Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
“There’s No Accounting for Quants”
Evan S. GraceDiane M. Garnick
Presentation to Northfield Conference
March 16, 2003
2
Efficiency at Risk
n Earnings Complexity is on the rise
§ Options Accounting
§ Pension Accounting
§ Transition Alternatives
n Fewer Analysts on the job
§ Record lay offs
§ Increase in names per analyst
§ GAAP not generally accepted
n Positioning for Neglect
§ Size is key
§ Less potential for IB business
Return of the Neglect and Revision Effects?
3
Theoretical Foundations of Earnings Complexity
nWhat are reasonable assumptions to make about option expenses?
nHas the actuarial smoothing embedded in pension accounting diluted representational faithfulness of corporate liabilities?
nWhat are the implications of alternative transition methods for accounting standards?
nWhat are the implications for investors?
Theory vs. Empirical Evidence
4
What are reasonable assumptions to make about options expenses?
Theoretical Foundations of Earnings Complexity
nWhat investors need to make rational decisions:
§ Transparency
§ Timely access to data
§ Representational faithfulness
5
What Do We Have? Many Options for Options
Theoretical Foundations of Earnings Complexity
nCalculation methods can vary;
§ Intrinsic Value
§ Fair Value
n If fair value is used, dates can vary;
§ Grant Date
§ Exercise Date
§ Vesting Date
nAvailability of data varies;
§ Appear on the face of the financials
§ Disclose in the footnotes
§ Long delays in data release diminishes its usefulness
§ Currently annually only, on the 10-K
6
Data are difficult to obtain; more difficult to analyze
Theoretical Foundations of Earnings Complexity
nRelatively unique: different treatments for same item
§ APB 25 (issued 1973)
§ Requires intrinsic value accounting– No charge for at-the-money option issuance
§ SFAS 123 (issued 1996)
§ “Prefers” fair value accounting as of grant date
§ Spread cost over vesting period
§ Can still use APB 25 if disclose impact, but no method proscribed– EIX Volatility range was 17% -52%
§ SFAS 148 (issued December 2002)
§ Better, more frequent disclosure
§ Three transition methods– Prospective, Modified Prospective, Retroactive Restatement
7
n Inefficient market reactions
§ Wal*Mart will begin expensing options as of 2/1/03
§ Net effect is a $0.02-$0.03 incremental expense for 2003
n The debate creates confusion in the marketplace
§ Bulls: it’s just an accounting change
§ Bears: it was always an expense, now we’re counting it
§ Quants: it’s a negative estimate revision
Theoretical Foundations of Earnings Complexity
What are the implications of alternative accounting methods?
8
Theoretical Foundations of Earnings Complexity
Inefficient Market ReactionsWal*Mart Relative to S&P500
90
92
94
96
98
100
102
10401
/31/
2003
02/0
1/20
03
02/0
2/20
03
02/0
3/20
03
02/0
4/20
03
02/0
5/20
03
02/0
6/20
03
02/0
7/20
03
02/0
8/20
03
02/0
9/20
03
02/1
0/20
03
02/1
1/20
03
02/1
2/20
03
02/1
3/20
03
02/1
4/20
03
02/1
5/20
03
02/1
6/20
03
02/1
7/20
03
02/1
8/20
03
02/1
9/20
03
02/2
0/20
03
Inde
xed
to 1
00 o
n Ja
nuar
y 31
, 200
3
S&P 500
Wal*Mart
WMT will begin expensing options as of 2/1/03 following FAS 123. This will result in a reduction to EPS estimated at $0.02-$0.03 for the year, equally distributed.
9
Theoretical Foundations of Earnings Complexity
nOptions Expensing regained importance as the percentage of earnings represented by options increased
Impact of Options on Core Earnings
Historical S&P 500
5.40
6.62
2.92
3.55
4.15
6.15
-1.27 -1.30 -1.32 -1.31 -1.28 -1.32
20%
45%
22%
37%
31%
24%
-2.00
-1.00
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
06/30/2002 03/31/2002 12/31/2001 09/30/2001 06/30/2001 03/31/2001
S&
P C
ore
Ear
nin
gs
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
Op
tio
n E
xpen
se %
Ear
nin
gs
Core EPSOption Expense PS
Percentage Options Represent
10
Theoretical Foundations of Earnings Complexity
nDifferent pattern of income recognition
§ Prospective: Expense gradually increases
§ Nearly all “early adopters” use this; to be phased out in 2003
§ Modified Prospective: Sudden one-time increase in expense
§ Retroactive Restatement: Restate history such that no change over time
§ Example: Coca Cola (source: Bear Stearns)
§ July 2002: Will use prospective method (only choice)– $0.01 per share impact in 2002
§ Post-FAS 148: Changed to modified prospective method– $0.11 per share impact in 2002– Old expense would have “ramped up” to $0.11 over time– Impact on profit growth, momentum, possibly revisions
Implications of Alternative Transition Methods
11
Theoretical Foundations of Earnings Complexity
n It’s not in the databases
§ Traditional quant solutions won’t work
§ Sell side
§ Accounting research suddenly popular, for some reason...
§ Big resource allocation to options
§ S&P
§ http://www/standardandpoors.com
§ Interns/Fundamental analysts
So what can you do about it?
12
Pensions: Income Statement Impact
n Accounting designed to smooth changes in projected benefit obligation funding status
§ Not necessarily a bad thing, but can be abused
§ Fine art of “Actuarial Smoothing”
n 52 companies in the S&P 500 have positive net pension income, but pension plans in deficit
n “Net pension charge/credit” is in operating income, but not a separate line item
§ Currently reported in either COGS or SG&A
n Example: Pension assumptions impacting EPS growth
§ Cinergy Net Income Growth ‘00-’01 = $42 million
§ Pension expense declined $8 million
§ 20% of net income growth due to pensions
13
Pensions: Balance Sheet Impact
n Theory: Matching Principle
§ Expense future pension commitments of current employees
§ Set aside current assets to pay for future obligations
n Reality: This asset often doesn’t reflect the true funding level
§ Substantial off balance sheet items
§ Unrecognized net actuarial loss
§ Unrecognized prior service cost
§ Unrecognized transition obligation/asset
n Example: GM is the poster child
§ Balance Sheet asset: “Prepaid benefit Cost” = $18.3 Billion
§ Off-Balance Sheet: “Funded Status” = -$12.6 Billion
14
Pensions: Funding Crisis?
nBalance Sheet Funding Status improving, but mixedHistorical Funded Status of Current S&P 500 Companies
205
17
88
108
54
28
184
20
97
139
44
16
147
54
181
80
29
9
0
50
100
150
200
250
< 50% 50% - 75% 75% - 100% 100% - 125% 125% - 150% > 150%
1991
19962001
15
Pensions: Funding Requirements
n Key Question: Are contributions necessary? If so, when?
§ ERISA requirement: underfunded if actuarial value of assets is less than 90% of current plan liabilities
§ Funding deficiency subject to excise tax of 5-10%
§ Need IRS Form 5500 to determine this– Filed 7-9.5 months after end of plan year
§ Different calculations for funding decision than for annual report– Different actuarial methods, assumptions, measurement dates
» Example: DJIA average interest rate for funding calculation 100 basis points lower than average interest rate for other financial reporting
§ PBGC Requirement: Underfunded plans must pay additional premiums
§ Additional $9/$1,000 underfunded vested benefits
16
Pensions: What’s a Quant To Do?n Set aside the question of whether pensions “matter” for the market
§ Different treatment of the same obligation for different stocks is the stock picker’s focus
n Simplest screen: Got a DB plan?
§ 142 S&P 500 companies with no DB plan (25 % of mcap)
§ 1,386 Russell 2000 companies with no DB plan (65 % of mcap)
n Size up the potential impact (all in Compustat)
§ Plan Assets/Market Cap
§ Expected Return of Plan Assets
§ Discount Rate
§ Interest Cost
§ Rate of Compensation Increase
17
Pensions: What’s a Quant To Do?
n Sell side
§ As before, sudden, inexplicable interest in accounting research...
§ Can provide line item data not easily available elsewhere
n Interns/Fundamental Analysts
§ Tear through Ks and Qs
§ Adjust valuation/growth metrics for pension distortions
18
Transition Alternatives
nCommon themes with Options and Pensions
§ Data availability is limited; accuracy is questionable
§ Accounting changes impair comparability
§ Cross-sectional and time series data set are impaired
19
Transition Alternatives
n Frequency of changes in accounting data impacts validity of back testsS&P500 Earnings Per Share
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
1990
Q1
1990
Q3
1991
Q1
1991
Q3
1992
Q1
1992
Q3
1993
Q1
1993
Q3
1994
Q1
1994
Q3
1995
Q1
1995
Q3
1996
Q1
1996
Q3
1997
Q1
1997
Q3
1998
Q1
1998
Q3
1999
Q1
1999
Q3
2000
Q1
2000
Q3
2001
Q1
2001
Q3
2002
Q1
2002
Q3
Some of the "lumpiness" in earnings is attributable to changes in accounting standards
FASB 112 Employers ' Accounting for Postemployment
Benef i ts—an amendment of FASB Statements No. 5 and 43
(November 1992)
FASB 118 Accounting by Creditors for Impairment of a Loan—Income Recognit ion and Disclosures—an amendment o f FASB Statement
No. 114 (October 1994)
FASB 123 Account ing for Stock-Based Compensat ion (October 1995)
FASB 128 Earnings per Share (February 1997)
FASB 142 Goodwil l and Other Intangible Assets (June 2001)
20
Transition Alternatives
S & P 5 0 0 R e p o r t e d v s . O p e r a t i n g E a r n i n g s
- 1 . 0 0
0 . 0 0
1 . 0 0
2 . 0 0
3 . 0 0
4 . 0 0
5 . 0 0
Mar
-77
Mar
-78
Mar
-79
Mar
-80
Mar
-81
Mar
-82
Mar
-83
Mar
-84
Mar
-85
Mar
-86
Mar
-87
Mar
-88
Mar
-89
Mar
-90
Mar
-91
Mar
-92
Mar
-93
Mar
-94
Mar
-95
Mar
-96
Mar
-97
Mar
-98
Mar
-99
Mar
-00
Mar
-01
Dif
fere
nce
R e p o r t e d e a r n i n g s a r e g r e a t e r t h a n o p e r a t i n g
O p e r a t i n g e a r n i n g s a r e g r e a t e r t h a n r e p o r t e d
T h e s p r e a d b e t w e e n o p e r a t i n g a n d r e p o r t e d e a r n i n g s h a s r e c e n t l y h i t a n a l l t i m e h i g h
nAlternative transition methods fuel credibility concerns§ The spread between reported and operating earnings is near an all time high
§ Integrity of models that rely on accounting data without regard to multiple transition alternatives needs to be questioned
21
Common Thread: What are Earnings?
nNo clear, agreed upon definition§ Operating?
§ Best reflection of “continuing operations” of the company
§ Subjective. Too many recurring non-recurring items
§ Reported?
§ Pure in some sense; it’s what actually happened
§ Can be misleading; want to know about core operations
§ Core?
§ Right idea, bad execution. Likely to be revised
22
Common Thread: What are Earnings?
nWould the real P/E ratio please stand up?S&P 500 Forward P/E Ratio
15.6
21.5
28.1
0
5
10
15
20
25
30
Operating Reported Core
EPS Metric
Fo
rward
P/E
23
Common Thread: What are Earnings?
nBack Test ImplicationsP/E vs. P/EBITDA
Calendar Year 2002 Performance
-60.00
-50.00
-40.00
-30.00
-20.00
-10.00
0.00
10.00Q1 Q5 Q1-Q5
Quartile
Tota
l Ret
urn
(%)
P/E
P/EBITDA
24
It’s Not Just Tech
n “Pro-forma” earnings have an impact across all sectorsS&P 500:
Percentage of Companies where GAAP=Pro Forma
0
10
20
30
40
50
60
70
80
90
100
1Q00 2Q00 3Q00 4Q00 1Q01 2Q01 3Q01 4Q01 1Q02 2Q02 3Q02 4Q02
25
What is to be Done?
nSector-Specific Models
§ Solves problem of cross-sector comparisons
§ Define sectors carefully: also includes large industries
§ Comparability across time can still be a challenge
§ Check for consistency in returns
§ Dynamic weighting schemes
nEarnings Quality Variables
n Independent variables additive to quant models
n Identify potential for manipulation
26
What is to be Done?
nTotal Net Accruals is a very promising factorNet Accruals Factor: Top Decile vs. Bottom Decile
0.00
50.00
100.00
150.00
200.00
250.00
300.00
350.00
400.00
450.00
91:4
92:1
92:2
92:3
92:4
93:1
93:2
93:3
93:4
94:1
94:2
94:3
94:4
95:1
95:2
95:3
95:4
96:1
96:2
96:3
96:4
97:1
97:2
97:3
97:4
98:1
98:2
98:3
98:4
99:1
99:2
99:3
99:4
00:1
00:2
00:3
00:4
27
What is to be Done?
nCall your Congressman (or FASB)§ Mandatory comment period for all new “FAS” standards
§ Typically one-sided commentary skews FASBs positions.
§ Have your voice heard!