Where not inconsistent with the guidance given here, the process and methods used to determine results under the Standard Scenario Method shall be the same as required in the calculation under the modeling methodology required by the Report. Any additional assumptions needed to apply the Standard Scenario Method to the inforce shall be explicitly documented.

Attachment Two

13


B) Results for the Standard Scenario Method.

The Standard Scenario Amount is equal to (1) + (2) – (3) where:

1) Is the sum of the Basic Adjusted Reserve as described in Section II for all contracts for which the Standard Scenario Amount is being determined,

2) Is zero or if greater the aggregate greatest present value for all contracts measured as of the end of each projection year of the negative of the Accumulated Net Revenue described below using the assumptions described in Subsection III(D) and a discount rate equal to the Accumulation Rate, AR. The Accumulated Net Revenue at the end of a projection year equals (i) + (ii) - (iii) where:

(i) Is the Accumulated Net Revenue at the end of the prior projection year accumulated at the rate AR to the end of the current projection year. The Accumulated Net Revenue at the beginning of the projection (i.e., time 0) is zero.

(ii) Are the margins generated during the projection year on account values as defined in paragraph III(D)(1) multiplied by one minus the tax rate and accumulated at rate AR to the end of current projection year, and

(iii) Are the contract benefits paid in excess of account value applied plus the Individual reinsurance premiums (ceded less assumed) less the Individual reinsurance benefits (ceded less assumed) payable or receivable during the projection year multiplied by one minus the tax rate and accumulated at rate AR to the end of current projection year. Individual reinsurance is defined in paragraph III(D)(2).

3) Is the value of approved hedges and Aggregate reinsurance as described in paragraph III(E)(2). Aggregate reinsurance is defined in paragraph III(D)(2).

C) The actuary shall determine the projected reinsurance premiums and benefits reflecting all treaty limitations and assuming any options in the treaty to the other party are exercised to decrease the value of reinsurance to the reporting company (e.g., options to increase premiums or terminate coverage). The positive value of any reinsurance treaty that is not guaranteed to the insurer or its successor shall be excluded from the value of reinsurance. The commissioner may require the exclusion of any portion of the value of reinsurance if the terms of the reinsurance treaties are too restrictive (e.g., time or amount limits on benefits correlate to the Standard Scenario Method).

D) Assumptions for Paragraph III (B) (2) Margins and Account Values.

1) Margins on Account Values. The bases for return assumptions on assets supporting account values are shown in Table I. The Initial returns shall be applied to the account values assigned to each asset class on the valuation date as immediate drops, resulting in the Account Values at time 0. The "Year 1" and "Year 2+" returns are gross annual effective rates of return and are used (along with other decrements and/or increases) to produce the Account Values as of the end of each projection year. For purposes of this section, money market funds shall be considered part of the Bond class.

The Fixed Fund rate is the greater of the minimum rate guaranteed in the contract or 3.5 percent but not greater than the current rates being credited to Fixed Funds on the valuation date.

Account Values shall be accumulated after the initial drop using the rates from Table I with appropriate reductions applied to the supporting assets. The appropriate reductions for account values supported by assets in the Equity, Bond or Balance Classes are all fund and contract charges according to the provisions of the funds and contracts. The appropriate reduction for Account Values supported by Fixed Funds is zero.

Attachment Two

14


The margins on Account Values are defined as follows:

a) During the Surrender Charge Period:

i. 0.10% of Account Value; plus

ii. The maximum of:

• 0.20% of Account Value; or

• Explicit and optional contract charges for guaranteed living and death benefits.

b) After the Surrender Charge Period:

i. The amount determined in (a) above; plus

ii. The lesser of:

• 0.65% of Account Values; and

• 50% of the excess, if any, of all contract charges over (a) above.

However, on fixed funds after the surrender charge period, a margin of up to the amount in (a) above plus 0.4% may be used.

Table I

Initial Year 1 Year 2+ Equity Class -20% 0% 3% Bond Class 0 0 4.85% Balanced Class -12% 0% 3.74% Fixed Separate Accounts and General Account

Fixed Fund Rate Fixed Fund Rate

2) Reinsurance Credit. Individual reinsurance is defined as reinsurance where the total premiums for and benefits of the reinsurance can be determined by applying the terms of the reinsurance to each contract covered without reference to the premiums or benefits of any other contract covered and summing the results over all contracts covered. Reinsurance that is not Individual reinsurance is Aggregate reinsurance.

Individual reinsurance premiums projected to be payable on ceded risk and receivable on assumed risk shall be included in the subparagraph III(B)(2)(iii). Similarly, Individual reinsurance benefits projected to be receivable on ceded risk and payable on assumed risk shall be included in subparagraph III(B)(2)(iii). No Aggregate reinsurance shall be included in subparagraph III(B)(2)(iii).

Attachment Two

15


3) Lapses, Partial Withdrawals, and Moneyness. Partial withdrawals elected as guaranteed living benefits or required contractually (e.g., a contract operating under an automatic withdrawal provision on the valuation date) are to be included in subparagraph III(B)(2)(iii). No other partial withdrawals, including free partial withdrawals, are to be included. All lapse rates shall be applied as full contract surrenders.

A contract is in the money (ITM) if it includes a guaranteed living benefit and at any time the portion of the future projected account value under the Standard Scenario Method required to obtain the benefit would be less than the value of the guaranteed benefit at the time of exercise or payment. If the projected account value is 90 percent of the value of the guaranteed benefit at the time of exercise or payment, the contract is said to be 10 percent in the money. If the income from applying the projected account value to guaranteed purchase rates exceeds the income from applying the projected benefit base to GMIB purchase rates for the same type of annuity, then there is no GMIB cost and the GMIB is not in the money. A contract not in the money is out of the money (OTM). If a contract has multiple living benefit guarantees then the contract is ITM to the extent that any of the living benefit guarantees are ITM. Lapses shall be at the annual effective rates given in Table II.

Table II – Lapse Assumptions

During Surrender Charge Period

After Surrender Charge Period

Death Benefit Only Contracts

5% 10%

All Guaranteed Living Benefits OTM

5% 10%

ITM< 10% 10% ≤ ITM< 20% 20% ≤ ITM Any Guaranteed Account Balance Benefits ITM

0% 0% 0% 0%

Any Other Guaranteed Living Benefits ITM

3% 7% 5% 2%

4) Account Transfers and Future Deposits. No transfers between funds shall be assumed to determine the greatest present value amount required under paragraph III(B)(2)

unless required by the contract (e.g., transfers from a dollar cost averaging fund or contractual rights given to the insurer to implement a contractually specified portfolio insurance management strategy or a contract operating under an automatic re-balancing option). When transfers must be modeled, to the extent not inconsistent with contract language, the allocation of transfers to funds must be in proportion to the contract's current allocation to funds.

Margins generated during a projection year on funds supporting account values are transferred to the Accumulation of Net Revenue at year-end and are subsequently accumulated at the Accumulation Rate. Assets for each class supporting account values are to be reduced in proportion to the amount held in each asset class at the time of transfer of margins or any portion of Account Value applied to the payment of benefits.

No future deposits shall be assumed unless required by the terms of the contract to prevent contract or guaranteed benefit lapse, in which case they must be modeled. When future deposits must be modeled, to the extent not inconsistent with contract language, the allocation of the deposit to funds must be in proportion to the contract’s current allocation to funds.

5) Mortality. Mortality at 80 percent of the 1994 MGDB tables through age 95 increasing by 1 percent each year to 100 percent of the 1994 MGDB table at age 115 shall be assumed in the projection used to the determine the greatest present value amount required under paragraph III(B)(2).

Attachment Two

16


6) Projection Frequency. The projection used to determine the greatest present value amount required under paragraph III(B)(2) shall be calculated using an annual or more frequent time step, such as quarterly. For time steps more frequent than annual, assets supporting Account Values at the start of each projection year may be retained in such funds until year-end (i.e., pre-tax margin earned during the year will earn the fund rates instead of the Discount Rate until year-end) or removed after each time step. However, the same approach shall be applied for all years. Subsequent to each projection year-end, Accumulated Net Revenues for the year shall earn the Accumulation Rate. Similarly, projected benefits, lapses, elections and other contract activity can be assumed to occur annually or at the end of each time step, but the approach shall be consistent for all years.

7) Surrender Charge Period. If the surrender charge for the contract is determined based on individual contributions or deposits to the contracts, the surrender charge

amortization period may be estimated for projection purposes. Such estimated period shall not be less than the remaining duration based on the normal amortization pattern for the remaining total contract charge assuming it resulted from a single deposit, plus one year.

8) Contract Holder Election Rates. Contract holder election rates to determine amounts in subparagraph III(B)(2)(iii) shall be 15 percent per annum for any elective ITM benefit except guaranteed withdrawal benefits, but only to the extent such election does not terminate a more valuable benefit subject to election. Guaranteed Minimum Death Benefits are not benefits subject to election. Exception: Contract holder election rates shall be 100 percent at the last opportunity to elect an ITM benefit, but only to the extent such election does not terminate a more valuable benefit subject to election. A benefit is more valuable if it is more ITM in absolute dollars using the definition of ITM in paragraph III(D)(3).

For guaranteed minimum withdrawal benefits, a partial withdrawal equal to the applicable percentage in Table III applied to the contract’s maximum allowable partial withdrawal shall be assumed in subparagraph III(B)(2)(iii). However, if the contract’s minimum allowable partial withdrawal exceeds the partial withdrawal from applying the rate in Table III to the contract’s maximum allowable partial withdrawal, then the contract’s minimum allowable partial withdrawal shall be assumed in subparagraph III(B)(2)(iii).

Table III – Guaranteed Withdrawal Assumptions

Attained Age Less than 50

Attained Age 50 to 59

Attained Age 60 or Greater

Withdrawals do not reduce other elective Guarantees that are in the money

50% 75% 100%

Withdrawals reduce elective Guarantees that are in the money

25% 50% 75%

9) GMIBs. For subparagraph III(B)(2)(iii), GMIB cost at the time of election shall be the excess, if positive, of the reserve required for the projected annuitization stream over the available account value. If the reserve required is less than the account value, the GMIB cost shall be zero. The reserve required shall be determined using the Annuity 2000 Mortality Table and a valuation interest rate equal to the Discount Rate. If more than one annuity option is available, chose the option with a reserve closest to the reserve for a life annuity with 10 years of certain payments.

10) Indices. If an interest index is required to determine projected benefits or reinsurance obligations, the index must assume interest rates have not changed since the last reported rates before the valuation date. If an equity index is required, the index shall be consistent with the last reported index before the valuation date, the initial drop in equity returns and the subsequent equity returns in the standard scenario projection up to the time the index is used. The sources of information and how the information is used to determine indexes shall be documented and, to the extent possible, consistent from year to year.

11) Taxes. All taxes shall be based on the enacted maximum federal corporate income tax rate.

E) Assumptions for use in paragraph III (B) (3).

Attachment Two

17


1) The Value of Aggregate Reinsurance. The value of Aggregate reinsurance is the discounted value, at rate AR of the excess of: a) the benefit payments from the reinsurance, over b) the reinsurance premiums, where (a) and (b) are determined under the assumptions described in Subsection III(D).

2) The Value of Approved Hedges. The value of approved hedges shall be calculated separately from the calculation in paragraph III(B)(2). The value of approved hedges is the difference between: a) the discounted value at rate AR of the after-tax cash flows from the approved hedges; less b) their statement values on the valuation date.

To be an approved hedge, a derivative or other investment has to be an actual asset held on the valuation date, be designated as a hedge for one or more contracts subject to the Standard Scenario, and be part of a clearly defined hedging strategy as described in the Report. If the approved hedge also supports contracts not subject to the Standard Scenario, then only that portion of the hedge designated for contracts subject to the Standard Scenario shall be included in the value of approved hedges. Approved hedges must be held in accordance with an investment policy that has been implemented for at least six months and has been approved by the Board of Directors or a subcommittee of Board members. A copy of the investment policy and the resolution approving the policy shall be maintained with the documentation of the Standard Scenario and available on request. Approved hedges must be held in accordance with a written investment strategy developed by management to implement the Board’s investment policy. A copy of the investment strategy on the valuation date, the most recent investment strategy presented to the Board if different and the most recent written report on the effectiveness of the strategy shall be maintained with the documentation of the Standard Scenario and available on request. The commissioner may require the exclusion of any portion of the value of approved hedges upon a finding that the company’s documentation, controls, measurement, execution of strategy or historical results are not adequate to support a future expectation of risk reduction commensurate with the value of approved hedges. The item being hedged, the contract guarantees, and the approved hedges are assumed to be accounted for at the average present value of the tail scenarios. The value of approved hedges for the standard scenario is the difference between an estimate of this “tail value” and the “fair value” of approved hedges. For this valuation to be consistent with the statement value of approved hedges, the statement value of approved hedges will need to be held at fair value with the immediate recognition of gains and losses. Accordingly, it is assumed that approved hedges are not subject to the IMR or the equity component of the AVR. Approved hedges need not satisfy SSAP No. 86. In particular, as gains and losses of approved hedges are recognized immediately, approved hedges need not satisfy the requirements for hedge accounting of fair value hedges. It is the combination of hedges and liabilities that determine which scenarios are the tail scenarios. In particular, scenarios where the hedging is least effective are likely to be tail scenarios and liabilities that are a left tail risk could in combination with hedges become a right tail risk. The cash flow projection for approved hedges that expire in less than one year from the valuation date should be based on holding the hedges to their expiration. For hedges with an expiration of more than one year, the value of hedges should be based on liquidation of the hedges one year from the valuation date. Where applicable, the liquidation value of hedges shall be consistent with Black-Scholes pricing, a risk free rate of DR, annual volatility implicit as of the valuation date in the statement value of the hedges under Black-Scholes pricing and a risk free rate of DR and the assumed returns in the Standard Scenario from the valuation date to the date of liquidation. There is no credit in the Standard Scenario for dynamic hedging beyond the credit that results from hedges actually held on the valuation date. There is no credit for hedges actually held on the valuation date that are not approved hedges as the commitment to maintain the level of risk reduction derived from such hedges is not adequate.

3) Retention of Components. For the Standard Scenario Amounts on the statement date the company should have available to the Commissioner the following values:

a) For runs A and B as defined in I(C) by contract and in aggregate the amounts determined in III(B)(1) and III(B)(2).

b) For run A the aggregate amounts determined in III(E)(1) and III(E)(2).

Smoothing and Transition Rules

Attachment Two

18


If a company is following a Clearly Defined Hedging Strategy (See “Recommended Approach for Setting Risk-Based Capital Requirements for Variable Annuities and Similar Products” presented by the American Academy of Actuaries’ Life Capital Adequacy Subcommittee to the National Association of Insurance Commissioner’s Capital Adequacy Task Force (June 2005) for the definition of this phrase) on some or all of its business, a decision should be made whether or not to smooth the TAR. In all cases where ‘cash value’ is to be used, the values used must be computed on a consistent basis for each block of business at successive year-ends. For deferred annuities with a cash value option, direct writers will use the cash value. For deferred annuities with no cash value option, or for reinsurance assumed through a treaty other than coinsurance, use the policyholder account value of the underlying contract. For payout annuities, or other annuities with no account value or cash value, use the amount as defined for variable payout annuities in the definition of Working Reserve. For any business reinsured under a coinsurance agreement that complies with all applicable reinsurance reserve credit “transfer of risk” requirements, the ceding company shall reduce the value in proportion to the business ceded while the assuming company shall use an amount consistent with the business assumed. A company who reported an amount in Line (37) last year may choose to smooth the Total Asset Requirement. A company is required to get approval from its domestic regulator prior to changing its decision about smoothing from the prior year. To implement smoothing, use the following steps. If a company does not qualify to smooth or a decision has been made not to smooth, go to the step “Reduction for Reported Statutory Reserves.” Instructions – 2007 and Later

1. Determine the Total Asset Requirement as the greater of that produced by the “Recommended Approach for Setting Risk-Based Capital Requirements for Variable Annuities and Similar Products” presented by the American Academy of Actuaries’ Life Capital Adequacy Subcommittee to the National Association of Insurance Commissioner’s Capital Adequacy Task Force (June 2005) or the value produced by the “Standard Scenario” as outlined above.

2. Determine the aggregate cash value for the contracts covered by the Stochastic modeling requirements. 3. Determine the ratio of TAR / CV for current year. 4. Determine the Total Asset Requirement as actually reported for the prior year Line (37). 5. Determine the aggregate cash value for the same contracts for the prior year-end. 6. Determine the ratio of TAR / CV for prior year. 7. Determine a ratio as 0.4*(6) plus 0.6*(3) {40% prior year ratio and 60% current year ratio}. 8. Determine TAR for current year as the product of (7) and (2) {adjust (2) to be actual 12/31 cash value}.

Reduction for Reported Statutory Reserves The amount of the TAR (post-Federal Income Tax) determined using the instructions for the applicable year is reduced by the reserve, net of reinsurance, for the business subject to this instruction reported in the current statutory annual statement. Allocation of Results to Line (35) and Line (37) See step (9) located in the overview section at the beginning of the instructions for this line.

Attachment Two

19


Instructions for 2020 & Later {Drafting Note: in the material that follows, Oliver Wyman’s proposed instructions are modified to present a more understandable requirement, but the only changes to actual requirements are for: F. Alternative Methodology, G. Phase-in, and I. Format of documentation. } Line (37) Cash Flow Modeling for RBC Requirements for Variable Annuities and Similar Products: Instructions for 2020 & Later {Drafting Note: in the material that follows, Oliver Wyman’s proposed instructions are modified to present a more understandable requirement, but the only changes to actual requirements are for: CF. Alternative Methodology, EG. Phase-in, F. Smoothing, and I. Format of documentation. } Overview Cash Flow Modeling for RBC Requirements for Variable Annuities and Similar Products: The amount reported on Line (37) is calculated using the 7-step process defined F below. This calculation applies to all policies and contracts that have been valued following the requirements of AG-43 or VM-21. For contracts whose reserve was determined using the Alternative Methodology (VM-21 Section 76) see step 3 while all other contracts follow steps 1 and 2, then all contracts follow steps 4 - 7. Step 1 CTE98: The first step is to determine CTE98by applying the one of the two methodologies described in Paragraph A below. Step 2 RBC: using the formulas in paragraph B, determine the RBC amount based on the amount calculated in step (A1). Floor this amount at $0. Step 3 Determine the RBC using the Alternative Methodology for any business subject to that requirements as described in Paragraph C. Step 4 As described in Paragraph D below, the RBC amount is the sum of the amounts determined in steps 2B and 3C above, but not less than zero. The Total Asset Requirement is the Reported Reserve based on the requirements of VM-21 prior to the application of any phase-in, plus the RBC amount. Step 5: For a company that has elected a Phase-in for reserves following VM-21 Section 2.B., the RBC amount is to be phased-in over the same time period following the requirements

in paragraph (E) below. Step 6 Apply the smoothing and transition rules (if applicable) to the RBC amount in step (4) or (5) as applicable. Step 7 Split the result from step 4, 5, or 6 (as appropriate) into an interest rate risk portion and a market risk portion as described in paragraph E. The interest rate portion of the risk should be included in Line (35) and the market risk portion, after applying smoothing and transition rules (if applicable), in Line (37). The lines on the alternative calculations page will not be required for 2016 or later.

Attachment Two

20


The RBC is calculated as follows: A. CTE (98) is calculated as follows: Except for policies and contracts subject to the Alternative Methodology (See E. below), apply the CTE methodology described in NAIC Valuation Manual VM-21 and calculate the CTE (98) as the numerical average of the 2 percent largest values of the Scenario Reserves, as defined by Section 4 of VM-21. In performing this calculation, the process and methods used to calculate the Scenario Reserves use the requirements of VM-21 and should be the same as used for the reserve calculations. The effect of Federal Income Tax should be handled following one of the following two methods

1. If using the Macro Tax Adjustment (MTA): The modeled cash flows will ignore the effect of Federal Income Tax. As a result, for each individual scenario, the numerical value of the Scenario Greatest Present Value used in this calculation should be identical to that for the same scenario in the Aggregate Reserve calculation under VM-21. Federal Income Tax is reflected later in the formula in paragraph B.1.

2. If using Specific Tax Recognition (STR): At the option of the company, CTE After-Tax (98) (CTEAT (98)) may be calculated using an

approach in which the effect of Federal Income Tax is reflected in the projection of Accumulated Deficiencies, as defined in Section 4.A. of VM-21, when calculating the Scenario Reserve for each scenario. To reflect the effect of Federal Income Tax, the company should find a reasonable and consistent basis for approximating the evolution of tax reserves in the projection, taking into account restrictions around the size of the tax reserves (e.g., that tax reserve must equal or exceed the cash surrender value for a given contract). The Accumulated Deficiency at the end of each projection year should also be discounted at a rate that reflects the projected after-tax discount rates in that year. In addition, the company should add the Tax Adjustment as described below to the calculated CTEAT (98) value.

3. A company that has elected to calculate CTEAT (98) using STR may not switch back to using MTA in the projection of Accumulated Deficiencies

without prominently disclosing that change in the certification and supporting memorandum. The company should also disclose the methodology adopted, and the rationale for its adoption, in the documentation required by Paragraph J below.

4. Application of the Tax Adjustment: Under the U.S. IRC, the tax reserve is defined. It can never exceed the statutory reserve nor be less than the

cash surrender value. If a company is using STR and if the company’s actual tax reserves exceed the projected tax reserves at the beginning of the projection, a tax adjustment is required.

The CTEAT (98) must be increased on an approximate basis to correct for the understatement of modeled tax expense. The additional taxable income at the time of claim will be realized over the projection and will be measured approximately using the duration to worst, i.e., the duration producing the lowest present value for each scenario. The method of developing the approximate tax adjustment is described below. The increase to CTEAT (98) may be approximated as the corporate tax rate times f times the difference between the company’s actual tax reserves and projected tax reserves at the start of the projections. For this calculation, f is calculated as follows: For the scenarios reflected in calculating CTE (98), the Scenario Greatest Present Value is determined and its associated projection duration is tabulated. At each such duration, the ratio of the number of contracts in force (or covered lives for group contracts) to the number of contracts in force (or covered lives) at the start of the modeling projection is calculated. The average ratio is then calculated over all CTE (98) scenarios and f is one minus this average ratio. If the Alternative Method is used, f is approximated as 0.5.

B. Determination of RBC amount using stochastic modeling:

Attachment Two

21


1. If using the MTA: Calculate the RBC Requirement by the following formula in which the statutory reserve is the actual reserve reported in the Annual Statement. in the second term – i.e., the difference between statutory reserves and tax reserves multiplied by the Federal Income Tax Rate – may not exceed the portion of the company’s non-admitted deferred tax assets attributable to the same portfolio of contracts to which VM-21 is applied in calculating statutory reserves:

25% × �(CTE (98) + Additional Standard Projection Amount − Statutory Reserve) × (1 − Federal Income Tax Rate)− (Statutory Reserve − Tax Reserve) × Federal Income Tax Rate�

2. If the company elects to use the STR: the RBC is determined by the following formula:

25% × (CTEAT (98) + Additional Standard Projection Amount × (1 − Federal Income Tax Rate) − Statutory Reserve) The Additional Standard Projection Amount is calculated using the methodology outlined in Section 6 of VM-21.

C. Determination of RBC using Alternative Methodology: This calculation applies to all policies and contracts that have been valued following the requirements of AG-43 or VM-21, for which the reserve was determined using the Alternative Methodology (VM-21 Section 76). The RBC amount is determined by applying the methodology as defined in Appendix 2 to these instructions to calculate the Total Asset Requirement for this block of business, then subtract the reserve reported for this block. D. The C-3 RBC amount is the sum of the amounts determined in steps paragraphs (B) and (C) above, but not less than zero. The Total Asset Requirement is defined as the Stochastic Reserve determined from VM-21 Section 4 plus the C-3 RBC amount. All values are prior to any consideration of Phase-in allowances for either reserve or RBC, or any RBC smoothing allowance. E. Phase in: A company that has elected to phase-in the effect of the new reserve requirements following VM-21 Section 2.B. shall phase in the effect on RBC over the same time period, using the following steps:

- 1. Begin with the C-3 RBC amount from step 8 for Dec. 31, 2019 LR027 Line (37) instructions for all business within the scope of the Variable Annuities modeling requirements as of 12/31/19. Add to this the amount of RBC computed in the same manner as the 2019 value for any reinsurance ceded that is expected to be recaptured in 2020 and in the scope of the Variable Annuities modeling requirements. This amount is 2019RBC

- 2. Determine the RBC amount as of 12/31/19 using steps A, B, C, and D for the same inforce business as in 1. Labeled as 2019RBC New - Determine PIA as the excess of (2019RBC New) over (2019RBC) - For 12/31/2020, compute the RBC following paragraphs A – D above, then subtract PIA times (2/3) - For 12/31/2021, compute the RBC following paragraphs A – D above, then subtract PIA times (1/3)

Guidance Note: For a company that has adopted a Phase-in for reserves longer than 3 years, adjust the above formula to reflect the actual period with uniform amortization amounts

during that period. F. Smoothing of C-3 RBC amount

A company should decide whether or not to smooth the C-3 RBC calculated in step (D) or (E) above to determine the amount in Line (37). . For any business reinsured under a coinsurance agreement that complies with all applicable reinsurance reserve credit “transfer of risk” requirements, the ceding company shall reduce the reserve in proportion to the business ceded while the assuming company shall use a reserve consistent with the business assumed.

Attachment Two

22


A company who reported an amount in Line (37) last year may choose to smooth the risk-based capital calculated in step (D) or (E) above. A company is required to get approval from its domestic regulator prior to changing its decision about smoothing from the prior year. In addition, a company that has elected to smooth the risk-based capital is required to get approval from its domestic regulator prior to smoothing if it has experienced a material change in its Clearly Defined Hedging Strategy from the prior year. For this purpose, a company’s Clearly Defined Hedging Strategy is considered to have experienced a material change if any of the items outlined in VM-21 Sections 9in the current year differs from that in the prior year. To implement smoothing, use the following steps. If a company does not qualify to smooth or a decision has been made not to smooth, go to the step (G)

1. Determine the risk-based capital amount calculated in step (D) or (E) above 2. Determine the aggregate reserve for the contracts covered by the Variable Annuity Stochastic modeling requirements. 3. Determine the ratio of the risk-based capital / reserve for current year. 4. Determine the risk-based capital as actually reported for the prior year Lines (35) plus (37). 5. Determine the aggregate reserve for the contracts in scope of these requirements for the prior year-end. 6. Determine the ratio of the risk-based capital / reserve for prior year. 7. Determine a ratio as 0.4*(6) plus 0.6*(3) {40% prior year ratio and 60% current year ratio}. 8. Determine the risk-based capital for current year as the product of (7) and (2) {adjust (2) to be actual 12/31 reserve}.

G. The amount determined in D., E., or F.. for the contracts shall be split into a component for interest rate risk and a component for market risk. Neither component may be less than

zero. The provision for the interest rate risk, if any, is to be reported in Line (35). The market risk component is reported in line (37).

This RBC is to be combined with the C-1cs component for covariance purposes. H. The way grouping (of funds and of contracts), sampling, number of scenarios, and simplification methods are handled is the responsibility of the company. However, all these

methods are subject to Actuarial Standards of Practice, supporting documentation and justification, and should be identical to those used in calculating the company’s statutory reserves following VM-21.

I. Certification of the work done to set the RBC amount for Variable Annuities and Similar products are the same as are required for reserves as part of VM-31. . The certification

should specify that the actuary is not opining on the adequacy of the company's surplus or its future financial condition.

The certification(s) should be submitted by hard copy with any state requiring an RBC hard copy. J. An actuarial memorandum should be constructed documenting the methodology and assumptions upon which the required capital for the variable annuities and similar products

is determined. The memorandum should be developed based on the requirements of VM-31. At the company’s option, the documentation of reserves and RBC may be merged into a single Actuarial Memorandum with any differences for RBC discussed in a separate section of the Memorandum. These differences will typically include the basis for considering federal income tax, whether or not smoothing was applied and whether or not a phase in was considered. This memorandum will be confidential and available to regulators upon request.

If the company elects to calculate CTEAT (98) using STR whereby the effect of Federal Income Tax is reflected in the projection of Accumulated Deficiencies, the company

should still disclose in the memorandum the Total Asset Requirement and C-3 risk-based capital that would be obtained if the company had elected to use the NTA method.

Attachment Two

23


The lines on the alternative calculations page will not be required for 2016 or later. The total of all annual statement reserves representing exposure to C–3 risk on Line (36) should equal the following: Exhibit 5, Column 2, Line 0199999 – Page 2, Column 3, Line 6 + Exhibit 5, Column 2, Line 0299999 + Exhibit 5, Column 2, Line 0399999 + Exhibit 7, Column 1, Line 14 + Separate Accounts Page 3, Column 3, Line 1 plus Line 2 after deducting (a) funds in unitized separate accounts with no underlying guaranteed minimum return and no

unreinsured guaranteed living benefits; (b) non-indexed separate accounts that are not cash flow tested with guarantees less than 4 percent; (c) non-cash-flow-tested experience rated pension reserves/liabilities; and (d) guaranteed indexed separate accounts using a Class II investment strategy.

– Non policyholder reserves reported on Exhibit 7 + Exhibit 5, Column 2, Line 0799997 + Schedule S, Part 1, Section 1, Column 12 – Schedule S, Part 3, Section 1, Column 14

Attachment Two

24


Appendix 1 – Cash Flow Testing Modeling for C-3 RBC

This appendix is applicable for all companies who do Cash Flow Testing for C-3 RBC for Certain Annuities and Single Premium Life products. The method of developing the C-3 component for these contracts is building on the work of the asset adequacy modeling, but using interest scenarios designed to help approximate the 95th percentile C- 3 risk. The C-3 component is to be calculated as the sum of four amounts, but subject to a minimum. The calculation is: (a) For Certain Annuities or Single Premium Life Insurance products other than equity-indexed products, whether written directly or assumed through reinsurance, that the company

tests for asset adequacy analysis using cash flow testing, an actuary should calculate the C-3 requirement based on the same cash flow models and assumptions used and same “as-of” date as for asset adequacy, but with a different set of interest scenarios and a different measurement of results. A weighted average of a subset of the scenario-specific results is used to determine the C-3 requirement. The result is to be divided by (1-enacted maximum federal corporate income tax rate) to put it on a pre-tax basis for LR027 Interest Rate Risk and Market Risk Column (2) Line (33).

If the “as-of” date of this testing is not Dec. 31, the ratio of the C-3 requirement to reserves on the “as-of” date is applied to the year-end reserves, similarly grouped, to determine the year-end C-3 requirement for this category.

(b) Equity-indexed products are to use the existing C-3 RBC factors, not the results of cash flow testing. (c) For all other products (either non-cash-flow-tested or those outside the product scope defined above) the C-3 requirements are calculated using current existing C-3 RBC factors

and instructions. (d) For callable/pre-payable assets (including IOs and similar investments other than those used for testing in component a) above, the after-tax C-3 requirement is 50.076.9 percent of

the excess, if any, of book/adjusted carrying value above current call price. The calculation is to be done on an asset-by-asset basis. For callable/pre-payable assets used for testing in component a) above as well as those used in C-3P2 testing, the C-3 factor requirement is zero.

The total C-3 component is the sum of (a), (b), (c) and (d), but not less than half the C-3 component based on current factors and instructions. • For this C-3 calculation, “Certain Annuities” means products with the characteristics of deferred and immediate annuities, structured settlements, guaranteed separate accounts

(excluding guaranteed indexed separate accounts following a Class II investment strategy) and GICs (including synthetic GICs and funding agreements). Debt incurred for funding an investment account is included if cash flow testing of the arrangement is required by the insurer’s state of domicile for asset adequacy analysis. The equity-based portions of variable products are not to be included, but guaranteed fixed options within such products are. See Appendix 1b for further discussion.

• The company may use either a standard 50 scenario set of interest rates or an alternative, but more conservative, 12 scenario set (for part a, above). It may use the smaller set for

some products and the larger one for others. Details of the cash flow testing for C-3 RBC methodology are contained in Appendix 1a.

Attachment Two

25


• In order to allow time for the additional work effort, an estimated value is permitted for the year-end annual statement. For the RBC electronic filing, the actual results of the cash flow testing for C-3 RBC will be required. If the actual RBC value exceeds that estimated earlier in the blanks filing by more than 5 percent, or if the actual value triggers regulatory action, a revised filing with the NAIC and the state of domicile is required by June 15; otherwise, re-filing is permitted but not required.

• The risk-based capital submission is to be accompanied by a statement from the appointed actuary certifying that in his or her opinion the assumptions used for these calculations

are not unreasonable for the products, scenarios and purpose being tested. This C-3 Assumption Statement is required from the appointed actuary even if the cash flow testing for C-3 RBC is done by a different actuary.

• The cash flow testing used for this purpose will use assumptions as to cash flows, assets associated with tested liabilities, future investment strategy, rate spreads, “as-of” date and how negative cash flow is reflected consistent with those used for cash flow testing for asset adequacy purposes (except that if negative cash flow is modeled by borrowing, the actuary needs to make sure that the amount and cost of borrowing are reasonable for that particular scenario of the C-3 testing). The other differences are the interest scenarios assumptions and how the results are used.

It is important that assumptions be reviewed for reasonableness under the severe scenarios used for C-3 RBC cash flow testing. The assumptions used for cash flow testing may need to be modified so as to produce reasonable results in severe scenarios.

• The actuary must also ensure that the cash flow testing used for the 50 or 12 scenarios does not double-count cash flow offsets to the interest rate risks. That is, that the calculations

do not reduce C-3 and another RBC component for the same margins. For example, certain reserve margins on some guaranteed separate account products serve an AVR role and are credited against the C-1o requirement. To that degree, these margins should be removed from the reserve used for C-3 RBC cash flow testing.

Attachment Two

26


Appendix 1a – Cash Flow Testing Modeling for C-3 RBC Methodology General Approach 1. The underlying asset and liability model(s) are those used for year-end Asset Adequacy Analysis cash flow testing, or a consistent model. 2. Run the scenarios (12 or 50) produced from the interest-rate scenario generator. 3. The statutory capital and surplus position, S(t), should be captured for every scenario for each calendar year-end of the testing horizon. The capital and surplus position is equal to

statutory assets less statutory liabilities for the portfolio. 4. For each scenario, the C-3 measure is the most negative of the series of present values S(t)*pv(t), where pv(t) is the accumulated discount factor for t years using 105 percent of the

after-tax one-year Treasury rates for that scenario. In other words:

∏ +=t

titpv1

1/(1)( )

5. Rank the scenario-specific C-3 measures in descending order, with scenario number 1’s measure being the positive capital amount needed to equal the very worst present value

measure. 6. Taking the weighted average of a subset of the scenario specific C-3 scores derives the final C-3 after-tax factor.

(a) For the 50 scenario set, the C-3 scores are multiplied by the following series of weights:

------------------------------------- Weighting Table -----------------------------------------

Scenario Rank: 17 16 15 14 13 12 11 10 9 8 7 6 5 Weight: 0.02 0.04 0.06 0.08 0.10 0.12 0.16 0.12 0.10 0.08 0.06 0.04 0.02

The sum of these products is the C-3 charge for the product.

(b) For the 12 scenario set, the charge is calculated as the average of the C-3 scores ranked 2 and 3, but cannot be less than half the worst scenario score.

7. If multiple asset/liability portfolios are tested and aggregated, an aggregate C-3 charge can be derived by first summing the S(t)'s from all the portfolios (by scenario) and then

following Steps 2 through 6 above. An alternative method is to calculate the C-3 score by scenario for each product, sum them by scenario, then order them by rank and apply the above weights.

Attachment Two

27


Single Scenario C-3 Measurement Considerations 1. GENERAL METHOD - This approach incorporates interim values, consistent with the approach used for bond, mortgage and mortality RBC factor quantification. The approach

establishes the risk measure in terms of an absolute level of risk (e.g., solvency) rather than volatility around an expected level of risk. It also recognizes reserve conservatism, to the degree that such conservatism hasn’t been used elsewhere.

2. INITIAL ASSETS = RESERVES - Consistent with appointed actuary practice, the cash flow models are run with initial assets equal to reserves; that is, no surplus assets are used. 3. AVR - Existing AVR-related assets should not be included in the initial assets used in the C-3 modeling. These assets are available for future credit loss deviations over and above

expected credit losses. These deviations are covered by C-1 risk capital. Similarly, future AVR contributions should not be modeled. However, the expected credit losses should be in the cash flow modeling. (Deviations from expected are covered by both the AVR and the C-1 risk capital.)

4. IMR - IMR assets should be used for C-3 modeling. (Also see #9 – Disinvestment Strategy.) 5. INTERIM MEASURE - Retained statutory surplus (i.e., statutory assets less statutory liabilities) is used as the year-to-year interim measure. 6. TESTING HORIZONS - Surplus adequacy should be tested over a period that extends to a point at which contributions to surplus on a closed block are immaterial in relationship

to the analysis. If some products are being cash flow tested for Asset Adequacy Analysis over a longer period than the 30 years generated by the interest-rate scenario generator, the scenario rates should be held constant at the year 30 level for all future years. A consistent testing horizon is important for all lines if the C-3 results from different lines of business are aggregated.

7. TAX TREATMENT - The tax treatment should be consistent with that used in Asset Adequacy Analysis. Appropriate disclosure of tax assumptions may be required.

8. REINVESTMENT STRATEGY - The reinvestment strategy should be that used in Asset Adequacy Analysis modeling. 9. DISINVESTMENT STRATEGY - In general, negative cash flows should be handled just as they are in the Asset Adequacy Analysis. The one caveat is, since the RBC scenarios

are more severe, models that depend on borrowing need to be reviewed to be confident that loans in the necessary volume are likely to be available under these circumstances at a rate consistent with the model’s assumptions. If not, adjustments need to be made.

If negative cash flows are handled by selling assets, then appropriate modeling of contributions and withdrawals to the IMR need to be reflected in the modeling.

10. STATUTORY PROFITS RETAINED - The measure is based on a profits retained model, anticipating that statutory net income earned one period is retained to support capital requirements in future periods. In other words, no stockholder dividends are withdrawn, but policyholder dividends, excess interest, declared rates, etc., are modeled realistically and assumed, paid or credited.

11. LIABILITY and ASSET ASSUMPTIONS - The liability and asset assumptions should be those used in Asset Adequacy Analysis modeling. Disclosure of these assumptions may

be required. 12. SENSITIVITY TESTING - Key assumptions shall be stress tested (e.g., lapses increased by 50 percent) to evaluate sensitivity of the resulting C-3 requirement to the various

assumptions made by the actuary. Disclosure of these results may be required.

Attachment Two

28


Appendix 1b - Frequently Asked Questions for Cash Flow Testing Modeling for C-3 RBC

1. Where can the scenario generator be found? What is needed to run it?

The scenario generator is a Microsoft Excel spreadsheet. By entering the Treasury yield curve at the date for which the testing is done, it will generate the sets of 50 or 12 scenarios. It requires Windows 95 or higher. This spreadsheet and instructions are available on the NAIC Web site at (http://www.naic.org/cmte_e_lrbc.htm). It is also available on diskette from the American Academy of Actuaries.

2. The results may include sensitive information in some instances. How can it be kept confidential?

As provided for in Section 8 of the Risk-Based Capital (RBC) For Insurers Model Act, all information in support of and provided in the RBC reports (to the extent the information therein is not required to be set forth in a publicly available annual statement schedule), with respect to any domestic or foreign insurer, which is filed with the commissioner constitute information that might be damaging to the insurer if made available to its competitors, and therefore shall be kept confidential by the commissioner. This information shall not be made public or be subject to subpoena, other than by the commissioner and then only for the purpose of enforcement actions taken by the commissioner under the Risk-Based Capital (RBC) For Insurers Model Act or any other provision of the insurance laws of the state.

3. The definition of the annuities category talks about “debt incurred for funding an investment account…” Could you give a specific description of what is intended?

One example is a situation where an insurer is borrowing under an advance agreement with a federal home loan bank, under which agreement collateral, on a current fair value basis, is required to be maintained with the bank. This arrangement has many of the characteristics of a GIC, but is classified as debt.

4. The instructions specify that assumptions consistent with those used for Asset Adequacy Analysis testing be used for C-3 RBC, but my company cash flow tests a combination of

universal life and annuities for that analysis and using the same assumptions will produce incorrect results. What was intended in this situation?

Where this situation exists, assumptions should be used for the risk-based capital work that are consistent with those used for the Asset Adequacy Cash Flow Testing. In other words, the assumptions used should be appropriate to the annuity component being evaluated for RBC and consistent with the overall assumption set used for Asset Adequacy Analysis.

Attachment Two

29


Appendix 2 – Appendix 8 from the June 2005 Academy document to be inserted here with adjustments for the new tax rates.

Appendix 2 – Alternative Method for GMDB Risks (2020 Instructions) {Drafting Note: the following is copied from the American Academy of Actuaries June 2005 Report to the NAIC Capital Adequacy Task Force This Appendix describes the Alternative Method for GMDB exposure in significant detail; how it is to be applied and how the factors were developed. Factor tables have been developed using the Conditional Tail Expectation (“CTE”) risk measure at two confidence levels: 65% and 90%. The latter is determined on an “after tax” basis and is required for the RBC C3 Phase II standard for Total Asset Requirement (“TAR”). The former is a pre-tax calculation and should assist the Variable Annuity Reserve Working Group (“VARWG”) in formulating a consistent “alternative method” for statutory reserves.

General

1. It is expected that the Alternative Method (“AltM”) will be applied on a policy-by-policy basis (i.e., seriatim). If the company adopts a cell-based approach, only materially similar contracts should be grouped together. Specifically, all policies comprising a “cell” must display substantially similar characteristics for those attributes expected to affect risk-based capital (e.g., definition of guaranteed benefits, attained age, policy duration, years-to-maturity, market-to-guaranteed value, asset mix, etc.).

2. The Alternative Method determines the TAR as the sum of the Cash Surrender Value and the following three (3) provisions, collectively referred to as the Additional Asset Requirement (“AAR”):

Provision for amortization of the outstanding (unamortized) surrender charges –“Charge Amortization” or “CA”; Provision for fixed dollar expenses/costs net of fixed dollar revenue – “Fixed Expenses” or “FE”; and Provision for claims (in excess of account value) under the guaranteed benefits net of available spread-based revenue (“margin offset”) –

“Guaranteed Cost” or “GC”.

All of these components reflect the impact of income taxes and are explained in more detail later in this Appendix.

The Risk Based Capital amount (C-3 RBC) is determined in aggregate for the block of policies as the TAR less the reserve determined based on Section 7 of VM-21.

Note the following regarding income taxes:

Attachment Two

30


The company determines the CA and FE amounts by projecting the inforce data and incorporating a 21% tax rate and a post-tax discount rate of 4.54% (= 5.75% x [1-21%]).

In determining the GC amounts, a “look-up” function is used which provides a GMDB Cost Factor “f” and Base Margin Offset Factor “g”. These factors (“f” and “g”) represent CTE90 factors on a post-tax basis where a 35% tax rates and 3.74% (= 5.75% x (1-35%)) discount rate has been used. The company needs to multiply these factors by (.79/.65) to adjust the factors for a 21% tax rate basis. It is noted that this adjustment overstates the impact of the lower tax rate as the impact of the higher discount rate has not been reflected.

3. The total AAR (in excess of cash surrender value) is the sum of the AAR calculations for each policy or cell. The result for any given policy (cell) may be negative, zero or positive.

4. For variable annuities without guarantees, the Alternative Method for capital uses the methodology which applied previously to all variable annuities. The charge is 11 percent of the difference between fund balance and cash surrender value if the current surrender charge is based on fund balance. If the current surrender charge is based on fund contributions, the charge is 2.4 percent of the difference for those contracts for which the fund balance exceeds the sum of premiums less withdrawals and 11 percent for those for which that is not the case. In all cases, the result is to be multiplied by 0.79 to adjust for Federal Income Tax. For in-scope contracts, such as many payout annuities with no cash surrender value and no performance guarantees, there is no capital charge.

5. For variable annuities with death benefit guarantees, the AAR for a given policy is equal to: ( ) GCFECAR ++× where:

CA (Charge Amortization) = Provision for amortization of the outstanding (unamortized) surrender charges

FE (Fixed Expense) = Provision for fixed dollar expenses/costs net of fixed dollar revenue

GC (Guaranteed Cost) = Provision for claims (in excess of account value) under the guaranteed benefits net of available spread-based revenue (“margin offset”)

The components CA, FE and GC are calculated separately. CA and FE are defined by deterministic “single-scenario” calculations which account for asset growth, interest, inflation and tax at prescribed rates. Mortality is ignored. However, the actuary determines the appropriate “prudent best estimate” lapses/withdrawal rates for the calculations. The components CA, FE and GC may be positive, zero or negative. is a “scaling factor” that depends on certain risk attributes θ~ for the policy and the product portfolio.

Attachment Two

31


6. The “Alternative Method” factors and formulas for GMDB risks (component GC) have been developed from stochastic testing using the 10,000 “Pre-packaged” scenarios (March 2005). The pre-packaged scenarios have been fully documented under separate cover – see http://www.actuary.org/pdf/life/c3supp_march05.pdf at the American Academy of Actuaries’ website.

7. The model assumptions for the AltM Factors (component GC) are documented in the section of this Appendix entitled Component GC.

8. The table of GC factors that has been developed assumes male mortality at 100% of the MGDB 94 ALB table. Companies using the Alternative Method may use these factors, or may use the procedure described in Methodology Note C3-04 to adjust for the actuary’s Prudent Best Estimate of mortality. Once a company uses the modified method for a block of business, the option to use the unadjusted table is no longer available for that part of its business. In applying the factors to actual inforce business, a 5-year age setback should be used for female annuitants.

9. There are five (5) major steps in using the GC factors to determine the “GC” component of the AAR for a given policy/cell:

a) Classifying the asset exposure;

b) Determining the risk attributes;

c) Retrieving the appropriate nodes from the factor grid;

d) Interpolating the nodal factors, where applicable (optional);

e) Applying the factors to the policy values.

Categorizing the asset value for the given policy or cell involves mapping the entire exposure to one of the eight (8) prescribed “fund classes”. Alternative Method factors are provided for each asset class.

The second step requires the company to determine (or derive) the appropriate attributes for the given policy or cell. These attributes are needed to calculate the required values and access the factor tables:

Product form (“Guarantee Definition”), P.

Adjustment to guaranteed value upon partial withdrawal (“GMDB Adjustment”), A.

Fund class, F.

Attained age of the annuitant, X.

Policy duration since issue, D.

Ratio of account value to guaranteed value, φ.

Attachment Two

32

http://www.actuary.org/pdf/life/c3supp_nov03.pdf


Total account charges, MER.

Other required policy values include:

Account value, AV.

Current guaranteed minimum death benefit, GMDB.

Net deposit value (sum of deposits less sum of withdrawals), NetDeposits1.

Net spread available to fund guaranteed benefits (“margin offset”), α.

The next steps – retrieving the appropriate nodes from the factor grid and interpolation – are explained in the section entitled Component GC of this Appendix. Tools are provided to assist the company in these efforts (see Appendix 9), but their use is not mandatory. This documentation is sufficiently detailed to permit the company to write its own lookup and extraction routines. A calculation example to demonstrate the application of the various component factors to sample policy values is shown in the section Component GC of this Appendix.

1 Net deposits are required only for certain policy forms (e.g., when the guaranteed benefit is capped as a multiple of net policy contributions).

Attachment Two

33


10. The total account charges should include all amounts assessed against policyholder accounts, expressed as a level spread per year (in basis points). This quantity is called the Management Expense Ratio (“MER”) and is defined as the average amount (in dollars) charged against policyholder funds in a given year divided by average account value. Normally, the MER would vary by fund class and be the sum of investment management fees, mortality & expense charges, guarantee fees/risk premiums, etc. The spread available to fund the GMDB costs (“margin offset”, denoted by α) should be net of spread-based costs and expenses (e.g., net of maintenance expenses, investment management fees, trail commissions, etc.), but may be increased for Revenue Sharing as can be reflected in modeling (i.e., had the Alternative Method not been elected) by adhering to the requirements set forth in section 6 of the Modeling Methodology. The section of this Appendix on Component GC describes how to determine MER and α . ‘Time-to-maturity’ is uniquely defined in the factor modeling by T = 95 − X. (This assumes an assumed maturity age of 95 and a current attained age of X.) Net deposits are used in determining benefit caps under the GMDB Roll-up and Enhanced Death Benefit (“EDB”) designs.

11. The GMDB definition for a given policy/cell may not exactly correspond to those provided. In some cases, it may be reasonable to use the factors/formulas for a different product form (e.g., for a “roll-up” GMDB policy near or beyond the maximum reset age or amount, the company should use the “return-of-premium” GMDB factors/formulas, possibly adjusting the guaranteed value to reflect further resets, if any). In other cases, the company might determine the RBC based on two different guarantee definitions and interpolate the results to obtain an appropriate value for the given policy/cell. However, if the policy form (definition of the guaranteed benefit) is sufficiently different from those provided and there is no practical or obvious way to obtain a good result from the prescribed factors/formulas, the company must select one of the following options:

a) Model the “C3 Phase II RBC” using stochastic projections according to the approved methodology;

b) Select factors/formulas from the prescribed set such that the values obtained conservatively estimate the required capital; or

c) Calculate company-specific factors or adjustments to the published factors based on stochastic testing of its actual business. This option is described more fully in the section of this Appendix on Component GC.

Attachment Two

34


12. The actuary must decide if existing reinsurance arrangements can be accommodated by a straight-forward adjustment to the factors and formulas (e.g., quota-share reinsurance without caps, floors or sliding scales would normally be reflected by a simple pro-rata adjustment to the “gross” GC results). For more complicated forms of reinsurance, the company will need to justify any adjustments or approximations by stochastic modeling. However, this modeling need not be performed on the whole portfolio, but can be undertaken on an appropriate set of representative policies. See the section of this Appendix on Component GC.

Component CA

Component CA provides for the amortization of the unamortized surrender charges using the actual surrender charge schedule applicable to the policy. Over time, the surrender charge is reduced and a portion of the charges in the policy are needed to fund the resulting increase in surrender value. This component can be interpreted as the “amount needed to amortize the unamortized surrender charge allowance for the persisting policies plus an implied borrowing cost”. By definition, the amortization for non-persisting lives in each time period is exactly offset by the collected surrender charge revenue (ignoring timing differences and any waiver upon death). The company must project the unamortized balance to the end of the surrender charge period and discount the year-by-year amortization under the following assumptions. All calculations should reflect the impact of income taxes.

Net asset return (i.e., after fees) as shown in Table 1 below. These rates roughly equate to an annualized 5th percentile return over a 10-year horizon2. The 10 year horizon was selected as a reasonable compromise between the length of a typical surrender charge period and the longer testing period usually needed to capture all the costs on "more expensive" portfolios (i.e., lower available spread, lower AV/GV ratio, older ages, etc.). Note, however, that it may not be necessary to use these returns if surrender charges are a function of deposits/premiums.

Income tax and discount rates (after-tax) as defined in Table 9 of this Appendix. .

The “Dynamic Lapse Multiplier” calculated at the valuation date (a function of Account Value (AV) ÷ Guaranteed Value (GV) ratio) is assumed to apply in each future year. This factor adjusts the lapse rate to reflect the antiselection present when the guarantee is in-the-money. Lapse rates may be lower when the guarantees have more value.

Surrender charges and free partial withdrawal provisions should be reflected as per the contract specifications.

“Prudent best estimate” lapse and withdrawal rates. Rates may vary according to the attributes of the business being valued, including, but not limited to, attained age, policy duration, etc.

For simplicity, mortality may be ignored in the calculations.

2 A 5th percentile return is consistent with the CTE90 risk measure adopted in the C3 Phase II RBC methodology.

Attachment Two

35


Unlike the GC component, which requires the actuary to map the entire contract exposure to a single “equivalent” asset class, the CA calculation separately projects each fund (as mapped to the 8 prescribed categories) using the net asset returns in Table 2-1.

Table 2-1: Net Asset Returns for “CA” Component

Asset Class/Fund Net Annualized Return

Fixed Account Guaranteed Rate

Money Market and Fixed Income 0%

Balanced −1%

Diversified Equity −2%

Diversified International Equity −3%

Intermediate Risk Equity −5%

Aggressive or Exotic Equity −8%

Component FE

Component FE establishes a provision for fixed dollar costs (i.e., allocated costs, including overhead and those expenses defined on a “per policy” basis) less any fixed dollar revenue (e.g., annual administrative charges or policy fees). The company must project fixed expenses net of any “fixed revenue” to the earlier of contract maturity or 30 years, and discount the year-by-year amounts under the following assumptions. All calculations should reflect the impact of income taxes.

Income tax and discount rates (after-tax) as defined in Table 9 of this Appendix.

The “Dynamic Lapse Multiplier” calculated at the valuation date (a function of MV÷GV ratio) is assumed to apply in each future year. This factor adjusts the lapse rate to reflect the antiselection present when the guarantee is in-the-money. Lapse rates may be lower when the guarantees have more value.

Per policy expenses are assumed to grow with inflation starting in the second projection year. The ultimate inflation rate of 3% per annum is reached in the 8th year after the valuation date. The company must grade linearly from the current inflation rate (“CIR”) to the ultimate rate. The CIR is the higher of 3% and the inflation rate assumed for expenses in the company’s most recent asset adequacy analysis for similar business.

Attachment Two

36


“Prudent best estimate” for policy termination (i.e., total surrender). Rates may vary according to the attributes of the business being valued, including, but not limited to, attained age, policy duration, etc. Partial withdrawals should be ignored as they do not affect survivorship.

For simplicity, mortality may be ignored in the calculations.

Component GC

The general format for GC may be written as: ( ) ( ) ( )ˆˆGC GV f AV g hθ θ θ= × − × × where GV = current guaranteed minimum death benefit, AV = current

account value and = ( )θαα ~ˆ

g× . The functions ( )f , ( )g and ( )h depend on the risk attributes of the policy θ and product portfolioθ . ( ) Rh =

was introduced in the “General” section as a “scaling factor”. α is the company-determined net spread (“margin offset”) available to fund the guaranteed benefits and 100ˆ =α basis points is the margin offset assumed in the development of the “Base” tabular factors. The functions ( )f , ( )g and ( )h are

more fully described later in this section.

Rearranging terms for GC, we have ( ) ( )[ ]θθ ~~ zAVGVfGC ×−×= . Admittedly, ( )θ~z is a complicated function that depends on the risk attribute sets

θ~ and θ , but conceptually we can view ( )θ~zAV × as a shock to the current account value (in anticipation of the adverse investment return scenarios that typically comprise the CTE(90) risk measure for the AAR) so that the term in the square brackets is a “modified net amount at risk”. Accordingly, ( )θ~f

can be loosely interpreted as a factor that adjusts for interest (i.e., discounting) and mortality (i.e., the probability of the annuitant dying).

In practice, ( )f , ( )g and ( )h are not functions in the typical sense, but values interpolated from the factor grid. The factor grid is a large pre-computed

table developed from stochastic modeling for a wide array of combinations of the risk attribute set. The risk attribute set is defined by those policy and/or product portfolio characteristics that affect the risk profile (exposure) of the business: attained age, policy duration, AV/GV ratio, fund class, etc.

Fund Categorization

The following criteria should be used to select the appropriate factors, parameters and formulas for the exposure represented by a specified guaranteed benefit. When available, the volatility of the long-term annualized total return for the fund(s) – or an appropriate benchmark – should conform to the limits presented. This calculation should be made over a reasonably long period, such as 25 to 30 years.

Attachment Two

37


Where data for the fund or benchmark are too sparse or unreliable, the fund exposure should be moved to the next higher volatility class than otherwise indicated. In reviewing the asset classifications, care should be taken to reflect any additional volatility of returns added by the presence of currency risk, liquidity (bid-ask) effects, short selling and speculative positions.

All exposures/funds must be categorized into one of the following eight (8) asset classes:

1. Fixed Account

2. Money Market

3. Fixed Income

4. Balanced

5. Diversified Equity

6. Diversified International Equity

7. Intermediate Risk Equity

8. Aggressive or Exotic Equity

Fixed Account. The fund is credited interest at guaranteed rates for a specified term or according to a ‘portfolio rate’ or ‘benchmark’ index. The funds offer a minimum positive guaranteed rate that is periodically adjusted according to company policy and market conditions.

Money Market/Short-Term. The fund is invested in money market instruments with an average remaining term-to-maturity of less than 365 days.

Fixed Income. The fund is invested primarily in investment grade fixed income securities. Up to 25% of the fund within this class may be invested in diversified equities or high-yield bonds. The expected volatility of the fund returns will be lower than the Balanced fund class.

Balanced. This class is a combination of fixed income securities with a larger equity component. The fixed income component should exceed 25% of the portfolio and may include high yield bonds as long as the total long-term volatility of the fund does not exceed the limits noted below. Additionally, any aggressive or ‘specialized’ equity component should not exceed one-third (33.3%) of the total equities held. Should the fund violate either of these constraints, it should be categorized as an equity fund. These funds usually have a long-term volatility in the range of 8% − 13%.

Diversified Equity. The fund is invested in a broad based mix of U.S. and foreign equities. The foreign equity component (maximum 25% of total holdings) must be comprised of liquid securities in well-developed markets. Funds in this category would exhibit long-term volatility comparable to that of the S&P500. These funds should usually have a long-term volatility in the range of 13% − 18%.

Attachment Two

38


Diversified International Equity. The fund is similar to the Diversified Equity class, except that the majority of fund holdings are in foreign securities. These funds should usually have a long-term volatility in the range of 14% − 19%.

Intermediate Risk Equity. The fund has a mix of characteristics from both the Diversified and Aggressive Equity Classes. These funds have a long-term volatility in the range of 19% − 25%.

Aggressive or Exotic Equity. This class comprises more volatile funds where risk can arise from: (a) underdeveloped markets, (b) uncertain markets, (c) high volatility of returns, (d) narrow focus (e.g., specific market sector), etc. The fund (or market benchmark) either does not have sufficient history to allow for the calculation of a long-term expected volatility, or the volatility is very high. This class would be used whenever the long-term expected annualized volatility is indeterminable or exceeds 25%.

THE SELECTION OF AN APPROPRIATE INVESTMENT TYPE SHOULD BE DONE AT THE LEVEL FOR WHICH THE GUARANTEE APPLIES. FOR GUARANTEES APPLYING ON A DEPOSIT-BY-DEPOSIT BASIS, THE FUND SELECTION IS STRAIGHTFORWARD. HOWEVER, WHERE THE GUARANTEE APPLIES ACROSS DEPOSITS OR FOR AN ENTIRE CONTRACT, THE APPROACH CAN BE MORE COMPLICATED. IN SUCH INSTANCES, THE APPROACH IS TO IDENTIFY FOR EACH POLICY WHERE THE “GROUPED FUND HOLDINGS” FIT WITHIN THE CATEGORIES LISTED AND TO CLASSIFY THE ASSOCIATED ASSETS ON THIS BASIS.

A seriatim process is used to identify the “grouped fund holdings”, to assess the risk profile of the current fund holdings (possibly calculating the expected long-term volatility of the funds held with reference to the indicated market proxies), and to classify the entire “asset exposure” into one of the specified choices. Here, “asset exposure” refers to the underlying assets (separate and/or general account investment options) on which the guarantee will be determined. For example, if the guarantee applies separately for each deposit year within the contract, then the classification process would be applied separately for the exposure of each deposit year.

In summary, mapping the benefit exposure (i.e., the asset exposure that applies to the calculation of the guaranteed minimum death benefits) to one of the prescribed asset classes is a multi-step process:

1. Map each separate and/or general account investment option to one of the prescribed asset classes. For some funds, this mapping will be obvious, but for others it will involve a review of the fund’s investment policy, performance benchmarks, composition and expected long-term volatility.

2. Combine the mapped exposure to determine the expected long-term “volatility of current fund holdings”. This will require a calculation based on the expected long-term volatilities for each fund and the correlations between the prescribed asset classes as given in Table 2-2.

3. Evaluate the asset composition and expected volatility (as calculated in step 2) of current holdings to determine the single asset class that best represents the exposure, with due consideration to the constraints and guidelines presented earlier in this section.

Attachment Two

39


In step 1., the company should use the fund’s actual experience (i.e., historical performance, inclusive of reinvestment) only as a guide in determining the expected long-term volatility. Due to limited data and changes in investment objectives, style and/or management (e.g., fund mergers, revised investment policy, different fund managers, etc.), the company may need to give more weight to the expected long-term volatility of the fund’s benchmarks. In general, the company should exercise caution and not be overly optimistic in assuming that future returns will consistently be less volatile than the underlying markets.

In step 2., the company should calculate the “volatility of current fund holdings” (σ for the exposure being categorized) by the following formula using the volatilities and correlations in Table 2.

∑∑= =

=n

i

n

jjiijji ww

1 1σσρσ

where ∑

=

kk

ii AV

AVw is the relative value of fund i expressed as a proportion of total contract value, ijρ is the correlation between asset classes i and j and

iσ is the volatility of asset class i (see Table 2). An example is provided at the end of this section.

Attachment Two

40


Table 2-2: Volatilities and Correlations for Prescribed Asset Classes

ANNUAL VOLATILITY FIXED

ACCOUNT MONEY MARKET

FIXED INCOME BALANCED DIVERSE

EQUITY INTL

EQUITY INTERM EQUITY

AGGR EQUITY

1.0% FIXED ACCOUNT 1 0.50 0.15 0 0 0 0 0

1.5% MONEY MARKET 0.50 1 0.20 0 0 0 0 0

5.0% FIXED INCOME 0.15 0.20 1 0.30 0.10 0.10 0.10 0.05

10.0% BALANCED 0 0 0.30 1 0.95 0.60 0.75 0.60

15.5% DIVERSE EQUITY 0 0 0.10 0.95 1 0.60 0.80 0.70

17.5% INTL EQUITY 0 0 0.10 0.60 0.60 1 0.50 0.60

21.5% INTERM EQUITY 0 0 0.10 0.75 0.80 0.50 1 0.70

26.0% AGGR EQUITY 0 0 0.05 0.60 0.70 0.60 0.70 1

Attachment Two

41


As an example, suppose three funds (Fixed Income, diversified U.S. Equity and Aggressive Equity) are offered to clients on a product with a contract level guarantee (i.e., across all funds held within the policy). The current fund holdings (in dollars) for five sample contracts are shown in Table 2-3.

TABLE 2-3: FUND CATEGORIZATION EXAMPLE

1 2 3 4 5

MV Fund X (Fixed Income): 5,000 4,000 8,000 - 5,000

MV Fund Y (Diversified Equity): 9,000 7,000 2,000 5,000 -

MV Fund Z (Aggressive Equity): 1,000 4,000 - 5,000 5,000

Total Market Value: 15,000 15,000 10,000 10,000 10,000

Total Equity Market Value: 10,000 11,000 2,000 10,000 5,000

Fixed Income % (A): 33% 27% 80% 0% 50%

Fixed Income Test (A>75%): No No Yes No No

Aggressive % of Equity (B): 10% 36% n/a 50% 100%

Balanced Test (A>25% & B<33.3%): Yes No n/a No No

Volatility of Current Fund Holdings: 10.9% 13.2% 5.3% 19.2% 13.4%

Fund Classification: Balanced Diversified* Fixed Income Intermediate Diversified

* Although the volatility suggests “Balanced Fund”, the Balanced Fund criteria were not met. Therefore, this ‘exposure’ is moved “up” to Diversified Equity. For those funds classified as Diversified Equity, additional analysis would be required to assess whether they should be instead designated as “Diversified International Equity”.

As an example, the “Volatility of Current Fund Holdings” for policy #1 is calculated as BA + where:

( ) ( ) ( )26.0155.07.0151

159226.005.005.0

151

1552155.005.01.0

159

1552

226.0

1512

155.01592

05.0155

××⋅⋅+××⋅⋅+××⋅⋅=

×+×+×=

B

A

Attachment Two

42


So the volatility for contract #1 = 0026.00092.0 + = 0.109 or 10.9%. Attachment Two

43


Derivation of Total Equivalent Account Charges (MER) and Margin Offset (α)

The total equivalent account charge (“MER”) is meant to capture all amounts that are deducted from policyholder funds, not only those that are commonly expressed as spread-based fees. The MER, expressed as an equivalent annual basis point charge against account value, should include (but not be limited to) the following: investment management fees, mortality & expense charges, administrative loads, policy fees and risk premiums. In light of the foregoing, it may be necessary to estimate the “equivalent MER” if there are fees withdrawn from policyholder accounts that are not expressed as basis point charges against account value.

The margin offset, α , represents the total amount available to fund the guaranteed benefit claims and amortization of the unamortized surrender charge allowance after considering most other policy expenses (including overhead). The margin offset, expressed as an equivalent annual basis point charge against account value, may include the effect of Revenue Sharing in the same manner as would be done for modeling as described in section 6 of the Modeling Methodology, except as may be thereby permitted, should be deemed “permanently available” in all future scenarios. However, the margin offset should not include per policy charges (e.g., annual policy fees) since these are included in FE. It is often helpful to interpret the margin offset as

XMER −=α + RS, where X is the sum of:

Investment management expenses and advisory fees;

Commissions, bonuses (dividends) and overrides;

Maintenance expenses, other than those included in FE; and

Unamortized acquisition costs not reflected in CA.

And RS is the Revenue Sharing to the extent permitted as described above.

Product Attributes and Factor Tables

The tabular approach for the GC component creates a multi-dimensional grid (array) by testing a very large number of combinations for the policy attributes. The results are expressed as factors. Given the seven (7) attributes for a policy (i.e., P, A, F, X, D, φ, MER), two factors are returned for ( )f

and ( )g . The factors are determined by looking up (based on a “key”) into the large, pre-computed multi-dimensional tables and using multi-dimensional

linear interpolation.

The policy attributes for constructing the test cases and the lookup keys are given in Table 2-4.

Attachment Two

44


As can be seen, there are 6 × 2 × 8 × 8 × 5 × 7 × 3 = 80,640 “nodes” in the factor grid. Interpolation is only permitted across the last four (4) dimensions: Attained Age (X), Policy Duration (D), AV÷GV Ratio (φ) and MER. The “MER Delta” is calculated based on the difference between the actual MER and that assumed in the factor testing (see Table 10), subject to a cap (floor) of 100 bps (−100 bps).

Functions are available to assist the company in applying the Alternative Method for GMDB risks. These functions perform the factor table lookups and associated multi-dimensional linear interpolations. Their use is not mandatory. Based on the information in this document, the company should be able to write its own lookup and retrieval routines. Interpolation in the factor tables is described further later in this section.

Table 2-4: Nodes of the Factor Grid

Policy Attribute Key : Possible Values & Description

Product Definition, P.

0 : 0 Return-of-premium. 1 : 1 Roll-up (3% per annum). 2 : 2 Roll-up (5% per annum). 3 : 3 Maximum Anniversary Value (MAV). 4 : 4 High of MAV and 5% Roll-up. 5 : 5 Enhanced Death Benefit (excl. GMDB)

GV Adjustment Upon Partial Withdrawal, A.

0 : 0 Pro-rata by market value. 1 : 1 Dollar-for-dollar.

Fund Class, F.

0 : 0 Fixed Account. 1 : 1 Money Market. 2 : 2 Fixed Income (Bond). 3 : 3 Balanced Asset Allocation. 4 : 4 Diversified Equity. 5 : 5 International Equity. 6 : 6 Intermediate Risk Equity. 7 : 7 Aggressive / Exotic Equity.

Attained Age (Last Birthday), X.

0 : 35 4 : 65 1 : 45 5 : 70 2 : 55 6 : 75 3 : 60 7 : 80

Policy Duration (years-since-issue), D.

0 : 0.5 1 : 3.5

Attachment Two

45


2 : 6.5 3 : 9.5 4 : 12.5

Account Value-to-Guaranteed Value Ratio, φ.

0 : 0.25 4 : 1.25 1 : 0.50 5 : 1.50 2 : 0.75 6 : 2.00 3 : 1.00

Annualized Account Charge Differential from Table 2-10 Assumptions (“MER Delta”)

0 : −100 bps 1 : +0 2 : +100

Attachment Two

46


A test case (i.e., a node on the multi-dimensional matrix of factors) can be uniquely identified by its key, which is the concatenation of the individual ‘policy attribute’ keys, prefixed by a leading ‘1’. For example, the key ‘12034121’ indicates the factor for a 5% roll-up GMDB, where the GV is adjusted pro-rata upon partial withdrawal, balanced asset allocation, attained age 65, policy duration 3.5, 75% AV/GV ratio and “equivalent” annualized fund based charges equal to the ‘base’ assumption (i.e., 250 bps p.a.).

The factors are contained in the file “C3-II GMDB Factors 100%Mort CTE(90) (2005-03-29).csv”, a comma-separated value text file. Each “row” represents the factors/parameters for a test policy as identified by the lookup keys shown in Table 2-4. Rows are terminated by new line and line feed characters.

Each row consists of 5 entries, described further below.

1 2 3 4 5 Test Case

Identifier (Key) Base GMDB Cost

Factor Base Margin Offset Factor

Scaling Adjustment (Intercept)

Scaling Adjustment (Slope)

GMDB Cost Factor. This is the term ( )θ~f in the formula for GC. The parameter set θ~ is defined by ( )MERDXFAP ,,,,,, φ . Here, φ is the AV/GV

ratio for the benefit exposure (e.g., policy) under consideration. The values in the factor grid represent CTE(90) of the sample distribution3 for the present value of guaranteed benefit cash flows (in excess of account value) in all future years (i.e., to the earlier of contract maturity and 30 years), normalized by guaranteed value.

Base Margin Offset Factor. This is the term ( )θ~g in the formula for GC. The parameter set θ~ is defined by ( )MERDXFAP ,,,,,, φ . Here, φ is the

AV/GV ratio for the benefit exposure (e.g., policy) under consideration. The values in the factor grid represent CTE(90) of the sample distribution for the present value of margin offset cash flows in all future years (i.e., to the earlier of contract maturity and 30 years), normalized by account value. Note that the Base Margin Offset Factors assume 100ˆ =α basis points of “margin offset” (net spread available to fund the guaranteed benefits).

3 Technically, the sample distribution for “present value of net cost” = PV[GMDB claims] – PV[Margin Offset] was used to determine the scenario results that comprise the CTE90 risk measure. Hence,

the “GMDB Cost Factors” and “Base Margin Offset Factors” are calculated from the same scenarios.

Attachment Two

47


All else being equal, the margin offset α has a profound effect on the resulting AAR. In comparing the Alternative Method against models for a variety of GMDB portfolios, it became clear that some adjustment factor would be required to “scale” the results to account for the diversification effects4 of

attained age, policy duration and AV/GV ratio. The testing examined 20.01 ==MER

W α and 60.02 ==

MERW α

, where α = available margin offset and

MER = total “equivalent” account based charges, in order to understand the interaction between the margin ratio (“W”) and AAR.

Based on this analysis, the Scaling Factor is defined as:

( ) 0 1ˆh R Wθ β β= = + ×

0β and 1β are respectively the intercept and slope for the linear relationship, defined by the parameter set θ = ( )ˆ, ,P F φ . Here, φ is 90% of the aggregate

AV/GV for the product form (i.e., not for the individual policy or cell) under consideration. In calculating the Scaling Factor directly from this linear function, the margin ratio “W” must be constrained5 to the range [ ]0.2,0.6 .

It is important to remember that ∑∑×=

GVAV

90.0φ for the product form being evaluated (e.g., all 5% Roll-up policies). The 90% factor is meant to reflect

the fact that the cost (payoff structure) for a basket of otherwise identical put options (e.g., GMDB) with varying degrees of in-the-moneyness (i.e., AV/GV ratios) is more left-skewed than the cost for a single put option at the “weighted average” asset-to-strike ratio.

To appreciate the foregoing comment, consider a basket of two 10-year European put options as shown in Table 2-5. These options are otherwise identical except for their “market-to-strike price” ratios. The option values are calculated assuming a 5% continuous risk-free rate and 16% annualized volatility. The combined option value of the portfolio is $9.00, equivalent to a single put option with S = $180.92 and X = $200. The market-to-strike (i.e., AV/GV) ratio is 0.905, which is less than the average AV/GV = 1 =

100$100$125$75$

++ .

4 By design, the Alternative Methodology does not directly capture the diversification benefits due to a varied asset profile and product mix. This is not a flaw of the methodology, but a consequence of

the structure. Specific assumptions would be required to capture such diversification effects. Unfortunately, such assumptions might not be applicable to a given company and could grossly over-estimate the ensuing reduction in required capital.

5 The scaling factors were developed by testing “margin ratios” 1 0.2W = and 2 0.6W = . Using values outside this range could give anomalous results.

Attachment Two

48


Table 2-5: Equivalent Single European Put Option

Equivalent Single Put

Option

Put Option A (“in-the-money”)

Put Option B (“out-of-the-

money”)

Market value (AV) $180.92 $75 $125

Strike price (GV) $200.00 $100 $100

Option Value $9.00 $7.52 $1.48

Scaling Adjustment (Intercept). The scaling factor ( )ˆh Rθ = is a linear function of W, the ratio of margin offset to MER. This is the intercept 0β that

defines the line.

Scaling Adjustment (Slope). The scaling factor ( )ˆh Rθ = is a linear function of W, the ratio of margin offset to MER. This is the slope 1β that defines

the line.

Table 2-6 shows the “Base Cost” and “Base Margin Offset” values from the factor grid for some sample policies. As mentioned earlier, the Base Margin

Offset factors assume 100 basis points of “available spread”. The “Margin Factors” are therefore scaled by the ratio100α

, where α = the actual margin

offset (in basis points per annum) for the policy being valued. Hence, the margin factor for the 7th sample policy is exactly half the factor for node 12044121 (the 4th sample policy in Table 6). That is, 0.02160 = 0.5 × 0.04319.

Attachment Two

49


Table 2-6: Sample Nodes on the Factor Grid

KEY GMDB TYPE

GV ADJUST

FUND CLASS AGE POLICY

DUR AV/GV MER (bps) OFFSET

COST FACTOR

MARGIN FACTOR

10132031 ROP $-for-$ Balanced Allocation 55 0.5 1.00 250 100 0.01073 0.04172



12044121 5% Rollup Pro-rata Diverse

Equity 65 3.5 0.75 250 100 0.18484 0.04319


Equity 65 3.5 1.00 250 100 0.12931 0.03944


Equity 65 3.5 1.25 250 100 0.08757 0.03707


Equity 65 3.5 0.75 250 50 0.18484 0.02160

Interpolation in the Factor Tables

Interpolation is only permitted across the last four (4) dimensions of the risk parameter setθ~ : Attained Age (X), Policy Duration (D), AV÷GV Ratio (φ) and MER. The “MER Delta” is calculated based on the difference between the actual MER and that assumed in the factor testing (see Table 2-10), subject to a cap (floor) of 100 bps (−100 bps). In general, the calculation for a single policy will require three applications of multi-dimensional linear interpolation between the 16 = 24 factors/values in the grid:

(1) To obtain the Base Factors ( )θ~f and ( )θ~g .

(2) To obtain the Scaling Factor ( )ˆ .h Rθ = .

Based on the input parameters, the supplied functions (see Appendix 9) will automatically perform the required lookups, interpolations and calculations for ( )ˆh Rθ = , including the constraints imposed on the margin ratio W. Use of the tools noted in Appendix 9 is not mandatory.

Multi-dimensional interpolation is an iterative extension of the familiar two-dimensional linear interpolation for a discrete function ( )xV :

Attachment Two

50


( ) ( ) ( ) ( )

xk

kkk

xx

andxVxVxV

−=

×+×−=+

+

+

1

11~

δξ

ξξδ

In the above formulation, ( )xV~ is assumed continuous and kx and 1+kx are defined values (“nodes”) for ( )xV . By definition, ( ) 1+≤+≤ kkk xxx δ so that

10 ≤≤ ξ . In effect, multi-dimensional interpolation repeatedly applies simple linear interpolation one dimension at a time until a single value is obtained.

Multi-dimensional interpolation across all four dimensions is not required. However, simple linear interpolation for AV÷GV Ratio (φ) is mandatory. In this case, the company must choose nodes for the other three (3) dimensions according to the following rules:

Risk Attribute (Dimension) Node Determination

Attained Age Use next higher attained age.

Policy Duration Use nearest.

MER Delta Use nearest (capped at +100 & floored at –100 bps. For example, if the actual policy/cell is attained age 62, policy duration 4.25 and MER Delta = +55 bps, the company should use the nodes defined by attained age 65, policy duration 3.5 and MER Delta = +100.

Table 2-7 provides an example of the fully interpolated results for a 5% Roll-up “Pro Rata” policy mapped to the Diversified Equity class (first row). While Table 2-7 does not demonstrate how to perform the multi-dimensional interpolation, it does show the required 16 nodes from the Base Factors. The margin offset is assumed to be 100 basis points.

Attachment Two

51


Table 2-7: Base Factors for a 5% Rollup GMDB Policy, Diversified Equity

Key Age Policy Dur

Policy Av/Gv

Mer (Bps)

Base Cost Factor

Base Margin Factor

INTERPOLATED 62 4.25 0.80 265 0.15010 0.04491

12043121 60 3.5 0.75 250 0.14634 0.04815 12043122 60 3.5 0.75 350 0.15914 0.04511 12043131 60 3.5 1.00 250 0.10263 0.04365 12043132 60 3.5 1.00 350 0.11859 0.04139 12043221 60 6.5 0.75 250 0.12946 0.04807 12043222 60 6.5 0.75 350 0.14206 0.04511 12043231 60 6.5 1.00 250 0.08825 0.04349 12043232 60 6.5 1.00 350 0.10331 0.04129

12044121 65 3.5 0.75 250 0.18484 0.04319 12044122 65 3.5 0.75 350 0.19940 0.04074 12044131 65 3.5 1.00 250 0.12931 0.03944 12044132 65 3.5 1.00 350 0.14747 0.03757 12044221 65 6.5 0.75 250 0.16829 0.04313 12044222 65 6.5 0.75 350 0.18263 0.04072 12044231 65 6.5 1.00 250 0.11509 0.03934 12044232 65 6.5 1.00 350 0.13245 0.03751

The interpolations required to compute the Scaling Factor are slightly different from those needed for the Base Factors. Specifically, the user should not interpolate the intercept and slope terms for each surrounding node, but rather interpolate the Scaling Factors applicable to each of the nodes.

Table 2-8 provides an example of the Scaling Factor for the sample policy given earlier in Table 2-7 (i.e., a 5% Roll-up “Pro Rata” policy mapped to the Diversified Equity class) as well as the nodes used in the interpolation. The aggregate AV/GV for the product portfolio (i.e., all 5% Roll-up policies combined) is 0.75; hence, 90% of this value is 0.675 as shown under “Adjusted Product AV/GV”. As before, the margin offset is 100 basis points per annum.

Attachment Two

52


Table 2-8: Interpolated Scaling Factors for a 5% Rollup GMDB Policy, Diversified Equity

Key Age Policy Dur

Adjusted Product Av/Gv

Mer (Bps) Intercept Slope Scaling

Factor

INTERPOLATED 62 4.25 0.675 265 n/a n/a 0.871996

12043111 60 3.5 0.50 250 0.855724 0.092887 0.892879 12043112 60 3.5 0.50 350 0.855724 0.092887 0.882263 12043121 60 3.5 0.75 250 0.834207 0.078812 0.865732 12043122 60 3.5 0.75 350 0.834207 0.078812 0.856725 12043211 60 6.5 0.50 250 0.855724 0.092887 0.892879 12043212 60 6.5 0.50 350 0.855724 0.092887 0.882263 12043221 60 6.5 0.75 250 0.834207 0.078812 0.865732 12043222 60 6.5 0.75 350 0.834207 0.078812 0.856725

12044111 65 3.5 0.50 250 0.855724 0.092887 0.892879 12044112 65 3.5 0.50 350 0.855724 0.092887 0.882263 12044121 65 3.5 0.75 250 0.834207 0.078812 0.865732 12044122 65 3.5 0.75 350 0.834207 0.078812 0.856725 12044211 65 6.5 0.50 250 0.855724 0.092887 0.892879 12044212 65 6.5 0.50 350 0.855724 0.092887 0.882263 12044221 65 6.5 0.75 250 0.834207 0.078812 0.865732 12044222 65 6.5 0.75 350 0.834207 0.078812 0.856725

Adjustments to GC for Product Variations & Risk Mitigation/Transfer

In some cases, it may be necessary for the company to make adjustments to the published factors due to:

1. A variation in product form wherein the definition of the guaranteed benefit is materially different from those for which factors are available (see Table 2-9); and/or

Attachment Two

53


2. A risk mitigation / management strategy that cannot be accommodated through a straight-forward and direct adjustment to the published values. Attachment Two

54


Any adjustments to the published factors must be fully documented and supported through stochastic analysismodeling. Such analysismodeling may require stochastic simulations, but would not ordinarily be based on full inforce projections. Instead, a representative “model office” should be sufficient. In the absence of material changes to the product design, risk management program and Alternative Method (including the published factors), the company would not be expected to redo this analysismodeling each year.

Note that minor variations in product design do not necessarily require additional effort. In some cases, it may be reasonable to use the factors/formulas for a different product form (e.g., for a “roll-up” GMDB policy near or beyond the maximum reset age or amount, the company should use the “return-of-premium” GMDB factors/formulas, possibly adjusting the guaranteed value to reflect further resets, if any). In other cases, the company might determine the RBC based on two different guarantee definitions and interpolate the results to obtain an appropriate value for the given policy/cell. Likewise, it may be possible to adjust the Alternative Method results for certain risk transfer arrangements without significant additional work (e.g., quota-share reinsurance without caps, floors or sliding scales would normally be reflected by a simple pro-rata adjustment to the “gross” GC results).

However, if the policy design is sufficiently different from those provided and/or the risk mitigation strategy is non-linear in its impact on the AAR, and there is no practical or obvious way to obtain a good result from the prescribed factors/formulas, the company must justify any adjustments or approximations by stochastic modeling. Notably this modeling need not be performed on the whole portfolio, but can be undertaken on an appropriate set of representative policies.

The remainder of this section suggests a process for adjusting the published “Cost” and “Margin Offset” factors due to a variation in product design (e.g., a “step-up” option at every 7th anniversary whereby the guaranteed value is reset to the account value, if higher). Note that the “Scaling Factors” (as determined by the slope and intercept terms in the factor table) would not be adjusted.

The steps for adjusting the published Cost and Margin Offset factors for product design variations are:

1. Select a policy design in the published tables that is similar to the product being valued. Execute cashflow projections using the documented assumptions (see Tables 2-9 and 2-10 ) and the pre-packaged scenarios from the prescribed generators for a set of representative cells (combinations of attained age, policy duration, asset class, AV/GV ratio and MER). These cells should correspond to nodes in the factor grid. Rank (order) the sample distribution of results for the present value of net cost6. Determine those scenarios which comprise CTE(90).

6 Present value of net cost = PV[ guaranteed benefit claims in excess of account value ] – PV[ margin offset ]. The discounting includes cashflows in all future years (i.e., to the earlier of contract

maturity and the end of the horizon).

Attachment Two

55


2. Using the results from step 1., average the present value of cost for the CTE(90) scenarios and divide by the current guaranteed value. For a the Jth cell, denote this value by JF . Similarly, average the present value of margin offset revenue for the same subset of scenarios and divide by

account value. For the Jth cell, denote this value by JG .

3. Extract the corresponding factors from the published grid. For each cell, calibrate to the published tables by defining a “model adjustment factor” (denoted by asterisk) separately for the “cost” and “margin offset” components:

( )J

J FfF θ~* = and ( )

JJ G

gG θ~ˆ* =

4. Execute “product specific” cashflow projections using the documented assumptions and pre-packaged scenarios from the prescribed generators for the same set of representative cells. Here, the company should model the actual product design. Rank (order) the sample distribution of results for the present value of net cost. Determine those scenarios which comprise CTE(90).

5. Using the results from step 4., average the present value of cost for the CTE(90) scenarios and divide by the current guaranteed value. For a the Jth cell, denote this value by JF . Similarly, average the present value of margin offset revenue for the same subset of scenarios and divide by

account value. For a the Jth cell, denote this value by JG .

6. To calculate the AAR for the specific product in question, the company should implement the Alternative Method as documented, but use *JJ FF ×

in place of ( )θ~f and *JJ GG × instead of ( )θ~g . The company must use the “Scaling Factors” for the product evaluated in step 1. (i.e., the product

used to calibrate the company’s cashflow model).

Attachment Two

56


Assumptions for the Alternative Method Published GMDB Factors

This subsection reviews the model assumptions used to develop the Alternative Method factors. Each node in the factor grid is effectively the modeled result for a given “cell”.

Table 2-9: Model Assumptions & Product Characteristics

Account Charges (MER) Vary by fund class. See Table 2-10 later in this section. Base Margin Offset 100 basis points per annum

GMDB Description

1. ROP = return of premium ROP. 2. ROLL = 5% roll-up, capped at 2.5 × premium, frozen at age 80. 3. MAV = annual ratchet (maximum anniversary value), frozen at age 80. 4. HIGH = Higher of 5% roll-up and annual ratchet frozen at age 80. 5. EDB = ROP + 40% Enhanced Death Benefit (capped at 40% of deposit).

Adjustment to GMDB Upon Partial Withdrawal “Pro-Rata by Market Value” and “Dollar-for-Dollar” are tested separately.

Surrender Charges Ignored (i.e., zero). Reflected in the “CA” component of the AAR.

Single Premium / Deposit $100,000. No future deposits; no intra-policy fund rebalancing.

Base Policy Lapse Rate • Pro-rata by MV: 10% p.a. at all policy durations (before dynamics) • Dollar-for-dollar: 2% p.a. at all policy durations (no dynamics)

Partial Withdrawals • Pro-rata by MV: None (i.e., zero) • Dollar-for-dollar: Flat 8% p.a. at all policy durations (as a % of AV). No dynamics or anti-selective behavior.

Mortality 100% of MGDB 94 ALB.

Gender /Age Distribution 100% male. Methodology accommodates different attained ages and policy durations. A 5-year age setback will be used for female annuitants.

Max. Annuitization Age All policies terminate at age 95.

Fixed Expenses, Annual Fees Ignored (i.e., zero). Reflected in the “FE” component of the AAR.

Income Tax Rate 2135%

Attachment Two

57


Discount Rate 3.74.54% (after-tax) effective = 5.75% pre-tax.

Dynamic Lapse Multiplier (Applies only to policies where GMDB is adjusted “pro-rata by MV” upon withdrawal)

U=1, L=0.5, M=1.25, D=1.1 Applied to the ‘Base Policy Lapse Rate’ (not withdrawals).

Attachment Two

58


Notes on GMDB Factor Development

The roll-up is continuous (not simple interest, not stepped at each anniversary) and is applied to the previous roll-up guaranteed value (i.e., not the contract guaranteed value under HIGH).

The Enhanced Death Benefit (“EDB”) is floored at zero. It pays out 40% of the gain in the policy upon death at time t:

( )[ ]DepositAVMAXDepositMINB tt −××= ,040.0,40.0 . The test policy also has a 100% return-of-premium GMDB, but the EDB Alternative

Factors will be net of the GMDB component. That is, the EDB factors are ‘stand-alone’ and applied in addition to the GMDB factors.

The “Base Policy Lapse Rate” is the rate of policy termination (total surrenders). Policy terminations (surrenders) are assumed to occur throughout the policy year (not only on anniversaries).

Partial withdrawals (if applicable) are assumed to occur at the end of each time period (quarterly).

Account charges (“MER”) represent the total amount (annualized, in basis points) assessed against policyholder funds (e.g., sum of investment management fees, mortality and expense charges, risk premiums, policy/administrative fees, etc.). They are assumed to occur throughout the policy year (not only on anniversaries).

Table 2-10: Account-Based Fund Charges (bps per annum)

Asset Class / Fund Account Value Charges (MER)

Fixed Account 0 Money Market 110 Fixed Income (Bond) 200 Balanced 250 Diversified Equity 250 Diversified International Equity 250 Intermediate Risk Equity 265 Aggressive or Exotic Equity 275

Attachment Two

59


Calculation Example

Continuing the previous example (see Tables 2-7 and 2-8) for a 5% Roll-up GMDB policy mapped to Diversified Equity, suppose we have the policy/product parameters as specified in Table 2-11.

Table 2-11: Sample Policy Results for 5% Roll-up GMDB, Diversified Equity

Parameter Value Description Deposit Value $100.00 Total deposits adjusted for partial withdrawals.

Account Value $98.43 Total account value at valuation date, in dollars.

GMDB $123.04 Current guaranteed minimum death benefit, in dollars.

Attained Age 62 Attained age at the valuation date (in years).

Policy Duration 4.25 Policy duration at the valuation date (in years).

GV Adjustment Pro-Rata GMDB adjusted pro-rata by MV upon partial withdrawal.

Fund Class Diversified Equity Contract exposure mapped to Diversified Equity as per the Fund Categorization instructions in the section of this Appendix on Component GC.

MER 265 Total charge against policyholder funds (bps).

ProductCode 2 Product Definition code as per lookup key in Table 4.

GVAdjust 0 GV Adjustment Upon Partial Withdrawal as per key in Table 2-4.

FundCode 4 Fund Class code as per lookup key in Table 2-4.

PolicyMVGV 0.800 Contract account value divided by GMDB.

AdjProductMVGV 0.675 90% of the aggregate AV/GV for the Product portfolio.

RC 150 Margin offset (basis points per annum).

Using the usual notation, ( ) ( ) ( )ˆˆGC GV f AV g hθ θ θ= × − × × .

Attachment Two

60


( )θ~f = 0.150099 = GetCostFactor(2, 0, 4, 62, 4.25, 0.8, 265)

( )θ~g = 0.067361 = GetMarginFactor(2, 0, 4, 62, 4.25, 0.8, 265, 150)

( )ˆh θ = 0.887663 = GetScalingFactor(2, 0, 4, 62, 4.25, 0.675, 265, 150)

Hence, GC = $12.58 = ( 123.04 × 0.150099 ) – ( 98.43 × 0.067361 × 0.887663 ). As a normalized value, this quantity is 12.78% of account value, 10.23% of guaranteed value and 51.1% of the current net amount at risk (Net amount at risk = GV – AV).

Note that ( ) ( ) 044907.0~ˆ

~ˆ100150

×=×= θααθ gg where ( )θ~g is “per 100 basis points” of available margin offset.

( )θ~g = 0.044907 = GetMarginFactor(2, 0, 4, 62, 4.25, 0.8, 265, 100)

Attachment Two

61


Attachment Two

62

3 I Proposed framework revisions I RBC C3 Charge

Attachment C

Attachment Three

63

Attachment Three

64

Attachment Three

65

Attachment Three

66

VM-21

© 2017 National Association of Insurance Commissioners 1-1

VM-21: Requirements for Principle-Based Reserves for Variable Annuities

Table of Contents

Section 1: Background ................................................................................................................... 21-1 Section 2: Scope and Effective Date .............................................................................................. 21-8 Section 3: Reserve Methodology .................................................................................................... 21-9 Section 3: Determination of CTE Amount Based on Projections ................................................ 21-10 Section 4: Reinsurance and Statutory Reporting Issues ............................................................... 21-19 Section 5: Standard Scenario Requirements ................................................................................. 21-21 Section 6: Alternative Methodology ............................................................................................ 21-33 Section 7: Scenario Calibration Criteria ....................................................................................... 21-50 Section 8: Allocation of the Aggregate Reserves to the Contract Level ...................................... 21-55 Section 9: Modeling of Hedges .................................................................................................... 21-57 Section 10: Certification Requirements .......................................................................................... 21-61 Section 11: Contract-Holder Behavior Assumptions ..................................................................... 21-66 Section 12: Specific Guidance and Requirements for Setting Prudent Estimate Mortality Assumptions21-72 Section 4: Determination of the Stochastic Reserve .................................................................... 21-11 Section 5: Reinsurance Ceded ...................................................................................................... 21-20 Section 1: Background

A. Purpose

These requirements establish the minimum reserve valuation standard for variable annuity contracts, and certain other policies and contracts (“contracts”) as defined in the Scope, issued on or after the operative date of the Valuation Manual as required by Model #820. These requirements constitute the Commissioners Annuity Reserve Valuation Method (CARVM) for variable annuity contracts by defining the assumptions and methodologies that will comply with Model #820. It also applies similar assumptions and methodologies to contracts that contain characteristics similar to those described in the scope but that are not directly subject to CARVM.all contracts encompassed by the Scope.

The contracts subject to these requirements may be aggregated with the contracts subject to Actuarial Guideline XLIII—CARVM for Variable Annuities (AG 43), published in Appendix C of the AP&P Manual, for purposes of performing and documenting the reserve calculations.

Guidance Note: It is intended that VM-21 reserve calculation requirements will mirrorin VM-21 also be used for those contracts issued prior to January 1, 2017 which are otherwise in the scope of VM-21. AG 43 references the calculation requirements of AG 43VM-21, and reserves for contracts subject to both VM-21 and AG 43 may be computed as a single group. If a company chooses to aggregate business subject to AG 43 with business subject to VM-21 in calculating the reserve, then the provisions in VM-G apply to this aggregate principle-based valuation.

B. Principles

The projection methodology used to calculate the CTE amountstochastic reserve, as well as the approach used to develop the Alternative Methodology, is based on the following set of principles. These principles should be followed when interpreting and applying the methodology in these requirements and analyzing the resulting reserves.

Guidance Note: The principles should be considered in their entirety, and it is required that companies meet these principles with respect to only those contracts that fall within the scope of these requirements and are in force as of the valuation date to which these requirements are applied.

Formatted

Commented [HR1]: Is the Table of Contents being left as TBD? There are currently more sections than this in the updated VM-21.

Commented [LE2]: Reference “Section 2.A” instead of “Scope”?

Commented [LE3]: Same comment as above.

Commented [HR4]: Do we still need to say “It is intended”? Since AG43 is being simultaneously revised, can’t we just say that these are also used for contracts subject to AG43?

Commented [LE5]: This is very confusing as written. I suggest something like this: “Effectively, through reference in AG 43, the reserve requirements in VM-21 also apply to those contracts issued prior to January 1, 2017 that would otherwise be encompassed by the scope of VM-21.” Then pick it up at “Reserves for contracts subject to both VM-21 and AG 43 may be computed as a single group. If a company…”

Commented [LE6]: Does deleting the word “only” address the concern? Or does that open up another can of worms?

Commented [HR7]: Any concern that this does not include AG43? That is, is AG43 clearly inheriting the calculation requirements but not necessarily the principles?

Attachment Four

67

2

Principle 1: The objective of the approach used to determine the CTE amountstochastic reserve is to quantify the amount of statutory reserves needed by the company to be able to meet contractual obligations in light of the risks to which the company is exposed.

Principle 2: The calculation of the CTE amountstochastic reserve is based on the results derived from an analysis of asset and liability cash flows produced by the application of a stochastic cash-flow model to equity return and interest rate scenarios. For each scenario, the greatest present value of accumulated surplus deficiency is calculated. The analysis reflects prudent estimate assumptions for deterministic variables and is performed in aggregate (subject to limitations related to contractual provisions) to allow the natural offset of risks within a given scenario. The methodology uses a projected total statutory balance sheet approach by including all projected income, benefit and expense items related to the business in the model and sets the CTE amountstochastic reserve at a degree of confidence using the CTE measure applied to the set of scenario specific greatest present values of accumulated statutory deficiencies that is deemed to be reasonably conservative over the span of economic cycles.

Guidance Note: Examples where full aggregation between contracts may not be possible include experience rated group contracts and the operation of reinsurance treaties.

Principle 3: The implementation of a model involves decisions about the experience assumptions and the modeling techniques to be used in measuring the risks to which the company is exposed. Generally, assumptions are to be based on the conservative end of the actuary’s confidence interval. The choice of a conservative estimate for each assumption may result in a distorted measure of the total risk. Conceptually, the choice of assumptions and the modeling decisions should be made so that the final result approximates what would be obtained for the CTE amountstochastic reserve at the required CTE level if it were possible to calculate results over the joint distribution of all future outcomes. In applying this concept to the actual calculation of the CTE amountstochastic reserve, the actuarycompany should be guided by evolving practice and expanding knowledge base in the measurement and management of risk.

Guidance Note: The intent of Principle 3 is to describe the conceptual framework for setting assumptions. Section 11 provides the requirements and guidance for setting contract-holder behavior assumptions and includes alternatives to this framework if the actuarycompany is unable to fully apply this principle.

Principle 4: While a stochastic cash-flow model attempts to include all real-world risks relevant to the objective of the stochastic cash-flow model and relationships among the risks, it will still contain limitations because it is only a model. The calculation of the CTE amountstochastic reserve is based on the results derived from the application of the stochastic cash-flow model to scenarios, while the actual statutory reserve needs of the company arise from the risks to which the company is (or will be) exposed in reality. Any disconnect between the model and reality should be reflected in setting prudent estimate assumptions to the extent not addressed by other means.

Principle 5: Neither a cash-flow scenario model nor a method based on factors calibrated to the results of a cash-flow scenario model can completely quantify a company’s exposure to risk. A model attempts to represent reality but will always remain an approximation thereto and, hence, uncertainty in future experience is an important consideration when determining the CTE amount.stochastic reserve. Therefore, the use of assumptions, methods, models, risk management strategies (e.g., hedging), derivative instruments, structured investments or any other risk transfer arrangements (such as reinsurance) that serve solely to reduce the calculated CTE amountstochastic reserve without also reducing risk on scenarios similar to those used in the actual cash-flow modeling are inconsistent with these principles. The use of assumptions and risk management strategies should be appropriate to the business and not merely constructed to exploit “foreknowledge” of the components of the required methodology.

Commented [JW8]: [1. JR Comment: This principle appears to equate the stochastic reserve with the statutory reserve. Since the statutory reserve = stochastic reserve + additional standard projection amount, this equation no longer holds.]

Commented [JW9]: [2. JR Comment: This is not a principle. It is a description of the methodology to be used.]

Commented [JW10]: [3. JR Comments: 1. This principle presumes that the purpose of prudent estimate assumptions is related to “any disconnect between the model and reality”. However, the real purpose of prudent estimate assumptions is to provide a hedge for the uncertainty around our estimate of future events. 2. Similar to principle 1, this appears to equate stochastic reserve and statutory reserve. 3. The second sentence does not seem relevant to the notion that a model has limitations, which is the basic premise of this principle.]

Commented [JW11]: [4. JR Comment: The term “cash-flow scenario model” should be replaced by “stochastic cash-flow model” (two places).]

Attachment Four

68

3

C. Risks Reflected and Risks Not Reflected

1. The risks reflected in the calculation of reserves under these requirements arise from actual or potential events or activities that are both:

a. Directly related to the contracts falling under the scope of these requirements or their supporting assets., and

b. Capable of materially affecting the reserve.

2. Categories and examples of risks reflected in the reserve calculations include, but are not necessarily limited to:

a. Asset risks

i. Separate account fund performance.

ii. Credit risks (e.g., default or rating downgrades).

iii. Commercial mortgage loan roll-over rates (roll-over of bullet loans).

iv. Uncertainty in the timing or duration of asset cash flows (e.g., shortening (prepayment risk) and lengthening (extension risk)).

v. Performance of equities, real estate and Schedule BA assets.

vi. Call risk on callable assets.

vii. Risk associated with hedge instrument (includes basis, gap, price, parameter estimation risks and variation in assumptions).

viii. Currency risk.

b. Liability risks

i. Reinsurer default, impairment or rating downgrade known to have occurred before or on the valuation date.

ii. Mortality/longevity, persistency/lapse, partial withdrawal and premium payment risks.

iii. Utilization risk associated with guaranteed living benefits.

iv. Anticipated mortality trends based on observed patterns of mortality improvement or deterioration, where permitted.

v. Annuitization risks.

vi. Additional premium dump-ins (high interest rate guarantees in low interest rate environments).

c. Combination risks

i. Risks modeled in the company’s risk assessment processes that are related to the contracts, as described above.

Commented [HR12]: Note that 1 and 2 address risks reflected, while 3 and 4 address risks not reflected.

Attachment Four

69

4

ii. Disintermediation risk (including such risk related to payment of surrender orpartial withdrawal benefits).

iii. Risks associated with revenue-sharing income.

3. The risks not necessarily reflected in the calculation of reserves under these requirements are:

a. Those not reflected in the determination of RBC.

b. Those reflected in the determination of RBC but arising from obligations of the company not directly related to the contracts falling under the scope of these requirements, or theirsupporting assets, as described above.

4. Categories and examples of risks not reflected in the reserve calculations include, but are notnecessarily limited to:

a. Asset risks

bi. Liquidity risks associated with a “run on the bank”.

cb. Liability risks

i. Reinsurer default, impairment or rating downgrade occurring after the valuationdate.

ii. Catastrophic events (e.g., epidemics or terrorist events).

iii. Major breakthroughs in life extension technology that have not yet fundamentally altered recently observed mortality experience.

iv. Significant future reserve increases as an unfavorable scenario is realized.

dc. General business risks

i. Deterioration of reputation.

ii. Future changes in anticipated experience (reparameterization in the case ofstochastic processes), which would be triggered if and when adverse modeledoutcomes were to actually occur.

iii. Poor management performance.

iv. The expense risks associated with fluctuating amounts of new business.

v. Risks associated with future economic viability of the company.

vi. Moral hazards.

vii. Fraud and theft.

DSection 2: Scope and Effective Date

A. Scope

Commented [HR13]: The analogous items have a period at the end.

Attachment Four

70

5

1. The following categories of annuities or product features issued on or after the operative date of the Valuation Manual, directly written or assumed through reinsurance, are covered by this section of the Valuation Manual:subject to the requirements of VM-21:

a. Variable deferred annuity contracts subject to the CARVM, whether or not such contracts contain GMDBs or VAGLBs.

b. Variable immediate annuity contracts, whether or not such contracts contain GMDBs or VAGLBs.

c. GroupAny group annuity contracts that are not subject to CARVM, but containcontract which containings guarantees similar in nature to GMDBs, VAGLBs or any combination thereof.

Guidance Note: The term “similar in nature” as used in Section D2.A.1.c and Section D2.A.1.d is intended to capture current products and benefits, as well as product and benefit designs that may emerge in the future. Examples of the currently known designs are listed in Section D2.A.1.d. Any product or benefit design that does not clearly fit the scope should be evaluated on a case-by-case basis taking into consideration factors that include, but are not limited to, the nature of the guarantees, the definitions of GMDB and VAGLB in Section E.1.a and Section E.1.bVM-01, and whether the contractual amounts paid in the absence of the guarantee are based on the investment performance of a market-value fund or market-value index (whether or not part of the company’s separate account).

d. AllAny other products thatpolicy or contract which contains guarantees similar in nature to GMDBs or VAGLBs, even if the insurer does not offer the mutual funds or, variable funds, or other supporting investments to which these guarantees relate, where there is no other explicit reserve requirement. If such a benefit is offered as part of a contract that has an explicit reserve requirement and that benefit does not currently have an explicit reserve requirement:

i. These requirements shall be applied to the benefit on a stand-alone basis (i.e., for purposes of the reserve calculation, the benefit shall be treated as a separate contract).

ii. The reserve for the underlying contract, excluding any benefits valued under i above, is determined according to the explicit reserve requirement.

iii. The reserve held for the contract shall be the sum of i and ii.

Guidance Note: For example, a group life contract that wraps a GMDB around a mutual fund generally would fall under the scope of these requirements since there is not an explicit reserve requirement for this type of group life contract. However, for an individual variable life contract with a GMDB and a benefit similar in nature to a VAGLB, the requirements generally would apply only to the VAGLB-type benefit, since there is an explicit reserve requirement that applies to the variable life contract and the GMDB.

2. These requirements do not apply to contracts falling under the scope of theVM- A–255: Modified Guaranteed Annuity Model Regulation (#255);Annuities; however, they do apply to contracts listed above that include one or more subaccounts containing features similar in nature to those contained in modified guaranteed annuities (MGAs) (e.g., market value adjustments).

Commented [LE14]: Grammar / wordsmith

Commented [LE15]: This reference doesn’t seem right to me. Should it refer to the definitions of VAGLBs that have now been moved to VM-01? Or is it referring to the example in the Guidance Note below? I would consider that “one example” rather than “a list”. Or perhaps it should refer to the Guidance Note in Section 2.A.3?

Commented [HR16]: Changed from a plural to singular noun, so add an “s” or we can change to “policies or contracts”. Or change “which contain” to “containing” which can pair with either singular or plural.

Commented [HR17]: Extra space

Attachment Four

71

6

3. Separate account productscontracts that guarantee an index and do not offer GMDBs or VAGLBs are excluded from the scope of these requirements.

Guidance Note: Current VAGLBs include Guaranteed Minimum Accumulation Benefits, Guaranteed Minimum Income Benefits, Guaranteed Minimum Withdrawal Benefits, Guaranteed Lifetime Withdrawal Benefits and Guaranteed Payout Annuity Floors. These requirements will be applied to future variations on these designs and to new guarantee designs.

E. Definitions

1. Definitions of Benefit Guarantees

a. The term “guaranteed minimum death benefit” (GMDB) means a guaranteed benefit providing, or resulting in the provision that, an amount payable on the death of a contract holder, annuitant, participant or insured will be increased and/or will be at least a minimum amount. Only such guarantees having the potential to produce a contractual total amount payable on death that exceeds the account value—or in the case of an annuity providing income payments, an amount payable on death other than continuation of any guaranteed income payments—are included in this definition. GMDBs that are based on a portion of the excess of the account value over the net of premiums paid less partial withdrawals made (e.g., an earnings enhanced death benefit) are also included in this definition.

b. The term “variable annuity guaranteed living benefit” (VAGLB) means a guaranteed benefit providing, or resulting in the provision that, one or more guaranteed benefit amounts payable or accruing to a living contract holder or living annuitant, under contractually specified conditions (e.g., at the end of a specified waiting period, upon annuitization or upon withdrawal of premium over a period of time) will increase contractual benefits should the contract value referenced by the guarantee (e.g., account value) fall below a given level or fail to achieve certain performance levels. Only such guarantees having the potential to provide benefits with a present value as of the benefit commencement date that exceeds the contract value referenced by the guarantee are included in this definition. Payout annuities without minimum payout or performance guarantees are neither considered to contain nor to be VAGLBs.

c. The term “guaranteed minimum income benefit” (GMIB) means a VAGLB design for which the benefit is contingent on annuitization of a variable deferred annuity or similar contract. The benefit is typically expressed as a contract-holder option, on one or more option dates, to have a minimum amount applied to provide periodic income using a specified purchase basis.

d. The term “guaranteed payout annuity floor” (GPAF) means a VAGLB design guaranteeing that one or more of the periodic payments under a variable immediate annuity will not be less than a minimum amount.

2. Definitions of Reserve Methodology Terminology

a. The term “scenario” means a set of asset growth rates and investment returns from which assets and liabilities supporting a set of contracts may be determined for each year of a projection.

b. The term “cash surrender value” means, for purposes of these requirements, the amount available to the contract holder upon surrender of the contract. Generally, it is equal to the account value less any applicable surrender charges, where the surrender charge reflects the availability of any free partial surrender options. For contracts where all or a portion of the amount available to the contract holder upon surrender is subject to a market value

Attachment Four

72

7

adjustment, however, the cash surrender value shall reflect the market value adjustment consistent with the required treatment of the underlying assets. That is, the cash surrender value shall reflect any market value adjustments where the underlying assets are reported at market value, but shall not reflect any market value adjustments where the underlying assets are reported at book value.

c. The term “scenario greatest present value” means the sum, for a given scenario, of:

i. The greatest of the present values, as of the projection start date, of the projected accumulated deficiencies for the scenario.

ii. The starting asset amount.

d. The term “conditional tail expectation (CTE) amount” means an amount equal to the numerical average of the 30% largest values of the scenario greatest present values.

e. The term “working reserve” means the assumed reserve used in the projections of accumulated deficiencies supporting the calculation of the scenario greatest present values. At any point in the projections, including at the start of the projection, the working reserve shall equal the projected cash surrender value.

For a variable payout annuity without a cash surrender value, the working reserve shall equal the present value, at the valuation interest rate and the valuation mortality table specified for such a product by Model #820, of future income payments projected using a return based on the valuation interest rate less appropriate asset-based charges. For annuitizations that occur during the projection, the valuation interest rate as of the current valuation date may be used in determining the working reserve. Alternatively, if an integrated model of equity returns and interest rates is used, a future estimate of valuation interest rates may be incorporated into the working reserve.

For contracts not covered above, the actuary shall determine the working reserve in a manner that is consistent with the above requirements.

f. The term “accumulated deficiency” means an amount measured as of the end of a projection year and equals the projected working reserve less the amount of projected assets, both as of the end of the projection year. Accumulated deficiencies may be positive or negative.

Guidance Note: A positive accumulated deficiency means there is a cumulative loss, and a negative accumulated deficiency means there is a cumulative gain.

g. The term “starting asset amount” means an amount equal to the value of the assets at the start of the projection, as defined in Section 3.D.1.

h. The term “anticipated experience” means the actuary’s reasonable estimate of future experience for a risk factor given all available, relevant information pertaining to the contingencies being valued.

i. The term “prudent estimate” means the basis upon which the actuary sets the deterministic assumptions to be used for projections. A prudent estimate assumption is to be set at the conservative end of the actuary’s confidence interval as to the true underlying probabilities for the parameter(s) in question, based on the availability of relevant experience and its degree of credibility.

Attachment Four

73

8

A prudent estimate assumption is developed by applying a margin for uncertainty to the anticipated experience assumption. The margin for uncertainty shall provide for estimation error and margins for adverse deviation. The resulting prudent estimate assumption shall be reasonably conservative over the span of economic cycles and over a plausible range of expected experience, in recognition of the principles described in Section 1.B. Recognizing that assumptions are simply assertions of future unknown experience, the margin should be directly related to uncertainty in the underlying risk factor. The greater the uncertainty, the larger the margin. Each margin should serve to increase the aggregate reserve that would otherwise be held in its absence (i.e., using only the anticipated experience assumption).

For example, assumptions for circumstances that have never been observed require more margins for error than those for which abundant and relevant experience data are available.

This means that valuation assumptions not stochastically modeled are to be consistent with the stated principles in Section 1.B, be based on any relevant and credible experience that is available, and should be set to produce, in concert with other prudent estimate assumptions, a CTE amount that is consistent with the stated CTE level.

The actuary shall follow the principles discussed in Section 11 and Section 12 in determining prudent estimate assumptions.

j. The term “gross wealth ratio” means the cumulative return for the indicated time period and percentile (e.g., 1.0 indicates that the index is at its original level).

k. The term “clearly defined hedging strategy” is a designation that applies to strategies undertaken by a company to manage risks through the future purchase or sale of hedging instruments and the opening and closing of hedging positions. In order to qualify as a clearly defined hedging strategy, the strategy must meet the principles outlined in Section 1.B (particularly Principle 5) and shall, at a minimum, identify:

i. The specific risks being hedged (e.g., delta, rho, vega, etc.).

ii. The hedge objectives.

iii. The risks not being hedged (e.g., variation from expected mortality, withdrawal, and other utilization or decrement rates assumed in the hedging strategy, etc.).

B. Effective Date and Phase -in

Option 1:

These requirements apply for valuation dates on or after January 1, 2020. A company may elect to phase in these requirements over a 36-month period beginning January 1, 2020. A company may elect a longer phase-in period, as long as up to 7 years, with approval of the domiciliary commissioner. The election of whether to phase in and the period of phase-in must be made prior to the 12/31/20December 31, 2020 valuation. At the company’s option, Aa phase-in may be terminated prior to the originally elected end of the phase-in period of phase-in at the company’s election; the reserve would then be equal to the unadjusted reserve calculated according to the then-current requirements of VM-21 applicable for valuation dates on or after January 1, 2020. The method to be used for the phase-in calculation is as follows:

1. Compute R1 = the reserves as of the valuation date following the applicable VM-21 requirements applicable for valuation dates on or after January 1, 2020 for all business in-force on the valuation date,

Commented [LE18]: Add dash between “Phase-in”

Commented [JW19]: ACLI comment #1: The ACLI supports the Phase-in Option 2, which is the simplified approach amortizing into income the initial dollar difference in reserves as of January 1, 2020. a. The purpose of a phase-in is to allow the company time to modify its product management before fully reflecting the financial impacts of the new framework. Neither approach produces a reserve that is “more correct” than the other. It is simply an organized transition between an existing method and a new method. b. The biggest problem that the Oliver Wyman study had identified with the old method is significant non-economic volatility, which was identified to be the primary reason for the use of captives. The new method has eliminated much of the non-economic volatility. c. If the phase-in relies on calculations from both the old method and the new method at each valuation date (Option 1), the non-economic volatility of the old method will continue to influence the required reserve during the phase-in period. We don’t believe this would be a good outcome. d. The recommendation Option 2 leverages the fact that the new method is the better method, then provides a simpler computation and smoother transition to this better method. e. Option 2 mitigates some of the duplicative effort that would exist under Option 1 for both regulators (audit/examination of two complex calculations and their interaction to understand reserves) and companies (administration and documentation of two modeling systems).

Commented [LE20]: Delete extra spaces

Commented [LE21]: wordsmith

Commented [HR22]: Consistency with other date formats shown here.

Commented [HR23]: Otherwise it sounds like we are ending phase-in before phasing in ends. Confusing. Other edits to this sentence for clarity.

Commented [LE24]: “then-current”, “applicable”, “these” – all so confusing. If all these terms refer to the same thing, we should stick with one term to avoid confusion, or my suggestion is to refer back to the effective date of the requirements or VM version of the requirements.

Commented [HR25]: Would it be more precise and easier to just say “R1 below”?

Commented [LE26]: Delete extra space

Commented [LE27]: Delete “s” at the end of “reserves”

Attachment Four

74

9

2. Separately, compute R2 = the reserves as of the valuation date following the calculation VM-21 requirements applicable from VM-21 in the 2019 NAIC Valuation Manual for the same in-force contracts used to compute R1, and

3. Compute the reported reserve as follows:

Reserve on a the valuation date = (A•*R1 + (B-A) •*R2)/B, where

• A is the number of months that has elapsed since December 31, 2019. For example, for the March 31, 2020 valuation, A = 3.

• B = 36 unless the company has obtained approval for a longer phase-in, in which case B = the number of months of approved phase-in.

Option 2:

These requirements apply for valuation dates on or after January 1, 2020. A company may elect to phase in these requirements over a 36-month period beginning January 1, 2020. A company may elect a longer phase-in period, as long as 7 years, with approval of the domiciliary commissioner. The election of whether to phase in and the period of phase-in must be made prior to the December 31, 2020 12/31/20 valuation. At the company’s option, aA phase-in may be terminated prior to the originally elected end of the phase-in period of phase-in at the company’s election; the reserve would then be equal to the unadjusted reserve calculated according to the then-current requirements of VM-21 applicable for valuation dates on or after January 1, 2020. The method to be used for the phase-in calculation is as follows:

1. Compute R1 = the reserves as of January 1, 20201/1/20 following these VM-21 requirements applicable for valuation dates on or after January 1, 2020 for all business in-force on the valuation date., The inforce used should include any reinsurance that is expected to be recaptured during 2020.

2. Separately, compute R2 = the reserves as of January 1, 20201/1/20 following the calculation VM-21 requirements applicable from VM-21 in the 2019 NAIC Valuation Manual for the same in-force contracts used to compute R1, and,

3. Determine the change in reserve requirements as C = R1 minus R2.

43. Compute the reported reserve on any valuation dates as follows:

Reported Reserve on a the valuation date = Reserve - (B-A) •*C)/B, where

• A is the number of months that has elapsed since December 31, 2019. For example, for the March 31, 2020 valuation, A = 3.

• B = 36 unless the company has obtained approval for a longer phase-in, in which case B = the number of months of approved phase-in.

• C = R1 minus R2.

A company may elect to apply these requirements applicable for valuation dates on or after January 1, 2020 as the NAIC requirements for the valuation on December 31, 2019. Any company so electing may not also elect theto phase-in period defined abovethese requirements.

Commented [HR28]: Suggest adding a comma here.

Commented [LE29]: Delete “s” at the end of “reserves”

Commented [LE30]: Change wording for consistency between R1 and R2 definitions

Formatted: Font: Italic

Commented [HR31]: “Valuation Manual” gets italicized throughout.

Commented [LE32]: Change “a valuation date” to “the valuation date”

Commented [HR33]: The VM sometimes used an asterisk and sometimes a dot for multiplication. Trying to consistently use dots now.

Commented [HR34]: Reads better with “the”.

Commented [HR35]: Add period.

Commented [HR36]: Consistency with other date formats shown here.

Commented [HR37]: Otherwise it sounds like we are ending phase-in before phasing in ends. Confusing. Other edits to this sentence for clarity.

Commented [LE38]: This wording is exactly the same between “Option 1” and “Option 2”. I recommend pulling this out of either option, and putting in before the “Option 1” header. Then to close the paragraph, say: “There are 2 options for the method to be used for the phase-in calculation. The 2 options are as follows:”. Then go to the “Option 1” header and straight into the calculation.

Commented [LE39]: Delete “s”

Commented [LE40]: “these” is unclear to me and should be clarified.

Formatted: Font: Times New Roman

Commented [HR41]: Suggest adding a comma here.

Commented [LE42]: Delete “s”

Formatted: Font: Italic

Commented [HR43]: Italicize Valuation Manual

Commented [LE44]: Period instead of comma


Commented [HR45]: Since adding “any”, delete the s.

Commented [LE46]: These changes made for consistency between Option 1 and Option 2

Commented [LE47]: What is this reserve supposed to be?

Commented [HR48]: Consistency

Commented [HR49]: It would read more easily if we defined C in a bullet below rather than in item #3 above.

Commented [HR50]: Unpaired parenthesis here.

Commented [HR51]: It does not seem right to say “NAIC requirements”. What do you really want to say here? “reserve requirements”? “Valuation Manual requirements”?

Commented [HR52]: Suggest so this reads more easily.

Attachment Four

75

10

Section 3iv. The financial instruments that will be used to hedge the risks.

v. The hedge trading rules, including the permitted tolerances from hedging objectives.

vi. The metric(s) for measuring hedging effectiveness.

vii. The criteria that will be used to measure hedging effectiveness.

viii. The frequency of measuring hedging effectiveness.

ix. The conditions under which hedging will not take place.

x. The person or persons responsible for implementing the hedging strategy. The hedge strategy may be dynamic, static or a combination thereof.

It is important to note that strategies involving the offsetting of the risks associated with variable annuity guarantees with other products outside of the scope of these requirements (e.g., equity-indexed annuities) do not currently qualify as a clearly defined hedging strategy under these requirements.

l. The term “revenue sharing,” for purposes of these requirements, means any arrangement or understanding by which an entity responsible for providing investment or other types of services makes payments to the company (or to one of its affiliates). Such payments are typically in exchange for administrative services provided by the company (or its affiliate), such as marketing, distribution and recordkeeping. Only payments that are attributable to charges or fees taken from the underlying variable funds or mutual funds supporting the contracts that fall under the scope of these requirements shall be included in the definition of revenue sharing.

m. The term “domiciliary commissioner,” for purposes of these requirements, means the chief insurance regulatory official of the state of domicile of the company.

n. The term “aggregate reserve” means the minimum reserve requirement as of the valuation date for the contracts falling within the scope of these requirements.

o. The term “1994 Variable Annuity Minimum Guaranteed Death Benefits (MGDB) Mortality Table” means the mortality table shown in Appendix 1.

Section 2: Reserve Methodology

A. General Description

The aggregate reserve for contracts falling within the scope of these requirements shall equal the CTE amount but not less than the standard scenario amount, where the aggregatestochastic reserve is calculated as the standard scenario amount(following the requirements of Section 4) plus the excess, if any, of additional standard projection amount (following the requirements of Section 6) less the PIMR included in the CTE amount over the standard scenario amount.starting assets.

B. Impact of Reinsurance Ceded

Where reinsurance is ceded for all or a portion of the contracts, bothall components in the above general description (and thus the aggregate reserve) shall be determined post-reinsurance ceded, that is, net of any reinsurance treaties that meet the statutory requirements that would allow the treaty to be accounted for as reinsurance.

Formatted: Font: Times New Roman, Bold

Commented [JW53]: ACLI comment #2: As a result of the direction of the changes to the Alternative Methodology, Section 3 A, C, and D need additional edits beyond those in the exposure, as outlined in Appendix A.

Commented [JW54]: Appendix F of ACLI comment letter contains additional non-substantive suggested edits to Section 3.A and 3.D

Commented [HR55]: Replace with draft Exposed with Section 7

Commented [LE56]: Do we need to change this reference to “net”? Or I have suggested clarifying that “net” is “post-reinsurance ceded”.

Attachment Four

76

11

An aggregate reserve before pre-reinsurance alsoceded shall also be calculated if needed for regulatory reporting or other purposes, using methods described in Section 45.

C. The Additional Standard ScenarioProjection Amount

The additional standard scenarioprojection amount is the aggregate of the reservesan additive factor, determined by applying one of the two standard scenarioprojection methods defined in Section 6, that is applied to the stochastic reserve to determine the aggregate reserve. The same method to each of the must be used for all contracts falling within the scope of these requirements. The a group of contracts that are aggregated together to determine the reserve. The company shall elect which method they will use to determine the additional standard scenario method is outlined in Section 5.projection amount. The company may not change that election for a future valuation without the approval of the domiciliary commissioner.

D. The CTE AmountStochastic Reserve

The CTE amountstochastic reserve shall be determined based on a projectionprojections of the contracts falling within the scope of these requirements, and the assets supporting these contracts, over a broad range of stochastically generated projection scenarios described in Section 8 and using prudent estimate assumptions. as required herein by this VM-21.

The stochastically generated projection scenarios shall meet the scenario calibration criteria described in Section 7.

The CTE amountThe stochastic reserve may be determined in aggregate for all contracts falling within the scope of these requirements (i.e., a single grouping). At) or, at the option of the company, it may be determined by applying the methodology outlined below to subgroupingssub-groupings of contracts, in which case the CTE amountstochastic reserve shall equal the sum of the amounts computed for each such subgrouping.

The CTE amount stochastic reserve for any group of contracts shall be determined using theas the CTE70 value of all scenario reserves computed following steps:

1. For each scenario, projected aggregate accumulated deficiencies are determined at the start of the projection (i.e., “time 0”) and at the end the requirements of each projection year as the sum of the accumulated deficiencies for each contract grouping.

The scenario greatest present value is determined for each scenario based on the sum of the aggregate accumulated deficiencies and aggregate starting asset amounts for the contracts for which the aggregate reserve is being computed.

Guidance Note: The scenario greatest present value is, therefore, based on the greatest projected accumulated deficiency, in aggregate, for all contracts for which the aggregate reserve is computed hereunder, rather than based on the sum of the greatest projected accumulated deficiency for each grouping of contracts.

2. The scenario greatest present values for all scenarios are then ranked from smallest to largest, and the CTE amount is the average of the largest 30% of these ranked values.

The projections shall be performed in accordance with Section 3. The actuary shall document the assumptions and procedures used for the projections and summarize the results obtained as described in Section 4 and Section 10..

E. Alternative Methodology (subject to further review)

Commented [HR57]: Since removing “if needed”, the rest of this seems unnecessary too.

Commented [JW58]: [5. JR Comment: the notion that pre- and post-reinsurance are using different methods is false. The methods are the same; the inputs are different.]

Commented [LE59]: This was in the original OW draft edits, not sure why it was taken out. I think it should be in there because it seems odd to say it is an additive factor, but not say what it is added to and why.

Commented [HR60]: Replace with Section reference?

Commented [LE61R60]: The “this” seems weird. Suggest replacing with “herein” if don’t want to add Section reference.

Commented [HR62]: Subgroup is not normally hyphenated in the VM.

Commented [HR63]: VM-20 uses the notation “CTE 70” (with a space). VM-21 uses it without a space “CTE70” and sometimes with parentheses “CTE(65)”. We should pick one and then be consistent throughout. To change things in VM-21 to be consistent with VM-20, we would need to modify: Sects 1-5: p. 11 of 27 (1 place), p. 14 of 27 (3 places) Sects 3, 7: p. 10 of 17 (4 places) Sect 6: pp. 5-7 (13 places) Sect 8: None Sect 9: p. 1 (3 places), pp. 4-5 (12 places), p. 6 (2 places), p. 7 (5 places) Sect 11-12: None Also shows up in footnote 1

Commented [LE64]: Seems like clarifying language was needed here.

Commented [HR65]: Extra space.

Commented [JW66]: [6. JR Comment: since the stochastic reserve is a weighted average of two CTE70’s in the case of a CDHS, this statement is no longer true. Replace with “The stochastic reserve for any group of contracts shall be determined following Section 4.”]

Commented [HR67]: Missing a space.

Attachment Four

77

12

For a group of variable deferred annuity contracts that contain either no guaranteed benefits or only GMDBs (i.e., no VAGLBs), the CTE amountstochastic reserve may be determined using the alternative methodology described in Section 67 rather than using the approach described in Section 23.D. However, in the event the approach described in Section 23.D has been used in prior valuations for that group of contracts, the Alternative Methodology may not be used without approval from the domiciliary commissioner.

The CTE amountstochastic reserve for the group of contracts to which the Alternative Methodology is applied shall not be less than the aggregate cash surrender value of those contracts.

The actuary shall document the assumptions and procedures used for the Alternative Methodology and summarize the results obtained as described in Section 4 and Section 10.

F. Allocation of Results to Contracts

The aggregate reserve shall be allocated to the contracts falling within the scope of these requirements using the method outlined in Section 89.

G. Reserve to Be Held in the General Account

The portion of the aggregate reserve held in the general account shall not be less than the excess of the aggregate reserve over the aggregate cash surrender value held in the separate account and attributable to the variable portion of all such contracts. For contracts for which a cash surrender value is not defined, the company shall substitute for cash surrender value held in the separate account the implicit amount for which the contract-holder is entitled to receive income based on the performance of the separate account. For example, for a variable payout annuity for which a specific number of units is payable, the implicit amount could be the present value of that number of units, discounted at the Aassumed iInvestment Rreturn and defined mortality, times the unit value as of the valuation date.

G.H. Documentation

A qualified actuary to whom responsibility for this group of contracts is assigned shall document the development of the reserves and provide the required certifications following the requirements of VM-31.

Section 34: Determination of CTE Amount Based on ProjectionsStochastic Reserve

A. Projection of Accumulated Deficiencies

1. General Description of Projection

The projection of accumulated deficiencies shall be made ignoring federal income tax in both cash flows and discount rates and reflect the dynamics of the expected cash flows for the entire group of contracts, reflecting all product features,— including theany guarantees provided under the contracts. Insurance company expenses (including overhead and investment expense), fund expenses, contractual fees and charges, revenue-sharing income received by the company (net of applicable expenses), and cash flows associated with any reinsurance or hedging instruments are

Guidance Nnote: this approach is equivalent to assuming that the separate account performance is equal to the Aassumed iInvestment rReturn.

Commented [JW68]: [7. JR Comments: 1. Section 3.D does not describe an approach, it simply points to Section 4, where the approach is described. Therefore the references to “Section 3.D” should be replaced by “Section 4” (two places). 2. The term “stochastic reserve” should not be used to describe the result of the Alternative Methodology.]

Commented [HR69]: Replace with draft Exposed with Section 7

Commented [LE70]: Delete extra space

Commented [LE71]: Lower case

Commented [LE72]: Capital “N” in “Guidance Note”, also lower case “air”

Commented [JW73]: [8. JR Comment: Can this be simplified to “The portion of the aggregate reserve held in the general account shall not be less than the excess of the aggregate reserve over the statement value of the assets held in the separate account on the valuation date”?]

Commented [HR74]: If deleting the dash, need a comma.

Attachment Four

78

13

to be reflected on a basis consistent with the requirements herein. Cash flows from any fixed account options also shall be included. Any market value adjustment assessed on projected withdrawals or surrenders also shall be included (whether or not the cash surrender value reflects market value adjustments). Throughout the projection, where estimates are used, such estimates shall be on a prudent estimate basisThroughout the projection, all assumptions shall be determined based on the requirements hereinthe requirements of this VM-21 as guided by Principle 3. Accumulated deficiencies shall be determined at the end of each projection year as the sum of the accumulated deficiencies for all contracts within each contract grouping.

Federal income tax shall not be included in the projection of accumulated deficiencies.

2. Grouping of Variable Funds and Subaccounts

The portion of the starting asset amount held in the separate account represented by the variable funds and the corresponding account values may be grouped for modeling using an approach that recognizes the investment guidelines and objectives of the funds. In assigning each variable fund and the variable subaccounts to a grouping for projection purposes, the fundamental characteristics of the fund shall be reflected, and the parameters shall have the appropriate relationship to the required calibration points of the S&P 500.stochastically generated projection scenarios described in Section 8. The grouping shall reflect characteristics of the efficient frontier (i.e., returns generally cannot be increased without assuming additional risk).

An appropriate proxy fund for each variable subaccount shall be designed in order to develop the investment return paths. The development of the scenarios for the proxy funds is a fundamental step in the modeling and can have a significant impact on results. As such, the actuarycompany must map each variable account to an appropriately crafted proxy fund normally expressed as a linear combination of recognized market indices (or, sub-indices). or funds.

3. Grouping of Contracts

Projections may be performed for each contract in force on the date of valuation or by groupingassigning contracts into representative cells of model plans using all characteristics and criteria having a material impact on the size of the reserve. Grouping shall be the responsibility of the actuary butAssigning contracts to model cells may not be done in a manner that intentionally understates the resulting reserve.

4. Modeling of Hedges

a. For a company that does not have a CDHS:

i. tThe company shall not consider the cash flows from any future hedge purchases or any rebalancing of existing hedge assets in its modeling.

ii. Existing hedging instruments that are currently held by the company in support of the contracts falling under the scope of these requirements shall be included in the starting assets. The hedge assets may then be considered in one of two ways:

a) Include the asset cash flows from any contractual payments and maturity values in the projection model, or

b) No hedge positions – in which caseReplace the hedge positions held on the valuation date are replaced with cash and/or other general account assets in an amount equal to the aggregate market value of these hedge positions. The cash may then be invested following the company’s investment strategy.

Commented [HR75]: Awkward phrasing.

Commented [JW76]: [9. JR Comments: 1. Concerning the second to last sentence: Principle 3 is not the only Principle addressing assumptions. I suggest deleting the sentence as unnecessary. 2. Concerning the last sentence, is there a presumption that accumulated deficiencies are calculated for each contract? If so, it would not be sufficient for the model office approach.]

Commented [JW77]: [10. JR Comment: This would be a good place to introduce the term “model office”, since it is an important consideration for the standard scenario projection.]

Commented [HR78]: Capitalize

Commented [LE79]: Delete space

Commented [HR80]: Does this presume a time zero purchase? Could it be made more clear? How does this differ from how the company may model any other cash presumed to be held at time 0 in the projection?

Attachment Four

79

14

A company may switch from method a) to b) at any time, but may only change from b) to a) with approval of the domiciliary commissioner.

b. For a companmy with a CDHS, Tthe detailed requirements for the modeling of hedges are defined in Section 9. The following paragraphs are an overviewa high level summary and do not supersede the detailed requirements.

ia. The appropriate costs and benefits of hedging instruments that are currently held by the company in support of the contracts falling under the scope of these requirements shall be included in the projections. used in the determination of the stochastic reserve.

iib. If the company is following a clearly defined hedging strategy and the hedging strategy meets the requirements of Section 910, tThe projections shall take into account the appropriate costs and benefits of hedge positions expected to be held in the future through the execution of that strategya CDHS. Because models do not always accurately portray the results of hedge programs, the company shall, through back-testing and other means, assess the accuracy of the hedge modeling. The company shall determine a stochastic reserve as the weighted average of two CTE values ; first, a CTE70 (“best efforts”) representing a company’s projection of all of the hedge cash flows including future hedge purchases, and a second CTE70 (“adjusted”) which shall use only hedge assets held by the company on the valuation date and no future hedge purchases. These are described more fully in Section [109]. The stochastic reserve shall be the weighted average of the two CTE70 values, where the weights reflect the error factor (E) determined following the guidance of Section [109.xC.xx4.]

Tostochastic reserve = CTE70 (“best efforts”) + E * Max (0,(CTE70 (“adjusted”) – CTE70 (“best efforts”)))

When computing CTE70 (“adjusted”), (see Section 10.X.x.), the degree either company should reflect one of the currently heldfollowing in the starting general account assets:

i. Any hedge positionsassets meeting the requirements described in Section 4.A.4.a.; or

ii. the Cash or other general account assets in an amount equal to the aggregate market value of the hedge positions expected to be held in the future introduce basis, gap, price or assumption risk, a suitable reduction for effectiveness of hedges shall be made. assets meeting the requirements described in Section 4.A.4.a.

The actuarycompany may switch from i. to ii. at its discretion, but may not switch from ii. to i. without approval from the domiciliary commissioner.

iiic. The company is responsible for verifying compliance with a clearly defined hedging strategyCDHS requirements and theany other requirements in Section 910 for all hedge instruments included in the projections.

While hedging strategies may change over time, any change in hedging strategy shall be documented and include an effective date of the change in strategy.

Commented [HR81]: Awkward phrasing

Commented [JW82]: [11. JR Comment: Contrary to the foregoing, the following comments constitute much more than an overview, and some will be repeated in Section 9. Therefore, consider retaining only statements that will not be covered in Section 9.]

Commented [HR83]: We deleted the reference to a particular strategy, so “that strategy” is now ambiguous.

Commented [HR84]: Correct Reference



Commented [HR87]: Section 9?

Attachment Four

80

15

ivd. The use of products not falling under the scope of these requirements (e.g., equity-indexed annuities) as a hedge shall not be recognized in the determination of accumulated deficiencies.

ce. These requirements govern the determination of contractminimum reserves and do not supersede any statutes, laws or regulations of any state or jurisdiction related to the use of derivative instruments for hedging purposes and should not be used in determining whether a company is permitted to use such instruments in any state or jurisdiction.

f. Upon request of the company’s domiciliary commissioner and for information purposes to show the effect of including future hedge positions in the projections, the company shall showdisclose the results of performing an additional set of projections reflecting only the hedges currently held by the company in support of the contracts falling under the scope of these requirements. Because this additional set of projections excludes some or all of the derivative instruments, the investment strategy used may not be modified to reflect the same as that used in the determinationabsence of the CTE amount.those future hedge instruments.

5. Revenue Sharing

a. Projections of accumulated deficiencies may include income from projected future revenue-sharing, net of applicable projected expenses (net revenue-sharing income) if each of the following requirements are met:

i. The net revenue-sharing income is received by the company.

Guidance Note: For purposes of this section, net revenue-sharing income is considered to be received by the company if it is paid directly to the company through a contractual agreement with either the entity providing the net revenue-sharing income or an affiliated company that receives the net revenue-sharing income. Net revenue-sharing income also would be considered to be received if it is paid to a subsidiary that is owned by the company and if 100% of the statutory income from that subsidiary is reported as statutory income of the company. In this case, the actuarycompany needs to assess the likelihood that future net revenue-sharing income is reduced due to the reported statutory income of the subsidiary being less than future net revenue-sharing income received.

ii. Signed contractual agreement or agreements are in place as of the valuation date and support the current payment of the net revenue-sharing income.

iii. The net revenue-sharing income is not already accounted for directly or indirectly as a company asset.

b. The amount of net revenue-sharing income to be used shall reflect the actuary’scompany’s assessment of factors that include, but are not limited to, the following (not all of these factors will necessarily be present in all situations):

i. The terms and limitations of the agreement(s), including anticipated revenue, associated expenses and any contingent payments incurred or made by either the company or the entity providing the net revenue-sharing as part of the agreement(s).

ii. The relationship between the company and the entity providing the net revenue-sharing income that might affect the likelihood of payment and the level of expenses.

Commented [JW88]: [12. JR Comments: 1. Which requirements are “These requirements”? 2. Why might they be interpreted as superseding state laws as indicated?]

Commented [LE89]: All?

Attachment Four

81

16

iii. The benefits and risks to both the company and the entity paying the net revenue-sharing income of continuing the arrangement.

iv. The likelihood that the company will collect the net revenue-sharing income during the term(s) of the agreement(s) and the likelihood of continuing to receive future revenue after the agreement(s) has ended.

v. The ability of the company to replace the services provided to it by the entity providing the net revenue-sharing income or to provide the services itself, along with the likelihood that the replaced or provided services will cost more to provide.

vi. The ability of the entity providing the net revenue-sharing income to replace the services provided to it by the company or to provide the services itself, along with the likelihood that the replaced or provided services will cost more to provide.

c. The amount of projected net revenue-sharing income also shall reflect a margin (which decreases the assumed net revenue-sharing income) directly related to the uncertainty of the revenue. The greater the uncertainty, the larger the margin. Such uncertainty is driven by many factors, including the potential for changes in the securities laws and regulations, mutual fund board responsibilities and actions, and industry trends. Since it is prudent to assume that uncertainty increases over time, a larger margin shall be applied as time that has elapsed in the projection increases.

d. All expenses required or assumed to be incurred by the company in conjunction with the arrangement providing the net revenue-sharing income, as well as any expenses assumed to be incurred by the company in conjunction with the assumed replacement of the services provided to it (as discussed in Section 34.A.5.b.v), shall be included in the projections as a company expense under the requirements of Section 34.A.1. In addition, expenses incurred by either the entity providing the net revenue-sharing income or an affiliate of the company shall be included in the applicable expenses discussed in Section 34.A.1 and Section 34.A.5.a that reduce the net revenue-sharing income.

e. The actuarycompany is responsible for reviewing the revenue-sharing agreements, and verifying compliance with these requirements and documenting the rationale for any source of net revenue-sharing income used in the projections..

f. The amount of net revenue-sharing income assumed in a given scenario shall not exceed the sum of (i) and (ii), where:

(i) Is the contractually guaranteed net revenue-sharing income projected under the scenario.

(ii) Is the actuary’scompany’s estimate of non-contractually guaranteed net revenue-sharing income before reflecting any margins for uncertainty multiplied by the following factors:

a) 1.000 in the first projection year.

b) 0.995 in the second projection year.

c) 0.890 in the third projection year.

d) 0.785 in the fourth projection year.

Attachment Four

82

17

e) 0.680 in the fifth projection year.

f) 0.5 in the sixth and all subsequent projection years. The resulting amount of non-contractually guaranteed net revenue-sharing income after application of this factor shall not exceed 0.25% per year on separate account assets in the sixth and all subsequent projection years.

6. Length of Projections

Projections of accumulated deficiencies shall be run for as many future years as needed so that no materially greater reserve value would result from longer projection periods.

7. Asset Valuation Reserve (AVR/)/Interest Maintenance Reserve (IMR)

The AVR and the IMR shall be handled consistently with the treatment in the company’s cash-flow testing.

B. Determination of Scenario Greatest Present ValuesReserve

1. Scenario Greatest Present Values

1. General

For a given scenario, the scenario greatest present valuereserve is the sum of:

a. The greatest present value, as of the projection start date, of the projected accumulated deficiencies defined in Section 1.E.2.f.; and

b. b. The starting asset amount.

When using the Ddirect iIteration mMethod described in Section 4.B.5, the scenario reserve will equal the final starting asset amount determined according to Section 4.B.5.

The scenario reserve for any given scenario shall not be less than the Ccash Ssurrender Vvalue in aggregate on the valuation date for the group of contracts modeled in the projection.

2. Discount Rates

In determining the scenario greatest present valuesreserve, accumulated deficiencies shall be discounted usingat the same interest rates at which positive cash flows are investednet asset earned rate on additional assets, as determineddefined in Section 4.B.34.D.4. Such interest rates

3. Additional Invested Assets

On the valuation date, the company shall be reduceddetermine the additional invested assets as the amount of assets needed, or an approximation of it, to reflectfund the present value of the accumulated deficiency. These assets may include only (i) general account assets available to the company on the valuation date that do not constitute part of the starting asset amount, and (ii) cash, and shall exclude separate account assets and policy loans. If the company elects not to include certain general account assets, or if the amount of additional available general account assets is lower than the amount needed to fund the present value of the accumulated deficiency, the company shall model cash assets to fill any deficiencies in the amount of additional invested assets. Any cash assumed will then be subject to the company’s investment policy as described in Section 4.D.4.a.

Commented [JW90]: [13. JR Comment: In light of the fact that the RBC charge and the aggregate reserve are on the same distribution, is this criterion sufficient?]

Commented [HR91]: AVR was removed from VM-20 because the prescribed defaults already got you to a CTE70 level to cover this risk. Now that VM-21 is also using the VM-20 prescribed default rates, AVR should be removed here as well. As background from around the time of forming VM-20, see: https://www.actuary.org/files/RBC_AVR_Presentation_to_NAIC_IRBCWG_12-5-13.pdf

Commented [HR92]: See comment above

Commented [JW93]: ACLI comment #3: This section should state that the amounts used should be adjusted to a pre-tax basis. This is consistent with VM-20, which uses a PIMR – PreTax IMR. The specific language is:

The AVR and IMR shall be handled consistently with the treatment in the company’s cash flow testing, and the amounts should be adjusted to a pre-tax basis.

Commented [JW94]: [14. JR Comment: John Bruins, Peter Tian and I developed a replacement for 3. and 4.]

Commented [JW95]: ACLI comment #4: In Section 4.B.3. and 4.B.4., the description of the additional assets and the Net Asset Earned Rate (NAER) is not clear. We have worked with John Robinson, MN, to develop language for those two paragraphs, and the recommended language is included as Appendix B to this letter. See comments submitted from Matt Kauffman, Moody’s Analytics related to these sections

Commented [HR96]: This is unclear. If the intent is to allow for an amount that is not sufficient in all scenarios, some guidance on what is sufficiently close needs to be added. Saying any approximation is allowed is not appropriate.

Attachment Four

83

https://www.actuary.org/files/RBC_AVR_Presentation_to_NAIC_IRBCWG_12-5-13.pdf

https://www.actuary.org/files/RBC_AVR_Presentation_to_NAIC_IRBCWG_12-5-13.pdf

18

At subsequent projection intervals, the “additional invested assets” shall equal the additional invested assets on the valuation date plus any reinvestment assets generated by cash flows from the initial additional invested assets.

4. Net Asset Earned Rate on Additional Assets

The net asset earned rate on additional assets shall represent the ratio of net investment earnings on additional invested assets to the amount of additional invested assets, as defined belowabove. All items reflected in the ratio are consistent with statutory asset valuation and accrual accounting, including reflection of due, accrued, or unearned investment income where appropriate. A vector of NAER’s will be produced for each scenario.

The net asset earned rate on additional assets for each projection interval shall be calculated in a manner that is consistent with the timing of cash flows and length of the projection interval of the related cash flow model. The net investment earnings included in the calculation shall be projected in a manner consistent with Section 4.D.4, reflecting expected credit losses. Note that the interest rates used do not include as prescribed in VM-20 Section 9.F. and anticipated investment expenses but without a reduction for federal income taxes.

5. Direct Iteration

In lieu of the method described in Sections 4.B.2, Section 4.B.3, and Section 4.B.4 above, the company may solve for the amount of starting assets which, when projected along with all contract cash flows, result in the defeasement of all projected future benefits and expenses atby the end of the projection horizonperiod. with no deficiencies at the end of any projection yearinterval during the projection period.

C. Projection Scenarios

1. Minimum Required Number of Scenarios

The number of scenarios for which projected greatest present values of accumulated deficienciesthe scenario reserve shall be computed shall be the responsibility of the actuarycompany and shall be considered to be sufficient if any resulting understatement in total reservesthe stochastic reserve, as compared with that resulting from running additional scenarios, is not material.

2. Economic Scenario Calibration CriteriaGeneration

U.S. ReturnsTreasury interest rate curves, as well as total investment return paths for the groupings of variable fundsgeneral account equity assets and separate account fund performance shall be determined on a stochastic basis suchusing the methodology described in Section 8. If the company uses a proprietary generator to develop scenarios, the company shall demonstrate

Guidance Note: As company’ies manage themselves differently, additional invested assets may include assets earmarked for the VA business, other assets including surplus assets, and cash. A company might start with additional invested assets equaling the total earmarked for the VA business and not in the starting assets, develop vectors of NAER’s by scenario, and model the scenario reserves. If those assets are not sufficient for all scenarios, the company could add assets backing surplus or cash, and update the NAER’s and scenario reserves until a sufficient amount of additional invested assets is determined for each scenario.

Commented [LE97]: Delete quotes, not sure why they are needed.

Commented [LE98]: Moved to “above”

Commented [LE99]: Delete all apostrophes

Commented [HR100]: Delete extra period.

Commented [HR101]: Delete extra “s”.


Formatted: Font: Times New Roman, 11 pt



Commented [HR103]: Suggest for time steps other than annual


Commented [JW104]: Appendix F of ACLI comment letter contains non-substantive suggested edit to 4.B.5.

Commented [JW105]: [15. JR Comment: Note that there are multiple ways to meet this criterion. For example, if a portfolio of amount of $1 million meets this criterion, then so does a portfolio that is 110% of that portfolio (provided it is available).]

Commented [JW106]: [16. JR Comment: What is the difference between an “investment return path” (See Section 4.A.2) and a “total investment return path”?]

Formatted: Font: Times New Roman, Bold






Attachment Four

84

19

that the resulting distribution of the gross wealth ratios of the scenarios meets the scenario calibration criteria specifiedrequirements described in Section 78.

D. Projection Assets

1. Starting Asset Amount

For the projections of accumulated deficiencies, the value of assets at the start of the projection shall be set equal to the approximate value of statutory reserves at the start of the projection. plus the allocated amount of PIMR attributable to the assets selected. Assets shall be valued consistently with their annual statement values, and shall include any hedge assets held in support of the guarantees in the contracts being valued1. The amount of such asset values shall equal the sum of the following items, all as of the start of the projection:

a. All of the separate account assets supporting the contracts.

b. An amount of assets held in the general account equal to the approximate value of statutory reserves as of the start of the projections less the amount in (a).

In many instances,If the amount of initial general account assets may beis negative, resulting inthe model should reflect a projected interest expense. General account assets chosen for use as described above shall be selected on a consistent basis from one reserve valuation hereunder to the next.

Any hedge assets meeting the requirements described in Section 3.A.4 shall be reflected in the projections and included with other general account assets under item (b). To the extent the sum of the value of such hedge assets and the value of, or cash or other general account assets in an amount equal to the aggregate market value of such hedge assets, and the value of separate account assets in item (a)supporting the contracts is greater than the approximate value of statutory reserves as of the start of the projections, then item (b) maythe company shall include enough negative general account assets or cash such that the sum of items (a) and (b)starting asset amount equals the approximate value of statutory reserves as of the start of the projections.

Guidance Note: Further elaboration on potential practices with regard to this issue may be included in a practice note.

The actuary shall document which assets were used as of the start of the projection and the approach used to determine which assets were chosen, as well as verify that the value of the assets equals the approximate value of statutory reserves at the start of the projection.

For an asset portfolio that supports policies and contracts that are: a) subject to, and b) not subject to these requirements, the company shall determine an equitable method to apportion the total amount of starting assets between a. and b.

2. Valuation of Projected Assets

For purposes of determining the projected accumulated deficiencies, the value of projected assets shall be determined in a manner consistent with their value at the start of the projection. However, for derivative instruments that are used in hedging and that are not assumed to be sold during a particular projection interval, the company may account for them at amortized cost in an appropriate manner deemed appropriateelected by the company.

For assets assumed to be purchased during a projection, the value shall be determined in a manner

1 Deferred hedge gains/losses developed under SSAP108 are not included in the value of the starting assets.

Commented [HR107]: Delete “s” since noun is plural.

Commented [JW108]: [17. JR Comment: Insert “Of”.]

Commented [JW109]: Appendix F of ACLI comment letter contains non-substantive suggested edit to 4.D.1.

Commented [JW110]: ACLI comment #5: Add pre-tax AVR, consistent with #3 above. “. . . plus the allocated amount of PIMR and AVR (pre-tax) attributable to the assets selected. “

Commented [HR111]: This sounds like the selected starting assets must include hedge assets. What about cash or general account assets being held in the amount of the hedge assets, in lieu of modeling held hedge assets? Do you mean “including” instead of “and shall include”, so that this says that hedge assets must also be valued consistently?

Commented [HR112]: After deleting the prior sentence, the “such” doesn’t make sense.

Commented [HR113]: Extra space.

Commented [LE114]: I prefer OW’s original wording: “For an asset portfolio that supports both policies and contracts that are subject and not subject to these requirements, the company shall determine an equitable method to apportion the total amount of starting assets between the subject and non-subject policies and contracts.”

Commented [HR115]: “deemed” appropriate is strange phrasing. Is it appropriate or not? We want to say any appropriate manner of amortization is allowed, at the company’s option.

Attachment Four

85

20

consistent with the value of assets at the start of the projection that have similar investment characteristics.

3. Separate Account Assets

For purposes of determining the starting asset amounts in Section 34.D.1 and the valuation of projected assets in Section 34.D.2, assets held in a separate account shall be summarized into asset categories determined by the actuarycompany as discussed in Section 34.A.2.

4. General Account Assets

a. General account assets shall be projected, net of projected defaults, using assumed investment returns consistent with their book value and expected to be realized in future periods as of the date of valuation. Initial assets that mature during the projection and positive cash flows projected for future periods shall be invested at interest rates, which, atin a manner that is representative of and consistent with the option of the actuary, are one ofcompany’s investment policy, subject to the following requirements:

i. The final maturities and cash flow structures of assets purchased in the model, such as the patterns of gross investment income and principal repayments or a fixed or floating rate interest basis, shall be determined by the company as part of the model representation;

ii. The combination of price and structure for fixed income investments and derivative instruments associated with fixed income investments shall appropriately reflect the projected U.S. Treasury curve along the relevant scenario and the requirements for gross asset spread assumptions stated below;

iii. For purchases of public non-callable corporate bonds, follow the requirements defined in VM-20 Sections 7.E, 7.F., and 9.F. The prescribed spreads reflect current market conditions as of the model start date and grade to long-term conditions based on historical data at the start of projection year four;

a. iv. The forward interest rates implied by the swap curve in effect as of the valuation date.

Guidance Note: The swap curve is based on the Federal Reserve H.15 interest swap rates. The rates are for a fixed rate payer in return for receiving three-month LIBOR. One place where these rates can be found is www.federalreserve.gov/releases/h15/default.htm.

b. The 200 interest rate scenarios available as prescribed for Phase I, C-3 RBC calculation, coupled with the separate account return scenarios by matching them up with the first 200 such scenarios and repeating this process until all separate account return scenarios have been matched with a Phase I scenario.

c. Interest rates developed for this purpose from a stochastic model that integrates the development of interest rates and the separate account returns.

When the option described in (a)—the forward interest rates implied by the swap curve—is used, an amount shall be subtracted from the interest rates to reflect the current market expectations about future interest rates using the process described in Section 3.E.1.

Commented [JW116]: [18. JR Comment: What is the difference between “assumed investment returns” and “investment return paths”?]


Attachment Four

86

21

The actuary may switch from (a) to (b), from (a) to (c) or from (b) to (c) from one valuation date to the next but may not switch in the other direction without approval from the domiciliary commissioner.

For transactions of derivative instruments associated with fixed income investments, reflect the prescribed assumptions in VM-20 Section 9.F. for interest rate swap spreads;

v. For purchases of other fixed income investments, if included in the model investment strategy, set assumed gross asset spreads over U.S. Treasuries in a manner that is consistent with, and results in reasonable relationships to, the prescribed spreads for public non-callable corporate bonds and interest rate swaps;

b. Notwithstanding the above requirements, the model investment strategy and any non-prescribed asset spreads shall be adjusted as necessary so that the aggregate reserve is not less than that which would be obtained by substituting an alternative investment strategy in which all fixed income reinvestment assets are public non-callable corporate bonds with gross asset spreads, asset default costs, and investment expenses by projection year that are consistent with a credit quality blend of 50% PBR credit rating 6 (A2/A) and 50% PBR credit rating 3 (Aa2/AA).

Drafting Note: this limitation is being referred to LATF for review.

Policy loans, equities and derivative instruments associated with the execution of a clearly defined hedging strategy (in compliance with VM-20 Section 7.L) are not affected by this requirement.

c. Any disinvestment shall be modeled in a manner that is consistent with the company’s investment policy and that reflects the company’s cost of borrowing where applicable, provided that the assumed cost of borrowing is not lower than the rate at which positive cash flows are reinvested in the same time period. Gross asset spreads used in computing market values of assets sold in the model shall be consistent with, but not necessarily the same as, the gross asset spreads in Section 4.D.4.a.iii and Section 4.D.4.a.v, recognizing that initial assets that mature during the projection may have different characteristics than modeled reinvestment assets.

5. Cash Flows from Invested Assets

a. Cash flows from general account fixed income assets, including starting and reinvestment assets, shall be reflected in the projection as follows:

i. Model gross investment income and principal repayments in accordance with the contractual provisions of each asset and in a manner consistent with each scenario.

ii. Reflect asset default costs as prescribed in VM-20 Section 9.F. and anticipated investment expenses through deductions to the gross investment income.

iii. Model the proceeds arising from modeled asset sales and determine the portion representing any realized capital gains and losses.

iv. Reflect any uncertainty in the timing and amounts of asset cash flows related to the paths of interest rates, equity returns, or other economic values directly in the projection of asset cash flows. Asset defaults are not subject to this requirement, since asset default assumptions must be determined by the prescribed method in VM-20 Sections 7.E, 7.F, and 9.F.

b. Cash flows from general account equity assets (i.e., non-fixed income assets having substantial volatility of returns such as common stocks and real estate), including starting and reinvestment assets, shall be reflected in the projection as follows:


Commented [HR119]: Based on VAWG referral for VM-20, need to add a constraint on the maturity or WAL of the alternative investment strategy that it is similar to the actual investment strategy.

Formatted: Font: Bold


Formatted: Indent: Left: 1.25"


Commented [HR120]: This is a VM-20 reference – need to add this so it does not look like a VM-21 reference. Or replace with reference to new VM-01 definition.


Commented [HR121]: This was part of the OW recommendation and makes sense to include.


Attachment Four

87

22

i. Determine the grouping for asset categories and the allocation of specific assets to each category in a manner that is consistent with that used for Sseparate Aaccount aAssets, as discussed in Section 4.A.2.

ii. Project the gross investment return including realized and unrealized capital gains in a manner that is consistent with the stochastically generated scenarios.

iii. Model the timing of an asset sale in a manner that is consistent with the investment policy of the company for that type of asset. Reflect expenses through a deduction to the gross investment return using prudent estimate assumptions.

E. Projection of Annuitization Benefits (Including GMIBs and GMWBs)

1. Assumed Annuitization Purchase Rates at Election

For purposes of projecting annuitization benefits (including annuitizations stemming from the election of a GMIB),) and withdrawal amounts from GMWBs, the projected annuitization purchase rates shall be determined assuming that market interest rates available at the time of election are the interest rates used to project general account assets, as determined in Section 3.D.4. However, where the interest rates used to project general account assets are based upon the forward interest rates implied by the swap curve in effect as of the valuation date (i.e., the option described in Section 3.D.4.a is used, herein referred to as a point estimate), the margin between the cost to purchase an annuity using the guaranteed purchase basis and the cost using the interest rates prevailing at the time of annuitization shall be adjusted as discussed below.4.D.4.

If a point estimate is being used, it is important that the margin assumed reflects the current market expectations about future interest rates at the time of annuitization, as described more fully below, and a downward adjustment to the interest rate assumed in the purchase rate basis. The latter adjustment is necessary since a greater proportion of contract holders will select an annuitization benefit when it is worth more than the cash surrender value than when it is not. As a practical matter, this effect can be approximated by using an interest rate assumption in the purchase rate basis that is 0.30% below that implied by the forward swap curve, as described below.

To calculate market expectations of future interest rates, the par or current coupon swap curve is used (documented daily in Federal Reserve H.15 with some interpolation needed). Deriving the expected rate curve from this swap curve at a future date involves the following steps:

a. Calculate the implied zero-coupon rates. This is a well-documented “bootstrap” process. For this process, we use the equation 100 = Cn * (v + v2 + … + vn) + 100vn where the “vt” terms are used to stand for the discount factors applicable to cash flows 1, 2, …, n years hence and Cn is the n-year swap rate. Each of these discount factors is based on the forward curve and, therefore, is based on a different rate (i.e., “v2” does not equal v times v). Given the one-year swap rate, one can solve for v. Given v and the two-year swap rate, one can then back into v2, and so on.

b. Convert the zero-coupon rates to one-year forward rates by calculating the discount factor needed to get from vt-1 to vt.

c. Develop the expected rate curve.

This recognizes that, for example, the five-year forward one-year rate is not the rate the market expects on one-year instruments five years from now. The reason is that as the bond gets shorter, the “risk premium” in the rate diminishes. This is sometimes characterized as “rolling down” the yield curve. Table A shows the historic average risk premium at various durations. From this table, one can see that to get the rate the market expects a one-year swap to have five years from now, one must subtract the risk premium associated with six-

Commented [LE123]: Not sure why this phrase is capitalized.

Attachment Four

88

23

year rates (0.95%) and add back that associated with one-year rates (0.50%). This results in a net reduction of 0.45%.

Table A: Risk Premium by Duration

Duration Risk Premium Duration Risk

Premium

1

2

3

4

0.500%

0.750%

0.750%

0.850%

6

7

8

9+

0.950%

1.000%

1.100%

1.150%

The Exhibit below combines the three steps. Column A through Column D convert the swap curve to the implied forward rate for each future payment date. Column E through Column H remove the current risk premium, add the risk premium t years in the future (the Exhibit shows the rate curve five years in the future), and uses that to get the discount factors to apply to the one-year, two-year, … five-year cash flows five years from now.

Exhibit: Derivation of Discount Rates Expected in the Future

Where interest rates are projected stochastically using an integrated model, although one would “expect” the interest rate n years hence to be that implied for an appropriate duration asset by the forward swap curve as described above, there is a steadily widening confidence interval about that point estimate with increasing time until the annuitization date. The “expected margin” in the purchase rate is less than that produced by the point estimate based on the expected rate, since a greater proportion of contract holders will have an annuitization benefit whose worth is in excess of cash surrender value when margins are low than when margins are high. As a practical matter, this effect can be approximated by using a purchase rate margin based on an earnings rate 0.30% below that implied by the

A B C D E F G H

1 2 3

Projection Years

Swap Curve Rate

PV of Zero Coupon

Forward 1 Year Rate

Risk Premium

Risk Premium 5 Years

Out

Expected Forward Rate in 5

Years

PV of Zero

Coupon in 5 Years

4 1 2.57% 0.97494 2.5700% 0.5000% 5 2 3.07% 0.94118 3.5879% 0.7500% 6 3 3.44% 0.90302 4.2251% 0.7500% 7 4 3.74% 0.86231 4.7208% 0.8500% 8 5 3.97% 0.82124 5.0010% 0.9000% 9 6 4.17% 0.77972 5.3249% 0.9500% 0.5000% 4.8749% 0.95352 10 7 4.34% 0.73868 5.5557% 1.0000% 0.7500% 5.3057% 0.90547 11 8 4.48% 0.69894 5.6860% 1.1000% 0.7500% 5.3360% 0.85961 12 9 4.60% 0.66050 5.8209% 1.1500% 0.8500% 5.5209% 0.81463 13 10 4.71% 0.62303 6.0131% 1.1500% 0.9000% 5.7631% 0.77024

14

Cell Formulas for Projection Year 10

=(1-B13 *SUM

($C$4:C12))/(1+B13)

=(C12/C13)-1 =E8 =D13-

E13+F13 =H12/(1+

G13)

Attachment Four

89

24

forward swap curve. If a stochastic model of interest rates is used instead of a point estimate, then no such adjustment is needed.

2. Projected Election of Guaranteed Minimum Income Benefit and Other Annuitization Options

a. For contracts projected to elect annuitization options (including annuitizations stemming from the election of a GMIB),) or for projections of GMWB benefits once the account value has been depleted, the projections may assume one of the following at the actuary’scompany’s option:

i. The contract is treated as if surrendered at an amount equal to the statutory reserve that would be required at such time for the payout annuity benefits. For GMWBs, the contract is treated as if surrendered at an amount equal to the statutory reserve that would be required at such time for the payout annuity benefits equivalent to the GMWB benefit payments.

ii. The contract is assumed to stay in force, and the projected periodic payments are paid, and the working reserve is equal to one of the following:

i. The statutory reserve required for the payout annuity, if it is a fixed payout annuity.

iii.ii. ii. If it is a variable payout annuity, the working reserve for a variable payout annuity.

a.b. If the projected payout annuity is a variable payout annuity containing a floor guarantee (such as a GPAF) under a specified contractual option, only option ii. under Section 4.E.2.a above shall be used.

c. Where mortality improvement is used to project future annuitization purchase rates, as discussed in (Section 4.E.1) above, mortality improvement also shall be reflected on a consistent basis in either the determination of the reserve in (Section 4.E.2.a.i) above or the projection of the periodic payments in Section 4.E.2.a.ii.

3. Projected Statutory Reserve for Payout Annuity Benefits

If the statutory reserve for payout annuity benefits referenced above in Section 4.E.2.a. requires a parameter that is not determined in a formulaic fashion such that, in reflecting the projected statutory reserve of a payout annuity benefit in the future, the company must make an assumption regarding this parameter and provide documentation in the VA ReportPBR Actuarial Report.

Guidance Note: F.

Relationship to RBC Requirements These requirements anticipate that the projections described herein may beare used for the determination of RBC for some or all of the contracts falling within the scope of these requirements. There are several differences between theseThese requirements and the RBC requirements, and among them for the topics covered within Section 4.A to 4.E are two major differences. First, the CTE level is different—CTE (70) for these requirements and CTE (90) for the RBC requirements. Second,identical. However, while the projections described in these requirements are performed on a basis that ignores federal income tax. That is, under these requirements, the accumulated deficiencies do not include, a company may elect to conduct the projections for calculating the RBC requirements by including projected federal income tax in the cash flows and reducing the

Commented [HR124]: Need to modify header to reflect GMWB

Commented [HR125]: Need this to remain as is to cover GMIB/other annuitization options.

Commented [HR126]: Adding this phrase makes this item now only about GMWBs. Need add unmodified sentence back to capture what this section originally covered.

Commented [HR127]: Extra period.


Commented [JW129]: ACLI suggested edit: “…provide documentation of the basis for this assumption in the…”

Commented [HR130]: Correct reference to VM-31 report. Also, could remove documentation reference here entirely, as this documentation requirement should be handled in VM-31.

Formatted: Font: Bold

Attachment Four

90

25

discount interest rates used to discount the scenario greatest present value (i.e., the interest rates determined in Section 3.D.4 contain no reduction for reflect the effect of federal income tax). Under the RBC requirements, the projections do include projected federal income tax, and the discount interest rates used as described in the RBC requirement do contain a reduction for federal income taxrequirements . A company that has elected to calculate RBC requirements in this manner may not switch back to using a calculation that ignores the effect of federal income tax without approval from the Ddomiciliary Ccommissioner.

1. To further aid the understanding of these requirements and any instructions relating to the RBC requirement, it is important to note the equivalence in meaning between the following terms, subject to the differences noted above:

a. The accumulated deficiency, the amount that is added to the starting asset amount in Section 2.D, is similar to the additional asset requirement referenced in the RBC requirement.

b. The CTE amount referenced in these requirements is similar to the total asset requirement referenced in the RBC requirement.

F. Compliance with ASOPs

When determining the CTE amountstochastic reserve using projections, the analysis shall conform to the ASOPs as promulgated from time to time by the ASB.

Under these requirements, thean actuary mustwill make various determinations, verifications and certifications. The company shall provide the actuary with the necessary information sufficient to permit the actuary to fulfill the responsibilities set forth in these requirements and responsibilities arising from each applicable ASOP, including ASOP No. 23, Data QualityASOPs.

H. Compliance with Principles

When determining the CTE amount using projections, any interpretation and application of the requirements of these requirements shall follow the principles discussed in Section 1.B.

Section 4: Reinsurance and Statutory Reporting Issues

.

Section 5: Reinsurance Ceded

A. Treatment of Reinsurance Ceded in the Aggregate Reserve

1. Aggregate Reserve Net ofPre- and Prior toPost- Reinsurance Ceded

As noted in Section 23.B, the minimum aggregate reserve is determined net of post-reinsurance ceded. Therefore, it is necessary to determine the components needed to determine the aggregate reserve (i.e., the additional standard scenarioprojection amount, and either the CTE amountstochastic reserve determined using projections and/or the CTE amountreserve determined using the Alternative Methodology) on a net of post-reinsurance ceded basis. In addition, as noted in Section 23.B, it may beis necessary to determine the aggregate reserve determined on a “direct”pre-reinsurance ceded basis, or prior to reflection of reinsurance ceded.. Where this is needed, each of these components shall be determined prior toignoring the effect of reinsurance ceded. Section 45.A.2 through Section 45.A.4 discuss methodsadjustments to inputs necessary to determine these components on both a “net of post-reinsurance” ceded and a “prior to pre-reinsurance” ceded basis. Note that due allowance for reasonable approximations may be used

Commented [HR131]: Delete extra space.

Commented [LE132]: Lower case “domiciliary commissioner” for consistency

Commented [HR133]: This is in the OW draft. Is this constraint being removed? It will also need to be in the RBC requirements, but it makes sense as part of this guidance note discussion.

Commented [HR134]: You discuss an/the actuary. Is this the qualified actuary? Or are we talking about actuaries that the qualified actuary may be relying on? In that case, should actuary be plural?

Commented [HR135]: Original was more appropriate here. “Must” is more accurate than “will”, because certifications are a requirement. “each applicable ASOP” is more correct than “applicable ASOPs”. And there is no revision in this restructure that makes ASOP 23 less important to call out than previously.

Commented [JW136]: [21. JR Comment: This Section 5 is needlessly long. The situation is that, if some or all of the contracts are reinsured, then the stochastic reserve, additional standard projection amount, and reserve under the Alternative Method must all be calculated on both pre-reinsurance and post-reinsurance bases, at least for the reinsured contracts. In calculating pre-reinsurance and post-reinsurance reserves, the method is the same; only the inputs are different. The only issue that needs to be addressed is where, in a pre-reinsurance situation, the starting asset amount may be in excess of the amount of assets actually held by the company.]

Commented [JW137]: [19. JR Comments: 1. The word “minimum” has not been used until now. I see no need to introduce it here. 2. If Section 3.B defines a post-reinsurance reserve, then that should be explicitly stated in Section 3.B.]

Commented [HR138]: Simplify alternative methodology references by making it determine a stochastic reserve and define the additional standard projection amount to be 0 for these policies.

Attachment Four

91

26

where appropriate.

2. CTE AmountStochastic Reserve Determined Using Projections

In order to determine the aggregate reserve net of post-reinsurance ceded, accumulated deficiencies, scenario greatest present valuesreserves and the resulting CTE amountstochastic reserve shall be determined reflecting the effects of reinsurance treaties that meet the statutory requirements that would allow the treaty to be accounted for as reinsurance within the projections.statutory accounting. This involves including, where appropriate, all anticipated reinsurance premiums or other costs and all reinsurance recoveries, where both premiums and recoveries are determined by recognizing any limitations in the reinsurance treaties, such as caps on recoveries or floors on premiums.

In order to determine the CTE amount prior to stochastic reserve pre-reinsurance ceded, accumulated deficiencies, scenario greatest present valuesreserves and the resulting CTE amountstochastic reserve shall be determined ignoring the effects of reinsurance ceded within the projections. One acceptable approach involves a projection based on the same starting asset amount as for the aggregate reserve net of post-reinsurance ceded and by ignoring, where appropriate, all anticipated reinsurance premiums or other costs and all reinsurance recoveries in the projections.

3. CTE Amount Reserve Determined using the Alternative Methodology (to be reviewed)

If a company chooses to use the Alternative Methodology, as allowed in Section 23.E, it is important to note that the methodology produces reserves on a prior to pre-reinsurance ceded basis. Therefore, where reinsurance is ceded, the Alternative Methodology must be modified to reflect the reinsurance costs and reinsurance recoveries under the reinsurance treaties in the determination of the aggregate reserve net of post-reinsurance ceded. In addition, the Alternative Methodology, unadjusted for reinsurance, shall be applied to the contracts falling under the scope of these requirements to determine the aggregate reserve prior to pre-reinsurance ceded.

4. Additional Standard ScenarioProjection Amount

Where reinsurance is ceded, the additional standard scenarioprojection amount shall be calculated as described in Section 56 to reflect the reinsurance costs and reinsurance recoveries under the reinsurance treaties. If it is necessary, the The additional standard scenarioprojection amount shall be also calculated prior to pre-reinsurance ceded using the methods described in Section 56, but ignoring the effects of the reinsurance ceded.

H.A. B. Aggregate Reserve to Be Held in the General Account

The amount of the reserve held in the general account shall not be less than the excess of the aggregate reserve over the sum of the basic reserve, as defined in Section 5.B, attributable to the variable portion of all such contracts.

C. Actuarial Certification and Memorandum

1. Actuarial Certification

Actuarial certification of the work done to determine the aggregate reserve shall be required. A qualified actuary (referred to throughout these requirements as “the actuary”) shall certify that the work performed has been done in a way that substantially complies with all applicable ASOPs. The scope of this certification does not include an opinion on the adequacy of the aggregate reserve, the company’s surplus or the company’s future financial condition. The actuary also shall note any

Commented [HR139]: This is either Reserve or Stochastic Reserve if we choose to more simply just define the additional standard projection amount as 0 for policies valued under the Alternative Methodology.

Commented [JW140]: [20. JR Comment: Is it the Alternative Method that must be modified, or the inputs into the Alternative Method? If it is the inputs, then it is false to say that the method produces pre-reinsurance reserves.]

Attachment Four

92

27

material change in the model or assumptions from that used previously and the estimated impact of such changes.

Section 10 contains more information on the contents of the required actuarial certification.

Guidance Note: The adequacy of total company reserves, which includes the aggregate reserve, is addressed in the company’s actuarial opinion as required by VM-30.

2. Required Memorandum

An actuarial memorandum shall be constructed documenting the methodology and assumptions upon which the aggregate reserve is determined. The memorandum also shall include sensitivity tests that the actuary feels appropriate, given the composition of the company’s block of business (i.e., identifying the key assumptions that, if changed, produce the largest changes in the aggregate reserve). This memorandum shall have the same confidential status as the actuarial memorandum supporting the actuarial opinion and shall be available to regulators upon request.

Section 10 contains more information on the contents of the required memorandum.

Guidance Note: This is consistent with Section 3A(4)(h) of Model #820, which states: “Except as provided in paragraphs (l), (m) and (n), documents, materials or other information in the possession or control of the Department of Insurance that are a memorandum in support of the opinion, and any other material provided by the company to the commissioner in connection with the memorandum, shall be confidential by law and privileged, shall not be subject to [insert open records, freedom of information, sunshine or other appropriate phrase], shall not be subject to subpoena, and shall not be subject to discovery or admissible in evidence in any private civil action. However, the commissioner is authorized to use the documents, materials or other information in the furtherance of any regulatory or legal action brought as a part of the commissioner’s official duties.”

3. CTE Amount Determined Using the Alternative Methodology

Where the Alternative Methodology is used, there is no need to discuss the underlying assumptions and model in the required memorandum. Certification that expense, revenue, fund mapping and product parameters have been properly reflected, however, shall be required.

Section 10 contains more information on the contents of the required actuarial certification and memorandum.

4. Material Changes

If there is a material change in results due to a change in assumptions from the previous year, the memorandum shall include a discussion of such change in assumptions and an estimate of the impact it has on the results.

Attachment Four

93


94

American Council of Life Insurers

101 Constitution Avenue, NW, Washington, DC 20001-2133

[email protected], 202-624-2169; JohnBruins,410-991-3996

www.acli.com

Brian Bayerle

Sr. Actuary

John Bruins

Consultant

January 22, 2019

Mr. Pete Weber

Chair – NAIC Variable Annuities Capital and Reserve (E/A) Subgroup

Re: Comments on Exposed Documents for Implementation of the VA Framework

Dear Mr. Weber:

The ACLI1 is pleased to submit the following comments to the Variable Annuities Capital and Reserve (E/A)

Subgroup (Joint WG) on behalf of our member companies regarding the exposed documents for the

implementation of the VA Framework.

ACLI generally supports all of the exposed documents as they do a good job overall of implementing the VA

framework developed by the VAIWG and adopted by the NAIC. The following are our substantive comments

on the documents, Appendix F contains a list of non-substantive or editorial changes.

VM-21 Sections 1-5

1. Section 2.B. Phase in

The ACLI supports the Phase-in Option 2, which is the simplified approach amortizing into income the

initial dollar difference in reserves as of January 1, 2020.

a. The purpose of a phase-in is to allow the company time to modify its product management before

fully reflecting the financial impacts of the new framework. Neither approach produces a reserve

that is “more correct” than the other. It is simply an organized transition between an existing

method and a new method.

b. The biggest problem that the Oliver Wyman study had identified with the old method is significant

non-economic volatility, which was identified to be the primary reason for the use of captives. The

new method has eliminated much of the non-economic volatility.

1 The American Council of Life Insurers (ACLI) advocates on behalf of 280 member companies dedicated to providing

products and services that promote consumers’ financial and retirement security. 90 million American families depend on

our members for life insurance, annuities, retirement plans, long-term care insurance, disability income insurance,

reinsurance, dental and vision and other supplemental benefits. ACLI represents member companies in state, federal and

international forums for public policy that supports the industry marketplace and the families that rely on life insurers’

products for peace of mind. ACLI members represent 95 percent of industry assets in the United States. Learn more at

www.acli.com.

Attachment Five

95

mailto:[email protected]


http://www.acli.com/

2

c. If the phase-in relies on calculations from both the old method and the new method at each

valuation date (Option 1), the non-economic volatility of the old method will continue to influence

the required reserve during the phase-in period. We don’t believe this would be a good outcome.

d. The recommendation Option 2 leverages the fact that the new method is the better method, then

provides a simpler computation and smoother transition to this better method.

e. Option 2 mitigates some of the duplicative effort that would exist under Option 1 for both

regulators (audit/examination of two complex calculations and their interaction to understand

reserves) and companies (administration and documentation of two modeling systems).

2. Section 3.A. through D: Alternative Methodology

As a result of the direction of the changes to the Alternative Methodology, Section 3 A, C, and D need

additional edits beyond those in the exposure, as outlined in Appendix A.

3. Section 4.A.7. – AVR and IMR

This section should state that the amounts used should be adjusted to a pre-tax basis. This is consistent

with VM-20, which uses a PIMR – PreTax IMR. The specific language is:

The AVR and IMR shall be handled consistently with the treatment in the company’s cash flow

testing, and the amounts should be adjusted to a pre-tax basis.

4. Section 4.B.3. & 4.

In Section 4.B.3. and 4.B.4., the description of the additional assets and the Net Asset Earned Rate

(NAER) is not clear. We have worked with John Robinson, MN, to develop language for those two

paragraphs, and the recommended language is included as Appendix B to this letter.

5. Section 4.D.1.

Add pre-tax AVR, consistent with #3 above.

“. . . plus the allocated amount of PIMR and AVR (pre-tax) attributable to the assets selected. “

VM-21 Section 6

6. Section 6.B.2.a. For the CSMP, there are requirements to define a particular market path for a company

to use in the Standard Projection. In the exposure, you had included a question of what to do if the

specified conditions were not met. We suggest that the process be to use the next closest scenario and

redo the steps. The specific language is included as Appendix C.

7. Section 6.B.5. Market Paths for CSMP

Numerous questions have been raised about the process to create the interest rate scenarios. Suggested

clarifying language is included in Appendix D. In addition, we ask that this be made an automatic

output of the prescribed generator to eliminate the potential misinterpretation of the intended scenarios.

8. Section 6.C.4. –

For a contract that has been taking routine conforming withdrawals, and makes a 1-year variation, that

single event should not override the prior election. The opening paragraph to 6.C.4. should be modified

to allow a single year variance from the required or elected pattern. Suggested language is:

4. Partial Withdrawals

Partial withdrawals required contractually or previously elected (e.g., a contract

operating under an automatic withdrawal provision, or that has voluntarily enrolled in

an automatic withdrawal program, on the valuation date) are to be deducted from the

Account Value in each projection interval consistent with the projection frequency

used, as described in Section 6.D., and according to the terms of the contract. However,

Attachment Five

96

3

if a GMWB or hybrid GMIB contract’s automatic withdrawals results in partial

withdrawal amounts in excess of the GMWB’s guaranteed maximum annual

withdrawal amount or the maximum amount above which withdrawals reduce the

GMIB basis by the same dollar amount as the withdrawal amount (the “dollar-for-

dollar maximum withdrawal amount”), such automatic withdrawals shall be revised

such that they equal the GMWB’s guaranteed maximum annual withdrawal amount or

the GMIB’s dollar-for-dollar maximum withdrawal amount.

Depending on the guaranteed benefit type, other partial withdrawals shall be projected

as follows but shall not exceed the free partial withdrawal amount above which

surrender charges are incurred:. Note that for contracts which have previously

commenced withdrawals but have a withdrawal in excess of the guaranteed

annual withdrawal amount in the policy year immediately preceding the valuation date,

companies may optionally treat such contracts as not having taken an excess

withdrawal when determining partial withdrawal assumptions.

9. Section 6.D. Projection Frequency

We recommend deleting this paragraph. There are several references in Sections 6.A. and 6.B. that the

standard projection uses the same modeling approach and parameters as for the stochastic reserve, but

substituting certain prescribed assumptions. (see 6.A.1.b., 6.B.1., 6.B.2.a.i., 6.B.2.b.). As such a

separate standard for projection frequency in not needed.

VM-21 Section 7

10. As part of the Framework change to the prescribed scenario generators, the pre-packaged scenarios that

had been a part of AG-43 were eliminated. This was not addressed in the drafting of Section 7,

therefore three modifications are needed to the exposed draft, as detailed in Appendix E.

VM-21 Section 8.E.

11. The reference in the last line to RBC LR027 Line 35 is incorrect, and should reference step 4 in the 7

step process defined in the new LR027 instructions. The wording would be:

At the option of the actuary, interest rates and total investment return scenarios for equity assets

and separate account fund performance may be generated in part or in full using non-prescribed

scenario generators in lieu of the prescribed generators, provided that the scenarios thus

generated do not result in a Total Asset Requirement (TAR) that is materially lower than the

TAR resulting from the use of the scenarios from the prescribed generators as defined in B, and

C. above. As defined in the RBC instructions, TAR is defined to be the sum of the reserve that

results from the application of these VM-21 requirements plus the RBC amount determined by

step 4 of the 7 step process in the Life RBC formula page LR027 Line (35) instructions for

2020.

We have no substantive comments at this time on VM-21 Section 9, the definitions in VM-01, or Actuarial

Guideline 43. We support these documents as drafted. We continue to work on possible recommendations to

modify the prescribed mortality, and certain other assumptions relating to salary deduction 403b business and

will communicate those when available.

Attachment Five

97

4

In the 6 appendices that follow are specific wording recommendations for items commented on above. We look

forward to continuing the work to implement the revised VA Framework for 2020.

Very truly yours,

Brian Bayerle John Bruins

Senior Actuary Consultant

202-624-2169 410-991-3996

[email protected] [email protected]

cc: Dave Fleming, NAIC

Attachment Five

98



Appendix A

5


The following are edits to the language included in the exposure of the Alternative Methodology,


The aggregate reserve for contracts falling within the scope of these requirements shall equal

the CTE amountstochastic reserve plusbut not less than the additional standard scenario

projection amount less any pre-tax Interest Maintenance Reserve (PIMR) and any Asset

Valuation Reserve for all contracts not valued under the Alternative Methodology

(Section 7), where the aggregate reserve is calculated as the standard scenario projection

amount plus the excess, if any, of the CTE amountstochastic reserve over the standard

scenarioprojection amount plus the reserve for any contracts determined using the

Alternative Methodology (following the requirements of Section 7)..

C. The Additional Standard Projection Amount

The additional standard projection amount is an additive factor, determined by applying one

of the two standard projection methods defined in Section 6. The same method must be

used for all contracts within a group of contracts that are aggregated together to

determine the reserve, and do not include any contracts whose reserve is determined

using the Alternative Methodology. The company shall elect which method they will

use to determine the additional standard projection amount. The company may not

change that election for a future valuation without the approval of the domiciliary

commissioner.

D. The Stochastic Reserve

The stochastic reserve shall be determined based on projections of the contracts falling

within the scope of these requirements excluding those contracts valued using the

Alternative Methodology, and the assets supporting these contracts, over a broad range

of stochastically generated projection scenarios described in Section 8 and using prudent

estimate assumptions as required by this VM-21.

The stochastic reserve may be determined in aggregate for all contracts falling within the

scope of these requirements (i.e., a single grouping) or, at the option of the company, it

may be determined by sub-groupings of contracts, in which case the stochastic reserve

shall equal the sum of the amounts computed for each such subgrouping.

The stochastic reserve for any group of contracts shall be determined as CTE70 of the scenario

reserves following the requirements of Section 4.

E. Alternative Methodology

For variable deferred annuity contracts that contain either no guaranteed benefits or only GMDBs

(i.e., no VAGLBs), the CTE amountreserve may be determined using the alternative methodology

Attachment Five

99

Appendix A

6

described in Section 76 rather than using the approach described in Sections 3.C and 23.D.

However, in the event the approach described in Sections 3.C and 23.D has been used in prior

valuations, the Alternative Methodology may not be used without approval from the domiciliary

commissioner.

The CTE amountreserve for the group of contracts to which the Alternative Methodology is applied

shall not be less than the aggregate cash surrender value of those contracts.

Attachment Five

100

Appendix B

7

Replace current paragraphs 4.A.3. & 4. with the following:

3. Determination of NAER on Additional Invested Asset Portfolio

a. The additional invested asset portfolio for a scenario is a portfolio of general account assets as

of the valuation date, outside of the starting asset portfolio, that is required in that projection

scenario so that the projection would not have a positive accumulated deficiency at the end of any

projection year. This portfolio may include only (i) general account assets available to the

company on the valuation date that do not constitute part of the starting asset portfolio; (ii) cash

assets, and (iii) reinvestment assets generated by investible cash from the initial additional

invested asset portfolio.

b. To determine the NAER on additional invested assets for a given scenario:

i. Project the additional invested asset portfolio as of the valuation date to the end of the

projection period,

a) investing any cash in the portfolio and reinvesting all investment proceeds

using the company’s investment policy;

b) without regard to any liability cash flows, and

c) incorporating the appropriate returns, defaults, and investment expenses for the

given scenario.

ii. If the value of the projected additional invested asset portfolio does not equal or

exceed the accumulated deficiencies at the end of each projection year for the scenario,

increase the size of the initial additional invested asset portfolio as of the valuation date,

and repeat the preceding step.

iii. Determine a vector of annual earned rates that replicates the growth in the additional

invested asset portfolio from the valuation date to the end of the projection period for the

scenario. This vector will be the NAER for the given scenario.

Guidance Note: Additional invested assets should be selected in a manner such that if the starting asset

portfolio were revised to include the additional invested assets, the projection would not be expected to

experience any positive accumulated deficiencies at the end of any projection year.

It is assumed that the accumulated deficiencies for this scenario projection are known.

Guidance Note: There are multiple ways to select the additional invested asset portfolio at the valuation

date. Similarly, there are multiple ways to determine the earned rate vector. The company should be

consistent in its choice of methods, from one valuation to the next.

Attachment Five

101

Appendix C

8

Section 6: Additional Standard Projection Requirements

B. Additional Standard Projection Amount - Application of the Standard Projection Method

2. Standard Projection Method

a. CSMP Method:

iii. Identify the path from the Company Standard Projection Set such that the

scenario reserve is closest to the CTE70 (adjusted), designated the Path A. This

scenario reserve shall be referred to as Company Amount A;

iv. Identify the following four market paths:

- two paths with the same starting interest rate as the Path A but equity shocks +/-

5% from that of Path A.

- two paths with the same equity fund returns as Path A but the next higher and

next lower interest rate shocks

From the four paths, identify the path whose reserve value is:

• If Company Amount A is lower than CTE70 (adjusted), the smallest

reserve value that is greater than CTE70 (adjusted);

• If Company Amount A is greater that CTE70 (adjusted), the greatest

reserve value that is less than CTE70 (adjusted).

This will be Path B, and the scenario reserve shall be referred to as Company

Amount B;

If none of the 4 paths satisfy the stated condition, discard the identified Path

A, and redo steps iii. and iv. using the scenario next closer to CTE70 to be

Path A in step iii.

Attachment Five

102

Appendix D

9

5. Market Paths for CSMP Method

If the company elects the CSMP method described in Section 6.B.2.a., the

Additional Standard Projection Amount shall be determined from the scenario

reserve calculated for the prescribed market paths defined below. Each

prescribed market path shall be defined by an initial equity fund stress and an

initial interest rate stress, after which both equity fund returns and interest rates

steadily recover and interest revert to the same long term mean. .

All combinations of prescribed equity fund return scenarios and interest rate

scenarios shall be considered prescribed Standard Projection market paths.

Accordingly, each company shall calculate scenario greatest present

valuesreserves for a minimum of 40 market paths.

a. Equity Fund Returns. Eight equity fund return market paths shall be

used. These scenarios differ only in the prescribed gross return in the

first projection year.

The eight prescribed gross returns for equity funds in the first projection

year shall be negative 25% to positive 10%, at 5% intervals. These gross

returns shall be projected to occur linearly over the full projection year.

After the first projection year, all prescribed equity fund return market

paths shall assume total gross returns of 3.0% per annum.

If the eight prescribed equity fund market paths are insufficient for a

company to calculate the Additional Standard Projection Amount via

steps (i) to (vi) outlined in Section 6.B.2.a., then the company shall

include additional equity fund market paths that increase or decrease the

prescribed gross returns in the first projection year by 5% increments at a

time.

b. Interest Rates. Five interest rate market paths shall be used. These market

paths differ only in the prescribed U.S. Treasury rates in the first

projection year.

The five prescribed interest rate market paths shall be generated using

the mean interest rate path embedded within the prescribed interest rate

scenario generator, using the prescribed parameters, described in Section

7.B. The mean interest rate path is the single market path generated if all

random variables in the prescribed interest rate scenario generator were

set to zero across all time periods.

The five prescribed interest rate market paths shall differ in the starting

U.S. Treasury rates used to generate the mean interest rate path.

Specifically, the following five sets of starting U.S. Treasury rates shall

be used:

(i) The actual U.S. Treasury rates as of the valuation date;

Attachment Five

103

Appendix D

10

(ii) The actual U.S. Treasury rates as of the valuation date, reduced at

each point on the term structure by 25% of the difference between the

U.S. Treasury rate as of the valuation date and 0.01%;

(iii) The actual U.S. Treasury rates as of the valuation date, reduced at



(iv) The actual U.S. Treasury rates as of the valuation date, reduced at



(v) The actual U.S. Treasury rates as of the valuation date, increased at


U.S. Treasury rate as of the valuation date and 0.01%.

For each of these five sets of starting U.S. Treasury rates, the prescribed

interest rate market path is defined as the interest rate path generated by

the prescribed interest rate scenario generator (described in Section 7.B)

when the applicable set of starting rates is the initial yield curve for the

generator and all random variables in the generator are set to zero across

all time periods. The starting U.S. Treasury rates should not change any

prescribed parameters in the generator, including the mean reversion

parameter.

If the five prescribed interest rate market paths are insufficient for a

company to calculate the Additional Standard Projection Amount via

steps (i) to (vi) outlined in Section 6.B.2.a., then the company shall

include additional interest rate market paths that increase or decrease the

prescribed starting U.S. Treasury rates at each point on the term structure

by increments equal to 25% of the difference between the U.S. Treasury

rate as of the valuation date and 0.01%. The lowest interest rate to be

used in this analysis is 0.01%.

For projecting swap rates along the prescribed interest rate market paths,

companies shall assume that the swap-to-Treasury spread term structure

in effect as of the valuation date persists throughout each market path.

Floor rates at 0.01%

Attachment Five

104

Appendix E

11

Section 7: Alternative Methodology

A. General Methodology

1. General Methodology Description

[Material not shown to conserve space]

1. Definitions of Terms Used in This Section

[paragraphs a – m not shown to conserve space]

n. Prepackaged Scenarios: Prepackaged scenarios are the year-by-year asset

returns that may be used (but are not mandated) in projections related to

the alternative methodology. These scenarios are available on the

Academy website.

no. Quota-Share Reinsurance: In this type of reinsurance treaty, the same

proportion is ceded on all cessions. The reinsurer assumes a set percentage

of risk for the same percentage of the premium, minus an allowance for

the ceding company’s expenses.

op. Resets: A reset benefit results in a future minimum guaranteed benefit

being set equal to the contract’s account value at previous set date(s) after

contract inception.

pq. Risk Mitigation Strategy: A risk mitigation strategy is a device to reduce

the probability and/or impact of a risk below an acceptable threshold.

qr. Risk Profile: Risk profile in these requirements relates to the prescribed

asset class categorized by the volatility of returns associated with that

class.

rs. Risk Transfer Arrangements: A risk transfer arrangement shifts risk

exposures (e.g., the responsibility to pay at least a portion of future

contingent claims) away from the original insurer.

st. Roll-Up: A roll-up benefit results in the guaranteed value associated with

a minimum contractual guarantee increasing at a contractually defined

interest rate.

tu. Volatility: Volatility refers to the annualized standard deviation of asset

returns.

C. Calculation of the GC Component

8. Adjusting F and G for Product Design Variations

This subsection describes the typical process for adjusting F and G factors due to

a variation in product design. Note that R (as determined by the slope and intercept

Attachment Five

105

Appendix E

12

terms in the factor table) would not be adjusted.

a. Select a contract design among those described in Section 6.C.4 that is

similar to the product being valued. Execute cash-flow projections using the

documented assumptions (see table of Liability Modeling Assumptions &

Product Characteristics in Section 6.E.1 and table of Asset-Based Fund

Charges in Section 6.E.2) and the prepackaged scenarios from the

prescribed generator for a set of representative cells (combinations of

attained age, contract duration, asset class, AV/GMDB ratio and asset-based

charges). These cells should correspond to nodes in the table of

precalculated factors. Rank (order) the sample distribution of results for the

present value of net cost. Determine those scenarios that comprise CTE (65).

d. Execute “product specific” cash-flow projections using the

documented assumptions and prepackaged scenarios from the

prescribed generators for the same set of representative cells. Here, the

company should model the actual product design. Rank (order) the

sample distribution of results for the present value of net cost.

Determine those scenarios that comprise CTE (65).

Attachment Five

106

Appendix F

13

Suggested Edits to various documents:

Section

Citation

Edits

3.A.

3.D.

The following are further edits to the language included in Appendix A.


The aggregate reserve for contracts falling within the scope of these

requirements shall equal the stochastic reserve plus the

additional standard projection amount for all groupings or

subgroupings of contracts not valued under the Alternative

Methodology (Section 7), plus the reserve for any contracts

determined using the Alternative Methodology (following the

requirements of Section 7) less the positive value of any Interest

Maintenance Reserve (PIMR) and pre-tax AVR.

D. The Stochastic Reserve

The stochastic reserve shall be determined based on projections of

the contracts falling within the scope of these requirements

excluding those contracts valued using the Alternative

Methodology, and the assets supporting these contracts, over a

broad range of stochastically generated projection scenarios

described in Section 8 and using prudent estimate assumptions

as required by this VM-21.

The stochastic reserve may be determined in aggregate for all

contracts falling within the scope of these requirements (i.e., a

single grouping) or, at the option of the company, it may be

determined by sub-groupings of contracts, in which case the

stochastic reserve shall equal the sum of the amounts

computed for each such subgrouping.

The stochastic reserve for any group of contracts shall be

determined as CTE70 of the scenario reserves following the

requirements of Section 4.

4.B.5. 5. Direct Iteration

In lieu of the method described in Sections 4.B.2, Section 4.B.3,

and Section 4.B.4 above, for each scenario, the company may

solve for the amount of starting assets which, when projected

along with all contract cash flows for the given scenario, result

in the defeasement of all projected future benefits and expenses

Attachment Five

107

Appendix F

14

at the end of the projection horizon with no deficiencies at the

end of any projection year during the projection period.

4.D.1. 1. Starting Asset Amount

a. For the projections of accumulated deficiencies, the value of

assets at the start of the projection shall be set equal to the

approximate value of statutory reserves at the start of the

projection plus the allocated amount of PIMR attributable to

the assets selected. Assets shall be valued consistently with

their annual statement values, and shall include any hedges

held in support of the guarantees in the contracts being

valued1. The amount of such asset values shall equal the sum

of the following items, all as of the start of the projection:

i. All of the separate account assets supporting the

contracts;

i.ii. Any hedge assets held in support of the contracts being

valued2; and

ii.iii. c. An amount of assets held in the general account

equal to the approximate value of statutory reserves as

of the start of the projections less the amounts in (ia)

and(ii).

b. If the amount of initial general account assets is negative, the

model should reflect a projected interest expense. General

account assets chosen for use as described above shall be

selected on a consistent basis from one reserve valuation

hereunder to the next.

c. To the extent the sum of the value of such hedge assets , or

cash or other general account assets in an amount equal to the

annual statement value of such hedge assets, and the value of

separate account assets supporting the contracts is greater than

the approximate value of statutory reserves as of the start of

the projections, then the company shall include enough

negative general account assets or cash such that the starting

asset amount equals the approximate value of statutory

reserves as of the start of the projections.

d. For an asset portfolio that supports some policies and

contracts that a) are subject to these requirements and b)

others that are not subject to these requirements, the company

2 Deferred hedge gains/losses developed under SSAP108 are not included in the value of the starting assets.

Attachment Five

108

Appendix F

15

shall determine an equitable method to apportion the total

amount of starting assets between a. and b. the two groups.

4.E.3. 3. Projected Statutory Reserve for Payout Annuity Benefits

If the statutory reserve for payout annuity benefits referenced

above in Section 4.E.2.a. requires a parameter that is not

determined in a formulaic fashion such that, in reflecting the

projected statutory reserve of a payout annuity benefit in the

future, the company must make an assumption regarding this

parameter and provide documentation of the basis for this

assumption in the VA Report.

6.A.2. 2. Inforce Used in the Standard Projection

If the Stochastic reserve is determined by the use of a model

office, which is a grouping of contracts into representative cells,

the model office shall be replaced with a seriatim in force prior

to the projection needed to calculate the additional standard

projection amount if the CSMP method described in Section

6.B.2.a. is used. If the company elects to calculate the

additional standard projection amount using the CTEPA method

described in Section 6.B.2.b., it may continue to use the same

model office grouping of contracts, or one that is no less

granular than the grouping that was used to determine the

stochastic reserve, provided that, on an annual basis, it can be

demonstrated that using such a grouped inforce does not

materially reduce the additional standard projection amount that

would result from using a seriatim inforce.

6.B.2.a.vi.

6.B.2.b.

6.B.2.c.

Change “Weighted Prescribed Projection Amount” to

“Prescribed Projections Amount” since the CTEPA

calculation is not a weighted average.

6.B.4. 4. Modeled Hedges

Cash flows associated with hedging shall be projected in the

same manner as that used in the calculation of the CTE70

(adjusted) as discussed in Section 9.C (or Section 4.A.4.a for a

company without a CDHS).

6.C.1.

6.C.4.

C. Assumptions

1. Assignment of Guaranteed Benefit Type

Attachment Five

109

Appendix F

16

Assumptions shall be set for each contract in accordance with

the contract’s guaranteed benefit type as defined in VM-01. If a

contract cannot be classified into any categories within a given

assumption, There may be benefits not described above. the

company shall determine the defined benefit type with the most

similar benefits and risk profile as the company’s benefit. The

choice of benefit type proxy shall be documented in the VA

Report.

4. Partial Withdrawals

j. There may be benefits not described above. The company

shall determine the defined benefit type with the most similar

benefits and risk profile as the company’s benefit. The choice

of benefit type proxy shall be documented in the VA Report.

6.C.2. 2. Maintenance Expenses

Maintenance expense assumptions shall be determined as the

sum of a. plus b. if the company is responsible for the

administration of the contract or c. if the company is not

responsible for the administration of the contract:

6.C.3. 3. Guarantee Actuarial Present Value

The Guarantee Actuarial Present Value (“GAPV”) represents

the actuarial present value of the lump sum or income

payments associated with a guaranteed benefit rider. For the

purpose of calculating the GAPV, such payments shall include

the portion that is paid out of the contract holder’s Account

Value.

6.C.5.h. h) Scale the remainder of the adjusted and scaled GAPV2 values

at all future initial withdrawal ages (i.e. all ages greater than the

initial withdrawal age that occurs immediately after the

termination of the guaranteed growth or the one-time bonus

with the greatest GAPV, as identified in the preceding step)

such that the sum of the revised GAPV2 values equals 0.95 for

tax-qualified GMWB policies, 0.80 for non-qualified GMWB

policies, 0.85 for tax-qualified hybrid GMIB policies, and 0.60

for non-qualified hybrid GMIB policies.

6.C.5.j. j) Scale the remained of the revised GAPV2 values at all future

initial withdrawal ages (i.e. all ages greater than 71, as

identified in the preceding step) such that the sum of the revised

GAPV2 values equals 0.95 for tax-qualified GMWB policies

and 0.85 for tax-qualified hybrid GMIB policies again.

Attachment Five

110

Appendix F

17

9.A. A. Initial Considerations

1. Subject to the guidance in Section 9.C.2., tThe appropriate costs

and benefits of hedging instruments that are currently held by the

company in support of the contracts falling under the scope of

these requirements (excluding those that involve the offsetting of

the risks associated with variable annuity guarantees with other

products outside of the scope of these requirements, such as

equity-indexed annuities) shall be included in the calculation of

the Stochastic reserve, determined in accordance with Section

3.D and Section 4.D. At the option of the actuary the full portfolio

of these hedge instruments may be substituted by cash or other

general account assets in an amount equal to their aggregate

market value in the starting assets; however, the actuary may not

conduct such substitution for individual hedge instruments.

9.C.3. 3. Because most models will include at least some approximations

or idealistic assumptions, CTE70 amount (best efforts) may

overstate the impact of the hedging strategy. To compensate for

potential overstatement of the impact of the hedging strategy, the

value for the Stochastic reserve is given by:

9.D. D. Additional Considerations for CTE70 (best efforts)

If the company is following a CDHS, fair value of the portfolio

of contracts falling within the scope of these requirements shall

be computed, and compared to the CTE (best efforts) and to the

CTE (adjusted) values. If the CTE (best efforts) is below both

the fair value and the CTE (adjusted) value, the company should

be prepared to explain why that result is reasonable.

For the purposes of this analysis, the stochastic reserves and fair

value calculations shall be done without requiring the scenario

reserve for any given scenario to be equal to or in excess of the

cash surrender value in aggregate in scope of the valuation date

for the group of contracts modeled in the projection.

Attachment Five

111


112

Moody’s Analytics

Richard Yule, FSA, MAAA (via 12/19 and 1/22 emails)

1) The meeting notes mention the situation when all CSPS scenario reserves are on the same side of CTE70. Similarly,what should be done if Path A is at the highest or lowest interest or equity rate? The new guidance says to look atscenarios with the next highest and lowest rates compared to Path A.

2) If all four potential Path B paths are on the same side of CTE70(adjusted) as Path A, this creates a similar situation.The guidance says to pick among the four potential Path B scenarios the one on the opposite side of CTE70(adjusted)from Path A.

3) Lapses are different in the policy year following the surrender charge period. There are many surrender chargestructures out there, it is very common for products to have each premium deposit carry its own surrender chargeperiod. This causes several situations that may not have been accounted for when basing lapses on the policy year inwhich surrender charges end. If a premium comes in halfway through a policy year, then the "surrender charge period"will also end halfway through the policy year. As written, the lapse rate would not go up for the rest of that policy yeareven though there are no surrender charges in effect, and then would go up the following policy year instead. Shouldwe be saying "year" instead of "policy year"?

4) The first paragraph of Section 6.C.2 says "or c. f the company", but should say "or c. if the company"

5) GAPVs are prescribed to include amounts paid from fund value. For GMABs, it seems ambiguous how the GAPVshould be calculated, especially when there are multiple reset periods in which the GMAB applies, for example if itguarantees a fund value every 10 years with an x% rollup rate applied so that the guarantee amount is larger the furtherinto the future you get. I’ve seen several different interpretations of what a GAPV for such a product could be. Theapproach that seems the most likely to be meaningful would be to determine a GAPV for each date in isolation, anddetermine the GAPV as the greatest of these GAPVs. I’d encourage some clarification of how the committee intendsGMAB GAPVs to be calculated.

6) Section 6.C.3.g says that GMDB GAPVs are calculated by assuming that they do not terminate. If a GMDB has aguaranteed accumulation rate of X%, would companies have to assume that the accumulation rate also does notterminate? This seems likely to give very high GAPVs for products with high guaranteed rates.

7) Similar to my comment 3 given previously, Sections 6.C.4.f and 6.C.4.h prescribe different behavior if withdrawalshappened in the policy year prior to that containing the valuation date. What if there were withdrawals earlier in thepolicy year that does contain the valuation date? If the valuation date is 6 months into my policy year, it seems unusualto assume that if a policyholder took a withdrawal 8 months ago they like to take withdrawals, but if they took it 2months ago then they are not. Should these say in the year prior to valuation, or perhaps “in the policy year prior to, orearlier within, the policy year containing the valuation date”?

8) Section 6.C.6.f gives a higher lapse rate in the year following maturity of a GMAB. For GMABs with multiple guaranteeperiods (as described above in comment 5), would the “maturity” be the last date on which the guarantee might apply?If it is every 10 years for the life of the policy, would the last date trigger the higher lapse, or would none of them? Thisalso ties into the definition of GMAB in VM-01, where a GMAB is defined as a benefit that guarantees “an amountpayable on the contractually determined maturity date of the benefit will be increased and/or will be at least aminimum amount”. Is each date a GMAB guarantees an account value as of considered a separate GMAB with its ownmaturity date?

Attachment Six

113

Matt Kauffman Comments/Questions on exposure draft of proposed VM-21 changes (Jan. 2019)

1. Starting Asset Amount

From section 4.D.1:

For the projections of accumulated deficiencies, the value of assets at the start of the projection shall be set equal to the approximate value of statutory reserves at the start of the projection plus the allocated amount of PIMR attributable to the assets selected.

From section 4.B.3:

On the valuation date, the company shall determine the additional invested assets as the amount of assets needed, or an approximation of it, to fund the present value of the accumulated deficiency. These assets may include only (i) general account assets available to the company on the valuation date that do not constitute part of the starting asset amount, and (ii) cash, and shall exclude separate account assets and policy loans. If the company elects not to include certain general account assets, or if the amount of additional available general account assets is lower than the amount needed to fund the present value of the accumulated deficiency, the company shall model cash assets to fill any deficiencies in the amount of additional invested assets. Any cash assumed will then be subject to the company’s investment policy as described in Section 4.D.4.a.

Comment/Question:

It is unclear to us how the starting assets should be adjusted with additional invested assets when using the scenario reserve method defined in section 4.B.2-4.B.4:

1. Should the starting assets be adjusted to the same amount across all stochastic scenarios (i.e. similar to VM-20)? If yes, does this mean it should be adjusted to the amount that will fund the present value of accumulated deficiency in the worst scenario (such that all other scenarios have accumulated surplus)? Or should it be adjusted to the amount that will fund the present value of accumulated deficiency on average across the worst 30% of scenarios (similar to the 2% asset collar requirement in VM-20)?

2. Should the starting assets be adjusted per individual stochastic scenario? If yes, then please clarify in what way this differs from the “direction iteration” method described in section 4.B.5.

3. Should the starting assets be adjusted only for some scenarios and not others (i.e. only the scenarios that have accumulated deficiency when using the initial approximated starting asset amount)?

Attachment Six

114

2. Net Earned Rate on Additional Invested Assets

From section 4.B.2:

In determining the scenario reserve, accumulated deficiencies shall be discounted at the net asset earned rate on additional assets, as defined in Section 4.B.4.

From section 4.B.4:

The net asset earned rate on additional assets shall represent the ratio of net investment earnings on additional invested assets to the amount of additional invested assets, as defined below. All items reflected in the ratio are consistent with statutory asset valuation and accrual accounting, including reflection of due, accrued, or unearned investment income where appropriate. A vector of NAER’s will be produced for each scenario.

Comment/Question:

The definition of “additional invested assets” appears to be heavily dependent upon the initial approximation of the “starting asset amount” as defined in section 4.D.1.

For example, let's assume you initially approximate that your Statutory reserve will be around $100 million, so you first run your model with starting assets totaling that amount. After running the model, the scenario reserve is determined to be $110 million, meaning your initial estimate of starting assets was too low. So you increase your starting assets to $110 million by adding $10 million of cash (or other general account assets) and rerun your model for a 2nd iteration. This would imply the "additional invested assets" in this example is $10 million of cash. The “additional invested assets”, including any future reinvestments purchased with the $10 million of cash, would need to be tracked separately from any other general account assets included in the “starting asset amount”, for the purposes of the net asset earned rate (NAER) calculation.

But what if your initial approximation had instead been exactly $110 million? Then your “additional invested assets” would be $0, and then how would you define a NAER on a portfolio of $0? The breakdown of the general account (GA) asset portfolio between the “starting asset amount” vs. “additional invested assets” appears arbitrarily dependent on the initial estimate of “starting asset amount”. For this reason, it seems troublesome to define the NAER based on only the “additional invested assets” portion of the GA asset portfolio rather than on the entire portfolio. Does the NAIC have any guidance for this issue?

Attachment Six

115

3. Bond fund and money market fund returns in the 40 market paths used for the CSMP method of calculating the Standard Scenario Amount

From section 6.B.5.c:

Bond fund returns Equal to the 5-year trailing average of the 5-year U.S. Treasury rate, plus an earned spread of 100 bps per annum. In the first projection year, additionally adjust the projected return by an amount equal to 20% of the prescribed gross equity fund return – with the same directionality, reflected in a linear fashion over the full projection year

Money market fund returns

Follow the three-month U.S. Treasury rate projected in the prescribed scenario

Comment/Question:

These functions appear to be a simplification of the formulas already provided by the prescribed Academy scenario generator for bond fund and money market fund returns.

For example, the Academy generator defines two bond funds.

• U.S. Intermediate Government, with a reference maturity of 7 years • U.S. Corporate Long, with a reference maturity of 10 years

Both bond funds have unique prescribed parameters. Similarly, the money market fund also has its own parameters. These funds also have a well-defined formula for calculating their returns based on the generated U.S. Treasury rates.

I would interpret the proposed VM-21 changes to override these Academy formulas for the standard scenario calculations, such that both bond fund returns would be the same for these paths (both based on the 5 year trailing average of 5-year U.S. Treasury rates).

Since these 40 paths already require the use of the Academy generator for producing the interest rates, why does the proposed regulation override the existing Academy generator formulas for bond and money market funds? If the intention is to introduce the additional adjustment of 20% of the equity fund returns in the first projection year (assuming some correlation between interest and equity rates), couldn’t that be done by layering this adjustment on top of the Academy’s bond fund return formula (with zero bond fund volatility)?

Attachment Six

116

VM-21: Requirements for Principle-Based Reserves for Variable Annuities



The aggregate reserve for contracts falling within the scope of these requirements shall equal the stochastic reserve (following the requirements of Section 4) plus the additional standard projection amount (following the requirements of Section 6) for all contracts not determined using the Alternative Methodology plus the reserve for any contracts determined using the Alternative Methodology(following the requirements of Section 7).

E. Alternative Methodology

For variable deferred annuity contracts that contain either no guaranteed benefits or only GMDBs (i.e.,no VAGLBs), the reserve may be determined using the alternative methodology described in Section 7 rather than using the approach described in Section 3.C and Section 3.D. However, in the event theapproach described in Section 3.C and Section 3.D has been used in prior valuations, the AlternativeMethodology may not be used without approval from the domiciliary commissioner.

The reserve for the group of contracts to which the Alternative Methodology is applied shall not be less than the aggregate cash surrender value of those contracts.

Section 7: Alternative Methodology

A. General Methodology

1. General Methodology Description

For variable deferred annuity contracts that either contain no guaranteed benefits or onlyGMDBs, including “earnings enhanced death benefits,” (i.e., no VAGLBs), the reserve may bedetermined by using the method outlined below rather than by using the approach described inSections 3.C and 3.D (i.e., based on projections), provided the approach described in Section3.D has not been used in prior valuations or else approval has been obtained from the domiciliary commissioner.

The reserve determined using the Alternative Methodology for a group of contracts with GMDBs shall be determined as the sum of amounts obtained by applying factors to each contract in force as of a valuation date and adding this to the contract’s cash surrender value. The amount that is added to an individual contract’s cash surrender value may be negative, zero or positive, thus resulting in a reserve for a given contract that could be less than, equal to or greater than the cash surrender value. The resulting reserve in aggregate shall not be less than the greater of the cash surrender value or the reserve determined by applying VM-C-33, each in aggregate for the group of contracts to which the Alternative Methodology is applied.

The reserve determined using the Alternative Methodology for a group of contracts that contain no guaranteed benefits shall be determined using an application of VM-C-33, as described below.

Guidance Note: The term “contracts that contain no guaranteed benefits” means that there are no guaranteed benefits at any time during the life of the contract (past, present or future).

For purposes of performing the Alternative Methodology, materially similar contracts within the group may be combined together into subgroups to facilitate application of the factors.

Formatted

Deleted: 2

Deleted: the CTE amountstochastic reserve but not less than the standard scenario projection amount, where the aggregate reserve is calculated as

Commented [HR1]: The second part of the description contradicts the first part, by adding the piece regarding the Alternative Methodology. Is there any reason to state this twice?

Deleted: scenario

Deleted:

Commented [HR3]: Otherwise, the first two items are not limited to those contracts not using the Alternative Methodology.

Deleted: plus the excess, if any, of the CTE amountstochastic reserve over the standard scenarioprojection amount

Commented [HR4]: Current phrasing was not accurate.

Commented [HR5]: Now that I am thinking of it, it seems simpler to let the Alternative Methodology calculate the stochastic reserve for those contracts, and define the additional standard projection amount to be 0 for those contracts.

Deleted: CTE amount

Deleted: 6


Deleted: s

Deleted: 2

Commented [LE7]: Put “Section” before each reference.

Deleted: s

Deleted: 2

Commented [LE8]: Same comment as above

Deleted: CTE amount

Commented [HR9]: Not referencing CTE amount any more.

Deleted: The actuary shall document the assumptions and procedures used for the Alternative Methodology and summarize the results obtained as described in Section 4 and Section 10.¶

Deleted: Section 6

Deleted: CTE amount

Deleted: 2

Commented [HR10]: As in Section 3, if you are having the Alternative Methodology replace the entirety of the reserve.

Deleted: 2

Deleted: CTE amount

Deleted: The resulting CTE amount shall not be less than the cash surrender value in aggregate for the group of contracts to which the Alternative Methodology is applied.¶Guidance Note:

Deleted: a

Deleted: CTE amount

Deleted: AG

Attachment Seven

117

Specifically, all contracts comprising a “subgroup” must display substantially similar characteristics for those attributes expected to affect reserves (e.g., definition of guaranteed benefits, attained age, contract duration, years-to-maturity, market-to-guaranteed value, asset mix, etc.). Grouping shall be the responsibility of the actuary but may not be done in a manner that intentionally understates the resulting reserve.

2. Definitions of Terms Used in This Section

a. Annualized Account Charge Differential: This term is the charge as percentage account value (revenue for the company) minus the expense as percentage of account value.

b. Asset Exposure: Asset exposure refers to the greatest possible loss to the insurance company from the value of assets underlying general or separate account contracts falling to zero.

c. Benchmark: Benchmarks have similar risk characteristics to the entity (e.g., asset class, index or fund) to be modeled.

d. Deterministic Calculations: In a deterministic calculation, a given event (e.g., asset returns going up by 7% and then down by 5%) is assumed to occur with certainty. In a stochastic calculation, events are assigned probabilities.

e. Foreign Securities: These are securities issued by entities outside the U.S.

f. Grouped Fund Holdings: Grouped fund holdings relate to guarantees that apply across multiple deposits or for an entire contract instead of on a deposit-by-deposit basis.

g. Guaranteed Value: The guaranteed value is the benefit base or a substitute for the account value (if greater than the account value) in the calculation of living benefits or death benefits. The methodology for setting the guaranteed value is defined in the variable annuity contract.

h. High-Yield Bonds: High-yield bonds are below investment grade, with NAIC ratings (if assigned) of 3, 4, 5 or 6. Compared to investment grade bonds, these bonds have higher risk of loss due to credit events. Funds predominately containing securities that are not NAIC rated as 1 or 2 (or similar agency ratings) are considered to be high-yield.

i. Investment Grade Fixed Income Securities: Securities with NAIC ratings of 1 or 2 are investment grade. Funds containing securities predominately with NAIC ratings of 1 or 2 or with similar agency ratings are considered to be investment grade.

j. Liquid Securities: These securities can be sold and converted into cash at a price close to its true value in a short period of time.

k. Margin Offset: Margin offset is the portion of charges plus any revenue-sharing allowed under Section 4.A.5 available to fund claims and amortization of the unamortized surrender charges allowance.

l. Multi-Point Linear Interpolation: This methodology is documented in mathematical literature and calculates factors based on multiple attributes categorized with discrete values where the attributes’ actual values may be between the discrete values.

m. Model Office: A model office converts many contracts with similar features into one contract with specific features for modeling purposes.

Deleted: 3

Attachment Seven

118

n. Prepackaged Scenarios: Prepackaged scenarios are the year-by-year asset returns that may be used (but are not mandated) in projections related to the alternative methodology. These scenarios are available on the Academy website.

o. Quota-Share Reinsurance: In this type of reinsurance treaty, the same proportion is ceded on all cessions. The reinsurer assumes a set percentage of risk for the same percentage of the premium, minus an allowance for the ceding company’s expenses.

p. Resets: A reset benefit results in a future minimum guaranteed benefit being set equal to the contract’s account value at previous set date(s) after contract inception.

q. Risk Mitigation Strategy: A risk mitigation strategy is a device to reduce the probability and/or impact of a risk below an acceptable threshold.

r. Risk Profile: Risk profile in these requirements relates to the prescribed asset class categorized by the volatility of returns associated with that class.

s. Risk Transfer Arrangements: A risk transfer arrangement shifts risk exposures (e.g., the responsibility to pay at least a portion of future contingent claims) away from the original insurer.

t. Roll-Up: A roll-up benefit results in the guaranteed value associated with a minimum contractual guarantee increasing at a contractually defined interest rate.

u. Volatility: Volatility refers to the annualized standard deviation of asset returns.

3. Contract-by-Contract Application for Contracts That Contain No Guaranteed Living or Death Benefits

The Alternative Methodology reserve for each contract that contains no guaranteed living or death benefits shall be determined by applying VM-C-33. The application shall assume a return on separate account assets equal to the valuation interest rate for a non-variable annuity with similar features issued during the first calendar quarter of the same calendar year less appropriate asset based charges. It also shall assume a return for any fixed separate account and general account options equal to the rates guaranteed under the contract.

The reserve for such contracts shall be included in the aggregate minimum comparison defined in Section 7.A.1.b.

4. Contract-by-Contract Application for Contracts That Contain GMDBs Only

For each contract, factors are used to determine a dollar amount, equal to R x (CA + FE) + GC (as described below), that is to be added to that contract’s cash surrender value as of the valuation date. The dollar amount to be added for any given contract may be negative, zero or positive. The factors that are applied to each contract shall reflect the following attributes as of the valuation date. a. The contractual features of the variable annuity product.

b. The actual issue age, period since issue, attained age, years-to-maturity and gender applicable to the contract.

c. The account value and composition by type of underlying variable or fixed fund.

d. Any surrender charges.

Deleted: AG

Deleted: year of issue

Deleted: no less than

Deleted: cash surrender value on the valuation date, as

Deleted: E.2

Commented [HR11]: Incorrect section reference. Just 7.A.1.

Attachment Seven

119

e. The GMDB and the type of adjustment made to the GMDB for partial withdrawals (e.g., proportional or dollar-for-dollar adjustment).

f. Expenses to be incurred and revenues to be received by the company as estimated on a prudent estimate basis as described in Section 4.A.1. and complying with the requirements for revenue sharing as described in Section 4.A.5.

5. Factor Components

Factors shall be applied to determine each of the following components.

Guidance Note: Material to assist in the calculation of the components is available on the Academy website at www.actuary.org/life/phase2.asp.

CA = Provision for amortization of the unamortized surrender charges calculated by the insurer based on each contract’s surrender charge schedule, using prescribed assumptions, except that lapse rates shall be based on the insurer’s prudent estimate, but with no provision for federal income taxes or mortality.

FE = Provision for fixed dollar expenses less fixed dollar revenue calculated using prescribed assumptions, the contract’s actual expense charges, the insurer’s anticipated actual expenses and lapse rates, both estimated on a prudent estimate basis, and with no provision for federal income taxes or mortality.

GC = Provision for the costs of providing the GMDB less net available spread-based charges determined by the formula F×GV-G×AV×R, where GV and AV are as defined in Section 7.C.1.

R = A scaling factor that is a linear function of the ratio of the margin offset to total account charges (W) and takes the form 𝑅𝑅(𝛽𝛽1,𝛽𝛽2) = 𝛽𝛽1 + 𝛽𝛽2 × 𝑊𝑊. The intercept and slope factors for this linear function may vary according to:

• Product type.

• Pro-rata or dollar-for-dollar reductions in guaranteed value following partial withdrawals.

• Fund class.

• Attained age.

• Contract duration.

• Asset-based charges.

• 90% of the ratio of account value to guaranteed value, determined in the aggregate for all contracts sharing the same product characteristics.

Tables of factors for F, G, β1 and β2 values reflecting a 65% confidence interval and ignoring federal income tax are available from the NAIC. In calculating 𝑅𝑅(𝛽𝛽1,𝛽𝛽2) directly from the linear function provided above, the margin ratio W must be constrained to values greater than or equal to 0.2 and less than or equal to 0.6.

Interpolated values of F, G and R (calculated using the linear function described above) for all contracts having the same product characteristics and asset class shall be derived from the pre-

Deleted: E.2.i

Commented [HR12]: The reference to the prudent estimate basis was taken out of Section 4.A.1. It may make more sense to put it back – it is in the OW draft.

Deleted: 3

Deleted: 6

Commented [HR13]: Correct section reference.

Deleted:

Attachment Seven

120

http://www.actuary.org/life/phase2.asp

calculated values using multi-point linear interpolation over the following four contract-level attributes:

a. Attained age.

b. Contract duration.

c. Ratio of account value to GMDB.

d. The total of all asset-based charges, including any fund management fees or allowances based on the underlying variable annuity funds received by the insurer.

The gross asset-based charges for a product shall equal the sum of all contractual asset-based charges plus fund management fees or allowances based on the underlying variable annuity funds received by the insurer determined by complying with the requirements for prudent estimate described in Section 1.E.2.i and revenue sharing described in Section 4.A.5. Net asset-based charges equal gross asset-based charges less any company expenses assumed to be incurred expressed as a percentage of account value. All expenses that would be assumed if a stochastic reserve were being computed as described in Section 4.A.1 should be reflected either in the calculation of the net asset based charges or in the expenses reflected in the calculation of the amount FE.

No adjustment is made for federal income taxes in any of the components listed above.

For purposes of determining the reserve using the Alternative Methodology, any interpretation and application of the requirements of these requirements shall follow the principles discussed in Section 1.B.

B. Calculation of CA and FE

1. General Description

Components CA and FE shall be calculated for each contract, thus reflecting the actual account value and GMDB, as of the valuation date, which is unique to each contract.

Components CA and FE are defined by deterministic “single-scenario” calculations that account for asset growth, interest and inflation at prescribed rates. Mortality is ignored for these two components. Lapse rates shall be determined on a prudent estimate basis as described in Section 1.E.2.i. Lapse rates shall be adjusted by the formula shown below (the dynamic lapse multiplier), which bases the relationship of the GMDB (denoted as GV in the formula) to the account value (denoted as AV in the formula) on the valuation date. Thus, projected lapse rates are smaller when the GMDB is greater than the account value and larger when the GMDB is less than the account value.

−×−= D

AVGVMLMAXUMIN 1,,λ , where U=1, L=0.5, M=1.25, and D=1.1.

Present values shall be computed over the period from the valuation date to contract maturity at a discount rate of 5.75%.

Projected fund performance underlying the account values is as shown in the table below. Unlike the GC component, which requires the entire account value to be mapped, using the fund categorization rules set forth in Section 7.D, to a single “equivalent” asset class (as described in Section 7.D.3), the CA and FE calculation separately projects each variable subaccount (as

Commented [HR14]: Incorrect section reference. Should now reference the VM-01 definition of pudent estimate.

Deleted: 3


Deleted: the

Deleted: CTE amount

Deleted: 3

Deleted: CTE amount

Commented [HR16]: Incorrect section reference. Definitions moved to VM-01.

Deleted: 6


Deleted: 6


Attachment Seven

121

mapped to the eight prescribed categories shown in Section 7.D using the net asset returns shown in the following table). If surrender charges are based wholly on deposits or premiums as opposed to account value, use of this table may not be necessary.

Asset Class/Fund Net Annualized Return

Fixed Account Guaranteed Rate

Money Market 0%

Fixed Income (Bond) 0%

Balanced -1%

Diversified Equity 2%

Diversified International Equity 3%

Intermediate Risk Equity 5%

Aggressive or Exotic Equity 8%

2. Component CA

Component CA is computed as the present value of the projected change in surrender charges plus the present value of an implied borrowing cost of 25 bps at the beginning of each future period applied to the surrender charge at such time.

This component can be interpreted as the “amount needed to amortize the unamortized surrender charge allowance for the persisting policies plus the implied borrowing cost.” By definition, the amortization for non-persisting lives in each time period is exactly offset by the collected surrender charge revenue (ignoring timing differences and any waiver upon death). The unamortized balance must be projected to the end of the surrender charge period using the net asset returns and Dynamic Lapse Multiplier, both as described above, and the year-by-year amortization discounted also as described above. For simplicity, mortality is ignored in the calculations. Surrender charges and free partial withdrawal provisions are as specified in the contract. Lapse and withdrawal rates are determined on a prudent estimate basis and may vary according to the attributes of the business being valued including, but not limited to, attained age, contract duration, etc.

1. Component FE

Component FE establishes a provision for fixed dollar expenses (e.g., allocated costs, including overhead expressed as “per contract” and those expenses defined on a “per contract” basis) less any fixed dollar revenue (e.g., annual administrative charges or contract fees) through the earlier of contract maturity or 30 years. FE is computed as the present value of the company’s assumed fixed expenses projected at an assumed annual rate of inflation starting in the second projection year. This rate grades uniformly from the current inflation rate (CIR) into an ultimate inflation rate of 3% per annum in the 8th year after the valuation date. The CIR is the greater of 3% and the inflation rate assumed for expenses in the company’s most recent asset adequacy analysis for similar business.

C. Calculation of the GC Component

1. GC Factors

GC is calculated as F×GV-G×AV×R, where GV is the amount of the GMDB and AV is the contract account value, both as of the valuation date. F, G and the slope and intercept for the

Deleted: 6


Deleted: <object>

Deleted: <object>

Attachment Seven

122

linear function used to determine R (identified symbolically as β1 and β2) are pre-calculated factors available from the NAIC and known herein as the “pre-calculated factors.” The factors shall be interpolated as described in Section 7.C.6 and modified as necessary as described in Section 7.C.7 and Section 7.C.8.

2. Five Steps

There are five major steps in determining the GC component for a given contract:

a. Classifying the asset exposure, as specified in Section 7.C.3.

b. Determining the risk attributes, as specified in Section 7.C.4 and Section 7.C.5.

c. Retrieving the appropriate nodal factors from the factor grid, as described in Section 7.C.5.

d. Interpolating the nodal factors, where applicable (optional), as described in Section 7.C.6.

e. Applying the factors to the contract values.

3. Classifying Asset Exposure

For purposes of calculating GC (unlike what is done for components CA and FE), the entire account value for each contract must be assigned to one of the eight prescribed fund classes shown in Section 7.D, using the fund categorization rules in Section 7.D.

2. Product Designs

Factors F, G and 𝑅𝑅(𝛽𝛽1,𝛽𝛽2) are available with the pre-calculated factors for the following GMDB product designs:

a. Return of premium (ROP).

b. Premiums less withdrawals accumulated at 3% per annum, capped at 2.5 times premiums less withdrawals, with no further increase beyond age 80 (ROLL3).

c. Premiums less withdrawals accumulated at 5% per annum, capped at 2.5 times premiums less withdrawals, with no further increase beyond age 80 (ROLL5).

d. An annual ratchet design (maximum anniversary value), for which the guaranteed benefit never decreases and is increased to equal the previous contract anniversary account value, if larger, with no further increases beyond age 80 (MAV).

e. A design having a guaranteed benefit equal to the larger of the benefits in designs c and d, above (HIGH).

f. An enhanced death benefit (EDB) equal to 40% of the net earnings on the account (i.e., 40% of account value less total premiums paid plus withdrawals made), with this latter benefit capped at 40% of premiums less withdrawals.

5. Other Attributes

Factors F, G and 𝑅𝑅(𝛽𝛽1,𝛽𝛽2) are available within the pre-calculated factors for the following set of attributes:

Deleted: 6


Deleted: 6


Deleted: 6


Deleted: 6


Deleted: 6


Deleted: 6


Deleted: 6


Deleted: 6


Deleted: 6

Deleted: 6


Commented [LE29]: The numbering is off. This should be numbered 4 not 2.

Attachment Seven

123

a. Two partial withdrawal rules—one for contracts having a pro-rata reduction in the GMDB and another for contracts having a dollar-for-dollar reduction.

b. The eight asset classes described in Section 7.D.2.

c. Eight attained ages, with a five-year age setback for females.

d. Five contract durations.

e. Seven values of GV/AV.

f. Three levels of asset-based income.

6. Interpolation of F, G and 𝑅𝑅(𝛽𝛽1,𝛽𝛽2)

a. Apply to a contract having the product characteristics listed in Section 7.E.1 and shall be determined by selecting values for the appropriate partial withdrawal rule and asset class and then using multipoint linear interpolation among published values for the last four attributes shown in Section 7.C.5.

b. Interpolation over all four dimensions is not required, but if not performed over one or more dimensions, the factor used must result in a conservative (higher) value of GC. However, simple linear interpolation using the AV÷GV ratio is mandatory. In this case, the company must choose nodes for the other three dimensions according to the following rules: next highest attained age, nearest duration and nearest annualized account charge differential, as listed in Section 7.E.3 (i.e., capped at +100 and floored at –100 bps).

c. For 𝑅𝑅(𝛽𝛽1,𝛽𝛽2), the interpolation should be performed on the scaling factors R calculated using β1, β2, using the ratio of margin offset to total asset charges (W), not on the factors β1 and β2 themselves.

d. An Excel workbook, Excel add-in and companion dynamic link library (.dll) program is available from the NAIC that can be used to determine the correct values and perform the multipoint linear interpolation.

e. Alternatively, published documentation can be referenced on performing multipoint linear interpolation and the required 16 values determined using a key that is documented in the table Components of Key Used for GC Factor Look-Up located in Section 7.E.3.

7. Adjustments to GC for Product Variations and Risk Mitigation/Transfer

In some cases, it may be necessary to make adjustments to the published factors due to:

a. A variation in product form wherein the definition of the guaranteed benefit is materially different from those for which factors are available. (See Section 7.C.8.)

b. A risk mitigation or other management strategy, other than a hedging strategy, that cannot be accommodated through a straightforward and direct adjustment to the published values.

Adjustments may not be made to GC for hedging strategies.

Any adjustments to the published factors must be fully documented and supported through stochastic analysis. Such analysis may require stochastic simulations, but

Deleted: ¶

Deleted: 6


Deleted: 6


Deleted: 6


Deleted: 6

Commented [LE33]: Correct reference

Deleted: 6


Deleted: 6


Attachment Seven

124

would not ordinarily be based on full in-force projections. Instead, a representative “model office” should be sufficient. Use of these adjusted factors must be supported by a periodic review of the appropriateness of the assumptions and methods used to perform the adjustments, with changes made to the adjustments when deemed necessary by such review.

Note that minor variations in product design do not necessarily require additional effort. In some cases, it may be reasonable to use the factors/formulas for a different product form (e.g., for a roll-up GMDB near or beyond the maximum reset age or amount, the ROP GMDB factors/formulas shall be used, possibly adjusting the guaranteed value to reflect further resets, if any). In other cases, the reserves may be based on two different guarantee definitions and the results interpolated to obtain an appropriate value for the given contract/cell. Likewise, it may be possible to adjust the Alternative Methodology results for certain risk transfer arrangements without significant additional work (e.g., quota-share reinsurance without caps, floors or sliding scales would normally be reflected by a simple pro-rata adjustment to the “gross” GC results).

However, if the contract design is sufficiently different from those provided and/or the risk mitigation strategy is nonlinear in its impact on the reserve, and there is no practical or obvious way to obtain a good result from the prescribed factors/formulas, any adjustments or approximations must be supported using stochastic modeling. Notably this modeling need not be performed on the whole portfolio, but can be undertaken on an appropriate set of representative policies.

8. Adjusting F and G for Product Design Variations

This subsection describes the typical process for adjusting F and G factors due to a variation in product design. Note that R (as determined by the slope and intercept terms in the factor table) would not be adjusted.

a. Select a contract design among those described in Section 7.C.4 that is similar to the product being valued. Execute cash-flow projections using the documented assumptions (see table of Liability Modeling Assumptions & Product Characteristics in Section 7.E.1 and table of Asset-Based Fund Charges in Section 7.E.2) and the prepackaged scenarios for a set of representative cells (combinations of attained age, contract duration, asset class, AV/GMDB ratio and asset-based charges). These cells should correspond to nodes in the table of precalculated factors. Rank (order) the sample distribution of results for the present value of net cost. Determine those scenarios that comprise CTE (65).

Guidance Note: Present value of net cost = PV [guaranteed benefit claims in excess of account value] – PV [margin offset]. The discounting includes cash flows in all future years (i.e., to the earlier of contract maturity and the end of the horizon).

b. Using the results from step 1, average the present value of cost for the CTE (65) scenarios and divide by the current guaranteed value. For the Jth cell, denote this value by FJ. Similarly, average the present value of the margin offset revenue for the same subset of scenarios and divide by account value. For the Jth cell, denote this value by GJ.

c. Extract the corresponding precalculated factors. For each cell, calibrate to the published tables by defining a “model adjustment factor” (denoted by asterisk) separately for the “cost” and “margin offset” components:

Deleted: CTE amount

Deleted: 6

Commented [LE36]: correct reference

Deleted: 6


Deleted: 6


Attachment Seven

125

𝐹𝐹𝐽𝐽∗ = 𝑓𝑓(𝜃𝜃�)𝐹𝐹𝐽𝐽

and 𝐺𝐺𝐽𝐽∗ = 𝑔𝑔�(𝜃𝜃�)𝐺𝐺𝐽𝐽

d. Execute “product specific” cash-flow projections using the documented assumptions and prepackaged scenarios for the same set of representative cells. Here, the company should model the actual product design. Rank (order) the sample distribution of results for the present value of net cost. Determine those scenarios that comprise CTE (65).

e. Using the results from step d, average the present value of cost for the CTE (65) scenarios and divide by the current guaranteed value. For the Jth cell, denote this value by 𝐹𝐹�𝐽𝐽. Similarly, average the present value of margin offset revenue for the same subset of scenarios and divide by account value. For the Jth cell, denote this value by ��𝐺𝐽𝐽 .

f. To calculate the reserve for the specific product in question, the company should implement the Alternative Methodology as documented, but use 𝐹𝐹�𝐽𝐽 × 𝐹𝐹𝐽𝐽∗ in place of F and ��𝐺𝐽𝐽 × 𝐺𝐺𝐽𝐽∗ instead of G. The same R factors as appropriate for the product evaluated in step 1 shall be used for this step (i.e., the product used to calibrate the cash-flow model).

9. Adjusting GC for Mortality Experience

The factors that have been developed for use in determining GC assume male mortality at 100% of the 1994 Variable Annuity MGDB ALB Mortality Table. Companies electing to use the Alternative Methodology that have not conducted an evaluation of their mortality experience shall use these factors. Other companies should use the procedure described below to adjust for the actuary’s prudent estimate of mortality. The development of prudent estimate mortality shall follow the requirements and guidance of Section 12. Once a company uses the modified method for a block of business, the option to use the unadjusted factors is no longer available for that part of its business. In applying the factors to actual in-force business, a five-year age setback should be used for female annuitants.

a. Develop a set of mortality assumptions based on prudent estimate. In setting these assumptions, the actuary shall be guided by the definition of prudent estimate and the principles discussed in Sections 10 and 11.

b. Calculate two sets of NSPs at each attained age: one valued using 100% of the 1994 Variable Annuity MGDB Age Last Birthday (ALB) Mortality Table (with the aforementioned five-year age setback for females) and the other using prudent estimate mortality. These calculations shall assume an interest rate of 3.75% and a lapse rate of 7% per year.

c. The GC factor is multiplied by the ratio, for the specific attained age being valued, of the NSP calculated using the prudent estimate mortality to the NSP calculated using the 1994 Variable Annuity MGDB ALB Mortality Table (with the aforementioned five-year age setback for females).

D. Fund Categorization

1. Criteria

The following criteria should be used to select the appropriate factors, parameters and formulas for the exposure represented by a specified guaranteed benefit. When available, the volatility of the long-term annualized total return for the fund(s)—or an appropriate benchmark—should

Deleted: CTE amount

Deleted: 11

Deleted: 12

Commented [HR39]: Original section references were correct – reject edits.

Attachment Seven

126

conform to the limits presented. For this purpose, “long-term” is defined as twice the average projection period that would be applied to test the product in a stochastic model (generally, at least 30 years).

Where data for the fund or benchmark are too sparse or unreliable, the fund exposure should be moved to the next higher volatility class than otherwise indicated. In reviewing the asset classifications, care should be taken to reflect any additional volatility of returns added by the presence of currency risk, liquidity (bid – ask) effects, short selling and speculative positions.

2. Asset Classes

Variable subaccounts must be categorized into one of the following eight asset classes. For purposes of calculating CA or FE, each contract will have one or more of the following asset classes represented, whereas for component GC, all subaccounts will be mapped into a single asset class.

a. Fixed account: This class is credited interest at guaranteed rates for a specified term or according to a “portfolio rate” or “benchmark” index. This class offers a minimum positive guaranteed rate that is periodically adjusted according to company policy and market conditions.

b. Money market/short-term: This class is invested in money market instruments with an average remaining term-to-maturity of less than 365 days.

c. Fixed income: This class is invested primarily in investment grade fixed income securities. Up to 25% of the funds within this class may be invested in diversified equities or high-yield bonds. The expected volatility of the returns for this class will be lower than the balanced fund class.

d. Balanced: This class is a combination of fixed income securities with a larger equity component. The fixed income component should exceed 25% of the portfolio. Additionally, any aggressive or “exotic” equity component should not exceed one-third (33.3%) of the total equities held. Should the fund violate either of these constraints, it should be categorized as an equity fund. This class usually has a long-term volatility in the range of 8%–13%.

e. Diversified equity: This class is invested in a broad-based mix of U.S. and foreign equities. The foreign equity component (maximum 25% of total holdings) must be comprised of liquid securities in well-developed markets. Funds in this class would exhibit long-term volatility comparable to that of the S&P 500. These funds should usually have a long-term volatility in the range of 13%–18%.

f. Diversified international equity: This class is similar to the diversified equity class, except that the majority of fund holdings are in foreign securities. This class should usually have a long-term volatility in the range of 14%–19%.

g. Intermediate risk equity: This class has a mix of characteristics from both the diversified and aggressive equity classes. This class has a long-term volatility in the range of 19%–25%.

h. Aggressive or exotic equity: This class comprises more volatile funds where risk can arise from: underdeveloped markets, uncertain markets, high volatility of returns, narrow focus (e.g., specific market sector), etc. This class (or market benchmark) either does not have sufficient history to allow for the calculation of a long-term expected

Attachment Seven

127

volatility, or the volatility is very high. This class would be used whenever the long-term expected annualized volatility is indeterminable or exceeds 25%.

3. Selecting Appropriate Investment Classes

The selection of an appropriate investment type should be done at the level for which the guarantee applies. For guarantees applying on a deposit-by-deposit basis, the fund selection is straightforward. However, where the guarantee applies across deposits or for an entire contract, the approach can be more complicated. In such instances, the approach is to identify for each contract where the “grouped holdings” fit within the categories listed and to classify the associated assets on this basis.

A seriatim process is used to identify the “grouped” fund holdings, to assess the risk profile of the current fund holdings (possibly calculating the expected long-term volatility of the funds held with reference to the indicated market proxies) and to classify the entire “asset exposure” into one of the specified choices. Here, “asset exposure” refers to the underlying assets (separate and/or general account investment options) on which the guarantee will be determined. For example, if the guarantee applies separately for each deposit year within the contract, then the classification process would be applied separately for the exposure of each deposit year.

In summary, mapping the benefit exposure (i.e., the asset exposure that applies to the calculation of the guaranteed minimum death benefits) to one of the prescribed asset classes is a multistep process:

a. Map each separate and/or general account investment option to one of the prescribed asset classes. For some funds, this mapping will be obvious, but for others, it will involve a review of the fund’s investment policy, performance benchmarks, composition and expected long-term volatility.

b. Combine the mapped exposure to determine the expected long-term “volatility of current fund holdings.” This will require a calculation based on the expected long-term volatility for each fund and the correlations between the prescribed asset classes as given in the table “Correlation Matrix for Prescribed Asset Classes” in Section 7.D.4.

c. Evaluate the asset composition and expected volatility (as calculated in step b) of current holdings to determine the single asset class that best represents the exposure, with due consideration to the constraints and guidelines presented earlier in this section.

d. In step a, the company should use the fund’s actual experience (i.e., historical performance, inclusive of reinvestment) only as a guide in determining the expected long-term volatility. Due to limited data and changes in investment objectives, style and/or management (e.g., fund mergers, revised investment policy, different fund managers, etc.), the company may need to give more weight to the expected long-term volatility of the fund’s benchmarks. In general, the company should exercise caution and not be overly optimistic in assuming that future returns will consistently be less volatile than the underlying markets.

e. In step b, the company should calculate the “volatility of current fund holdings” (for the exposure being categorized) by the following formula:

∑∑= =

=n

i

n

jjiijji ww

1 1σσρσ

Using the volatilities and correlations in the following table where

Deleted: 6


Deleted: ,

Attachment Seven

128

∑

=

kk

ii AV

AVw

is the relative value of fund i expressed as a proportion of total contract value, ijρ is

the correlation between asset classes i and j, and iσ is the volatility of asset class i. An example is provided after the table.

4. Correlation Matrix for Prescribed Asset Classes

Annual Volatility Fixed

Account Money Market

Fixed Income Balanced Diverse

Equity Intl Equity Interm Equity

Aggr Equity

1.0% Fixed Account 1 0.50 0.15 0 0 0 0 0

1.5% Money Market 0.50 1 0.20 0 0 0 0 0

5.0% Fixed Income 0.15 0.20 1 0.30 0.10 0.10 0.10 0.05

10.0% Balanced 0 0 0.30 1 0.95 0.60 0.75 0.60

15.5% Diverse Equity 0 0 0.10 0.95 1 0.60 0.80 0.70

17.5% Intl Equity 0 0 0.10 0.60 0.60 1 0.50 0.60

21.5% Interm Equity 0 0 0.10 0.75 0.80 0.50 1 0.70

26.0% Aggr Equity 0 0 0.05 0.60 0.70 0.60 0.70 1

Attachment Seven

129

5. Fund Categorization Example

As an example, suppose three funds (fixed income, diversified U.S. equity and aggressive equity) are offered to clients on a product with a contract level guarantee (i.e., across all funds held within the contract). The current fund holdings (in dollars) for five sample contracts are shown in the following table:

1 2 3 4 5

MV Fund X (Fixed Income) 5,000 4,000 8,000 - 5,000

MV Fund Y (Diversified Equity) 9,000 7,000 2,000 5,000 -

MV Fund Z (Aggressive Equity) 1,000 4,000 - 5,000 5,000

Total Market Value 15,000 15,000 10,000 10,000 10,000

Total Equity Market Value 10,000 11,000 2,000 10,000 5,000

Fixed Income % (A) 33% 27% 80% 0% 50%

Fixed Income Test (A > 75%) No No Yes No No

Aggressive % of Equity (B) 10% 36% n/a 50% 100% Balanced Test (A > 25% & B < 33.3%) Yes No n/a No No

Volatility of Current Fund Holdings 10.9% 13.2% 5.3% 19.2% 13.4%

Fund Classification Balanced Diversified1 Fixed Income Intermediate Diversified

As an example, the “volatility of current fund holdings” for contract #1 is calculated as √𝐴𝐴 + 𝐵𝐵 where:

( ) ( ) ( )26.0155.07.0151

159226.005.005.0

151

1552155.005.01.0

159

1552

226.0

1512

155.01592

05.0155

××

⋅⋅+××

⋅⋅+××

⋅⋅=

×+

×+

×=

B

A

So, the volatility for contract #1 = √0.0092 + 0.0026 = 0.109 or 10.9%

1 Although the volatility suggests “balanced fund,” the balanced fund criteria were not met. Therefore, this “exposure” is moved “up” to diversified equity. For those funds classified as diversified equity, additional analysis would be required to assess whether they should be instead designated as “diversified international equity.”

Deleted: Section Break (Next Page)

Attachment Seven

130

E. Tables

1. Liability Modeling Assumptions and Product Characteristics used for GC Factors

Asset Based Charges (MER) Vary by fund class. See Section 7.E.2.

Base Margin Offset 100 bps per annum.

GMDB Description

1. ROP = return of premium. 2. ROLL3 = 3% roll-up, capped at 2.5×premium, frozen at age 80. 3. ROLL5 = 5% roll-up, capped at 2.5×premium, frozen at age 80. 4. MAV = annual ratchet (maximum anniversary value), frozen at age 80. 5. HIGH = higher of 5% roll-up and annual ratchet. 6. EDB = 40% enhanced death benefit (capped at 40% of deposit). Note that the

pre-calculated factors were originally calculated with a combined ROP benefit, but they have been adjusted to remove the effect of the ROP. Thus, the factors for this benefit five are solely for the EDB.

Adjustment to GMDB Upon Partial Withdrawal Separate factors for “pro-rata by market value” and “dollar-for-dollar.”

Surrender Charges Ignored (i.e., zero). Included in the CA component.

Single Premium/Deposit $100,000. No future deposits; no intra-contract fund rebalancing.

Base Contract Lapse Rate (Total Surrenders)

Pro-rata by MV: 10% p.a. at all contact durations (before dynamics). Dollar-for-dollar: 2% p.a. at all contract durations (no dynamics).

Partial Withdrawals Pro-rata by MV: None (i.e., zero). Dollar-for-dollar: Flat 8% p.a. at all contract durations (as a % of AV). No dynamics or anti-selective behavior.

Mortality

100% of the 1994 Variable Annuity MGDB Mortality Table (MGDB 94 ALB). For reference, 1000qx rates at ages 65 and 70 for 100% of MGDB 94 ALB Male are 18.191 and 29.363, respectively. Note: Section 7.C.9 allows modification to this assumption.

Gender/Age Distribution 100% male. Methodology accommodates different attained ages. A five-year age setback will be used for female annuitants.

Max. Annuitization Age All policies terminate at age 95.

Fixed Expenses Ignored (i.e., zero). Included in the FE component.

Annual Fee and Waiver Ignored (i.e., zero). Included in the FE component.

Discount Rate 5.75% pre-tax.

Dynamic Lapse Multiplier (Applies only to policies where GMDB is adjusted “pro-rata by MV” upon withdrawal)

−×−= D

AVGVMLMAXUMIN 1,,λ

U = 1, L = 0.5, M = 1.25, D = 1.1 Applied to the “Base Contract Lapse Rate.” Does not apply to partial withdrawals.

Deleted: 6


Deleted: 6


Attachment Seven

131

2. Asset-Based Fund Charges (bps per annum)

3. Components of Key Used for GC Factor Look-Up

(First Digit always “1”) Contract Attribute Key: Possible Values and Description

Product Definition, P

0 : 0 Return-of-premium. 1 : 1 Roll-up (3% per annum). 2 : 2 Roll-up (5% per annum). 3 : 3 Maximum anniversary value (MAV). 4 : 4 High of MAV and 5% roll-up. 5 : 5 Enhanced death benefit (excludes the ROP GMDB,

which would have to be added separately if the contract in question has an ROP).

GV Adjustment Upon Partial Withdrawal, A

0 : 0 Pro-rata by market value. 1 : 1 Dollar-for-dollar.

Fund Class, F

0 : 0 Fixed Account. 1 : 1 Money Market. 2 : 2 Fixed Income (Bond). 3 : 3 Balanced Asset Allocation. 4 : 4 Diversified Equity. 5 : 5 International Equity. 6 : 6 Intermediate Risk Equity. 7 : 7 Aggressive/Exotic Equity.

Attained Age (Last Birthday), X

0 : 35 4 : 65 1 : 45 5 : 70 2 : 55 6 : 75 3 : 60 7 : 80

Contract Duration (years-since-issue), D

0 : 0.5 3 : 9.5 1 : 3.5 4 : 12.5 2 : 6.5

Account Value-to-Guaranteed Value Ratio, φ

0 : 0.25 4 : 1.25 1 : 0.50 5 : 1.50 2 : 0.75 6 : 2.00 3 : 1.00

Annualized Account Charge Differential from Section 7.E.2 Assumptions

0 : −100 bps 1 : +0 2 : +100

Asset Class/Fund Account Value Charge Fixed Account 0 Money Market 110 Fixed Income (Bond) 200 Balanced 250 Diversified Equity 250 Diversified International Equity 250 Intermediate Risk Equity 265 Aggressive or Exotic Equity 275

Deleted: 1 : 3.5

Deleted: 3 : 9.5

Deleted: 4 : 12.5

Commented [HR43]: Contract Duration should be arranged to read down Column 1, then down Column 2, like all the other panels

Deleted: 6


Attachment Seven

132