The Federal Home Loan Bank System: The Lender of Next-to-Last Resort?

ADAM ASHCRAFT

MORTEN L. BECH

W. SCOTT FRAME

The Federal Home Loan Bank System: The Lender

of Next-to-Last Resort?

The Federal Home Loan Bank (FHLB) System is a large cooperatively ownedgovernment-sponsored liquidity facility that lends predominately to U.S.depository institutions. This paper documents the significant role played bythe FHLB System at the outset of the recent financial crisis and providesevidence on the uses of FHLB funding by member banks and thrifts duringthat time. We then compare lending activity by the FHLB System and theFederal Reserve during 2007 and 2008, discuss the types of institutionsseeking government-sponsored liquidity at various times, and identify thetrade-offs faced by borrowers eligible to tap liquidity from both facilities.

JEL codes: E4, E5, G21, G28Keywords: government-sponsored enterprise, lender of last resort,

liquidity.

IN JULY 2007, THE CREDIT RATING AGENCIES (Standard & Poors,Moody’s, and Fitch) responded to the rapid deterioration in the performance of re-cently originated U.S. subprime mortgages by taking a historical downgrade actionon the entire sector of associated mortgage-backed securities (MBS). This downgradehad global implications.

The views expressed do not necessarily reflect those of the Federal Reserve Banks of Atlanta or NewYork or the Federal Reserve System. Helpful comments have been provided by Nuno Cassola, DeborahLucas (the editor), Stuart Gabriel, Thomas Mosk, Frank Packer, Ned Prescott, Martijn Schrijvers, LarryWall, Larry White, an anonymous reviewer, and seminar participants at the 2009 CEPR/JFI Conference onthe Financial Crisis, the 2009 Federal Reserve Bank of Chicago Bank Structure Conference, the 2009 FMAEuropean Conference, the Banque de France, the Federal Reserve Banks of Atlanta, Boston, Dallas, NewYork, and Philadelphia, and Tilburg University. We thank Dennis Kuo for excellent research assistance.

ADAM ASHCRAFT is a Vice President at the Research Department, Federal Reserve Bank ofNew York (E-mail: [email protected]). MORTEN L. BECH is a Research Officer at theResearch Department, Federal Reserve Bank of New York (E-mail: [email protected]).W. SCOTT FRAME is a Financial Economist and Policy Advisor at the Research Department,Federal Reserve Bank of Atlanta (E-mail: [email protected]).

Received December 3, 2008; and accepted in revised form November 16, 2009.

Journal of Money, Credit and Banking, Vol. 42, No. 4 (June 2010)C© 2010 The Ohio State University

552 : MONEY, CREDIT AND BANKING

Many of the very largest U.S. and European financial institutions were directlyexposed to the subprime mortgage market through loans to subprime originatorsand investments in the senior tranches of subprime MBS and collateralized debtobligations, the latter of which was largely secured by the subordinate tranches ofsubprime MBS. These same institutions were also indirectly exposed through theirsponsorship of structured investment vehicles and asset-backed commercial paper(ABCP) conduits that purchased subprime MBS, as well as through exposures totheir trading counterparties who had similar problems.

The ratings action also triggered a loss of confidence by investors in a broad arrayof structured finance products, and related selling and hedging activity put additionaldownward pressure on the prices of a broad range of asset-backed securities. Mark-to-market accounting rules, in turn, resulted in the recognition of large accountinglosses and a material deterioration in capital positions of the most affected institu-tions. Uncertainty about the ultimate level of losses faced by individual institutionsprompted investors to pull their funding from potentially affected institutions, therebyleading to a sharp increase in the cost and a significant reduction in the availabilityof term funding. This stress in term funding markets was key because the inability ofinstitutions to access term credit—concurrent with the breakdown of the originate-to-distribute model of financial intermediation that left them with unexpected assets ontheir balance sheets—would impair the ability of these institutions to originate newcredit and amplify the effect of the correction in the mortgage and housing markets.

When faced with such liquidity shocks, conventional wisdom holds that agovernment-sponsored liquidity provider (e.g., the central bank) should be avail-able to act as a lender of last resort by providing required funding on a collateralizedbasis.1 During the ongoing financial crisis, the Federal Reserve has indeed played therole of a lender of last resort and has provided substantial amounts of liquidity to thefinancial system. However, at the outset of the liquidity crisis, the Federal Reservesaw very little demand for primary credit through its Discount Window—even afterlowering the discount rate from 100 basis points to 50 basis points above the federalfunds target.2 Some observers attributed the lack of Discount Window lending duringthis period to the notion of there being a “stigma” to such borrowing insofar as itwould send an adverse signal about the financial viability of the borrower.3 However,

1. Freixas et al. (1999) define the role of the lender of last resort to be “the discretionary provision ofliquidity to a financial institution (or the market as a whole) by the central bank in reaction to an adverseshock that causes an abnormal increase in the demand for liquidity which cannot be met by an alternativesource.” While history provides some examples of the lenders of last resort being private entities (e.g.,clearing houses in the United States prior to the establishment of the Federal Reserve) or even privateindividuals (J.P. Morgan in 1907), we consider the lender of last resort to be either part of the governmentor operating with explicit or implicit governmental backing.

2. The Discount Window is historically the principal mechanism through which the Federal Reserveperforms its lender of last resort function. The Discount Window is considered to be a “Lombard Facility”—meaning that eligible depository institutions can freely access central bank credit at a penalty rate withappropriate collateral. The Discount Window began operating this way in 2003. See Madigan and Nelson(2002) for an overview.

3. Prior to 2003, Discount Window loans were offered at below market rates, necessitating intensesupervisory scrutiny and moral suasion to limit the market’s use of such loans (Furfine 2003). According

ADAM ASHCRAFT, MORTEN L. BECH, AND W. SCOTT FRAME : 553

the lack of borrowing from the Discount Window can also be explained by the presenceof an alternative, lower cost government-sponsored liquidity backstop: the FederalHome Loan Bank System (FHLB System).

The FHLB System is a large, complex, and understudied U.S. government-sponsored enterprise (GSE) that was created in the midst of the Great Depression.This housing GSE consists of 12 cooperatively owned wholesale banks that act asa general source of liquidity to over 8,000 member financial institutions, which arecommercial banks, thrifts, credit unions, and insurance companies. This liquidity isprimarily provided through “advances” or (over) collateralized lending to members.FHLB advances are available in various maturities, can carry either fixed or variableinterest rates, and may include embedded options.

During the second half of 2007, the FHLB System increased its advance lendingby $235 billion to $875 billion by the end of that year (a 36.7% increase). Ten FHLBmembers alone accounted for almost $150 billion of this new advance lending. FHLBadvance activity is largely funded by debt that investors perceive to be implicitlyguaranteed by the U.S. government due to the FHLB System’s GSE status. FHLBdebt securities are considered Federal Agency securities, like those issued by FannieMae and Freddie Mac. At the outset of the financial crisis, money market investors ranaway from debt issued or sponsored by depository institutions and into instrumentsguaranteed explicitly or implicitly by the U.S. Treasury. As a result, the FHLB Systemwas able to reintermediate term funding to member depository institutions throughadvances.4

However, it became clear in December 2007—and again in March 2008—that theresponse of the FHLB System was not enough to ease all of the liquidity stress in termfunding markets. Institutions ineligible for FHLB membership, such as foreign banksand primary dealers, continued to have significant demands for dollar funding andwere not able to borrow from the Federal Reserve. In response, the Federal Reserveintroduced several new liquidity facilities aimed at reducing funding pressures, suchas: the Term Auction Facility, liquidity swaps with the European Central Bank andthe Swiss National Bank, the Term Security Lending Facility, and the Primary DealerCredit Facility.

During the recent financial crisis, the liquidity facilities of the Federal Reserveand the FHLB System have at the same time competed with and complemented eachother. The FHLB System took the early lead; it was not until May 2008 that theFederal Reserve became the largest government-sponsored liquidity facility in termsof crisis-related lending to the financial system. Indeed, FHLB lending grew more

to Peristiani (1998), market participants refrained from requesting Discount Window loans because of theperception that it would send a negative signal to the Federal Reserve, bank supervisors, and eventuallythe market at large. Furfine (2003) presents evidence that despite the policy change that allowed marketparticipants to access central bank credit with no questions asked, the perception of stigma remains.

4. Interestingly, the reintermediation of credit through the FHLBs during the fall of 2007 was quitedifferent from what occurred during the last major global liquidity event: the Asian financial crisis. Duringthe fall of 1998, money market investors ran from short-term paper issued by the corporate sector anddeposited their funds with the banking system. Banks, in turn, relent those funds to corporations throughbackup lines of credit (e.g., Gatev, Schuermann, and Strahan 2005).


modestly in 2008: from $875 billion to $1,012 billion at the end of the third quarterbefore falling to $929 billion at year-end. FHLB advances outstanding continued torecede during the first half of 2009. Hence, we view the FHLB System as the lenderof next-to-last resort.

The objective of our paper is threefold. First, we seek to document and understandthe role played by the FHLB System in the ongoing financial crisis. To this end, weprovide a brief overview of this larger sibling to the more well-known housing GSEs:Freddie Mac and Fannie Mae. We then document FHLB advance activity during thesecond half of 2007 and analyze how these funds were used by commercial banksand thrifts.

Second, we want to understand the interplay between the liquidity facilities pro-vided by the FHLB System and the Federal Reserve, respectively. We do so bycomparing quantities and prices. As a general reluctance to lend among privateagents emerged at the outset of the crisis, investors placed a premium on FHLBdebt due to its implicit government backing. This demand kept FHLB funding costsand advance rates down—making the FHLB System an especially attractive source ofwholesale funding. The FHLB System continued to see strong demand for advancesthrough the end of 2007 despite substantial cuts in the Federal Reserve’s discountrate relative to the federal funds target. The Federal Reserve eventually devised newprograms to meet the liquidity demands of financial institutions without FHLB ac-cess. However, following heightened concerns about the financial health of FannieMae and Freddie Mac during the summer of 2008, the FHLB System found itself“guilty by association” and saw its borrowing costs (and hence advance rates) riseand its own availability of term funding limited. As a result, the Discount Windowbecame the more attractively priced liquidity facility and saw a significant increase inborrowings.

Finally, we wish to draw insights and lessons from this episode in order to framea discussion for how to think about the lender of last resort role in a modernizedfinancial regulatory structure. During the first several months of the financial crisis,the FHLB System provided more U.S. dollar liquidity to the financial system than theFederal Reserve. And even after the Federal Reserve surpassed the FHLB System interms of total liquidity provided, the FHLB System continued to be the largest lenderto U.S. depository institutions, as much of the Federal Reserve’s liquidity operationsbenefited nondepository or foreign financial institutions. However, developments inthe Federal Agency debt market during the summer and fall of 2008 demonstratedthe limitations of relying on a lender of last resort with only implicit governmentguarantees. Following the Lehman Brothers bankruptcy, the Federal Reserve signif-icantly expanded its balance sheet in order to meet the explosive demand for U.S.dollar liquidity.

The organization of the paper closely follows these objectives. We begin withan overview of the FHLB System, continue with an analysis of the uses of FHLBadvances by depository institutions at the outset of the financial crisis, and then providea detailed comparison of the liquidity facilities of the FHLB System and the FederalReserve.


TABLE 1

FEDERAL HOME LOAN BANK SIZE AND MEMBERSHIP BY DISTRICT AS OF DECEMBER 31, 2008

Membership concentration

(five largest members)Total assets

FHLB $ billion Number of members Share of capital Share of advances

Atlanta $208.6 1,238 43.9% 51.3%Boston $80.4 461 51.5% 53.6%Chicago $92.1 816 29.9% 39.9%Cincinnati $98.2 728 51.4% 59.6%Dallas $78.9 923 51.1% 59.2%Des Moines $68.1 1,245 40.1% 40.1%Indianapolis $56.9 424 45.4% 48.4%New York $137.5 311 47.8% 50.4%Pittsburgh $90.8 323 57.8% 63.1%San Francisco $321.2 430 72.2% 77.8%Seattle $58.4 381 58.4% 68.4%Topeka $58.6 872 39.2% 49.9%

SOURCE: Federal Home Loan Banks 2008 Combined Financial Report, authors’ calculations.

1. THE FHLB SYSTEM

The FHLB System is composed of 12 regional FHLBs and an Office of Financethat acts as the FHLBs’ gateway to the capital markets. Each FHLB is a separate legalentity and has its own management, employees, board of directors, and financial state-ments. FHLBs are cooperatively owned by their member commercial banks, thrifts,credit unions, and insurance companies headquartered within the distinct geographicarea that the FHLB has been assigned to serve. Members must either maintain at least10% of their asset portfolios in mortgage-related assets or be designated as “com-munity financial institutions.”5 The FHLB System was originally created in 1932 toprimarily serve the thrift (or savings and loan) industry, which at that time did not haveaccess to the Federal Reserve’s Discount Window.6,7 In 1989, following the savingsand loan crisis, FHLB membership was expanded to include commercial banks andcredit unions. As of year-end 2008, the FHLB System had 8,152 financial institutionmembers—86% of which were commercial banks or thrifts.

Table 1 presents the relative sizes (in terms of total assets) and numbers of membersfor each of the 12 FHLBs as of December 31, 2008. The FHLB of San Francisco isby far the largest institution ($321.2 billion), accounting for almost a quarter of the

5. “Community financial institutions” are defined at 12 U.S.C. § 1422(13).

6. In the presidential statement about the signing of the FHLB Act in 1932, Herbert Hoover noted:“Its purpose is to establish a series of discount banks for home mortgages, performing a function forhomeowners somewhat similar to that performed in the commercial field by the Federal Reserve banksthrough their discount facilities.” See http://www.presidency.ucsb.edu/ws/?pid=23176.

7. The Depository Institutions Deregulation and Monetary Control Act of 1980 opened the DiscountWindow to all banks, savings and loan associations, savings banks, and credit unions holding transactionsaccounts and nonpersonal time deposits.


TABLE 2

FEDERAL HOME LOAN BANK SYSTEM COMBINED BALANCE SHEET AS OF DECEMBER 31, 2008

Amount ($ billions) Share of assets (%)

AssetsAdvances $928.6 68.8Cash and investments 326.7 24.2Mortgage loans (net) $87.4 6.5Other assets $6.4 0.5

Total assets $1,349.1 100.0

Liabilities and capitalConsolidated obligations (net) $1,258.3 93.3Other liabilities $39.4 2.9Membership capital stock $49.6 3.7Retained earnings $2.9 0.2Other comprehensive income ($1.1) (0.1)

Total liabilities and capital $1,349.1 100.0

SOURCE: Federal Home Loan Banks 2008 Combined Financial Report, authors’ calculations.

FHLB System’s assets. The FHLBs of Des Moines and Atlanta each have about 15%of the total FHLB System membership. The table also shows the extent to which eachbank’s business is dominated by its largest members. The percentage of each FHLB’sadvances (loans to members) that is accounted for by its five largest users range from39.9% (the FHLB of Chicago) to 77.8% (the FHLB of San Francisco).8

The FHLB System is often viewed as a whole because virtually all FHLB financingtakes the form of consolidated obligations for which the 12 institutions are jointlyand severally liable. Hence, Table 2 provides the consolidated balance sheet of the 12FHLBs as of December 31, 2008. Advances constitute 68.8% of the FHLB System’s$1,349.1 billion in total assets, cash and investments another 24.2%, and holdingsof residential mortgages are 6.5% of total assets. On the liability side of the balancesheet, consolidated obligations constitute 93.3% of total assets. The FHLB System’scapital is only 3.8% of assets, and almost all of that is the members’ contributedcapital; retained earnings are only 0.2% of assets. The FHLB System is thus a verylarge and highly leveraged financial institution.

The FHLB System is considered a GSE since it has been expressly created by anact of Congress (The Federal Home Loan Bank Act of 1932) that includes severalinstitutional benefits designed to reduce their operating costs. In this way, the FHLBSystem is similar to the other two housing GSEs—Fannie Mae and Freddie Mac.Certain charter provisions combined with past government actions, have created aperception in financial markets that GSE obligations are implicitly guaranteed by thefederal government.9 This, in turn, allows these institutions to finance their activities

8. Similarly, the percentage of each bank’s capital that is accounted for by its five largest membersranges from 30.0% (the FHLB of Chicago) to 74.9% (the FHLB of San Francisco).

9. Special privileges accruing to the FHLB System include: a provision authorizing the Treasurysecretary to purchase up to $4 billion of FHLB securities, the treatment of FHLB securities as “governmentsecurities” under the Securities and Exchange Act of 1934, the statutory ability to use the Federal Reserve as


by issuing debt on more favorable terms than any AAA-rated private corporation.10

Housing GSEs also accrue cost savings through an exemption from federal corporateincome taxes and an exemption from Securities and Exchange Commission registra-tion requirements for their debt securities. Key differences between the FHLB Systemand Fannie Mae/Freddie Mac relate to their primary functions (collateralized lendingvia advances versus issuing credit guarantees on mortgage-backed securities) andownership structure (cooperative versus publicly held corporations). The $1.3 trillionin total assets controlled by the FHLB System as of December 31, 2008 exceededthose for either Fannie Mae or Freddie Mac at that time ($912 billion and $851 billion,respectively).11

It is understood that explicit or implicit government guarantees of financial insti-tution liabilities will distort the risk-taking incentives of the insured institutions ina way that increases the probability of financial distress.12 Recognizing this poten-tial moral hazard, the federal government regulates the FHLB System for “safetyand soundness” through the Federal Housing Finance Agency (FHFA), which alsohas responsibility for Fannie Mae and Freddie Mac. The FHFA is an independentagency within the executive branch that was created in July 2008 with the passageof the Housing and Economic Recovery Act of 2008. Previously, the Federal Hous-ing Finance Board had sole responsibility for supervising the FHLB System. Likeother financial regulators, the FHFA is authorized to set capital standards, conductexaminations, and take certain enforcement actions if unsafe or unsound practices areidentified.13

The stated public purpose of the FHLB System is to provide their members withfinancial products and services, most notably advances, to assist and enhance themembers’ financing of housing and community lending.14 One important empiricalquestion relates to what types of assets FHLB advances ultimately fund on memberbalance sheets. While members must post collateral to secure their advances andthat collateral is typically residential mortgage related (whole loans or mortgage-backed securities), money is fungible; there is no reason why the members wouldnecessarily use the borrowed funds for further housing loans or other designateduses. Indeed, empirical evidence provided by Frame, Hancock, and Passmore (2007)suggests that FHLB advances are just as likely to fund other types of bank credit asto fund residential mortgages.

its fiscal agent (like the Treasury), and an exemption from the bankruptcy code by way of being considered“federal instrumentalities.”

10. For estimates of this debt funding advantage, see Ambrose and Warga (1996, 2002), Nothaft, Pearce,and Stevanovic (2002), and Passmore, Sherlund, and Burgess (2005).

11. Fannie Mae and Freddie Mac also maintain large volumes of credit guarantees on mortgage-backedsecurities. These guarantees (net of securities held on their own balance sheets) totaled $2.3 trillion (FannieMae) and $1.8 trillion (Freddie Mac) as of December 31, 2008.

12. Flannery and Frame (2006) identify and analyze FHLB System risk-taking incentives and comparethem to those faced by Fannie Mae and Freddie Mac.

13. The regulations currently applying to the FHLB System are those previously promulgated by theFederal Housing Finance Board. These regulations are codified at 12 C.F.R. § 900–999.

14. See 12 U.S.C. § 1430(a)(2).


Another important question relates to whether the benefits of FHLB membershipflow to members and, if so, whether it flows further still to consumers—especiallymortgage borrowers. In one study, the U.S. Congressional Budget Office (2004)estimated that the FHLB System accrued $3.4 billion in implicit federal supportin 2003 and that $0.2 billion of that accrued to conforming mortgage borrowers whilethe remainder was captured by various FHLB stakeholders. Presumably, most ofthese benefits accrue to the FHLBs member-owners who, in turn, pass them on totheir customers. However, some benefits may be captured by FHLB management andshareholders. A more comprehensive analysis of the distribution of FHLB benefitswould be a welcome addition to the literature.

2. THE ROLE OF FHLB ADVANCES AT THE OUTSETOF THE LIQUIDITY CRISIS

Advances are historically the primary activity conducted by the FHLBs. Advancesare available in various maturities, carry fixed or variable rates of interest, sometimescontain embedded options, and are fully collateralized. In terms of maturities, as ofyear-end 2008, about 43% of advances were due in less than 1 year, about 41% weredue in 1–5 years, and about 16% due thereafter. At the same time, put and call optionsthat can alter the duration and yield of an advance were included in about 21% ofthe FHLBs’ combined advance book. The most common forms of advance collateralare residential mortgage-related assets (whole loans and mortgage-backed securities)and U.S. Treasury and Federal Agency securities.15

Beyond the explicit collateral and a member’s capital subscription, the FHLBs alsohave priority over the claims of depositors and almost all other creditors (including theFederal Deposit Insurance Corporation, or FDIC) in the event of a member’s default;this is often described as a “super-lien.”16 Taken together, these features help to explainwhy none of the FHLBs has ever suffered a loss on an advance. Unfortunately, froma public policy perspective, the combination of overcollateralization and the super-lien can create an incentive for the FHLBs to provide their members with morecredit than is socially optimal. This is due to the fact that these provisions reduce theFHLBs’ incentives to screen and monitor their members and the pledged collateral.This arrangement also serves to weaken the claims of existing private creditors andexpose the FDIC to increased losses in the event of failure (Stojanovic, Vaughan, andYeager 2008). Furthermore, the 12 FHLBs are not subject to regulatory “loan to oneborrower limits.”17

15. See 12 U.S.C. § 1430(a)(3) for a complete list of eligible collateral. Federal Agency securities aregenerally synonymous with debt and MBS issued by government-sponsored enterprises.

16. See 12 U.S.C. § 1430(e).

17. Some individual FHLBs have internal limits that are set in the range of 30%–50% of member totalassets. By contrast, national banks limit loans to one borrower at 25% of bank total equity—with not morethan 15% of bank equity being unsecured (12 U.S.C. § 84(a)).


600

650

700

750

800

850

900

950

1,000

1,050

Jul-08Jan-08Jul-07Jan-07Jul-06Jan-06

$ B

illi

on

s

Published Estimated

FIG. 1. Federal Home Loan Bank Advances Outstanding: Monthly Observations, January 2006 to December 2008.

SOURCE: Federal Home Loan Banks’ Office of Finance.

NOTE: Estimated numbers are based on linear interpolation prior to September 2007. After September 2007, interpolation

is based on the pattern of FHLBs’ net issuance of consolidated obligations.

FHLB advances are generally viewed as an attractive source of wholesale funds.Advance interest rates are set by the individual FHLBs and reflect a mark-up to thecost of Federal Agency debt funding secured by the Office of Finance. However, inorder to receive an advance, a member must also purchase FHLB stock in an amountranging from 2% to 6% of the advance (as dictated by the individual FHLB’s capitalplan). While FHLB stock typically pays a dividend, to the extent this pay-out fallsbelow the members’ marginal investment opportunity the stock purchase requirementcan create an opportunity cost. Generally speaking, there is a positive relationshipbetween the interest rates charged on advances and dividend rates across FHLBs,with differences presumably reflecting efficiencies and the collective preferences ofthe membership.

Advances grew rapidly during the 1990s and early 2000s following the introduc-tion of commercial banks as FHLB System members. However, from the end of 2005through the first half of 2007, the level of outstanding FHLB advances oscillated withina narrow range of $620 to $640 billion (see Figure 1). The amount of outstanding


TABLE 3

LARGEST DOLLAR INCREASES IN ADVANCES BY FHLB MEMBERS: 2007:Q2 TO 2007:Q4

Advances/assetsChange in advances Advances Assets

Q2–Q4 2007 Q4 2007 Q4 2007 Q2 2007 Q4 2007

$ Billions PercentWashington Mutual 42.4 63.9 325.8 6.9 19.6Bank of America 25.5 57.2 1,667.4 2.2 3.4Countrywide, FSB 18.9 47.7 121.1 28.9 39.4Merrill Lynch 11.7 11.7 115.8 – 10.1Wachovia Corp. 11.3 41.9 740.8 4.5 5.7Wells Fargo & Co 11.1 11.3 529.8 0.0 2.1Citigroup 8.3 102.0 1,351.3 7.6 7.6PNC 7.3 8.7 131.2 1.2 6.6Capital One 5.9 6.8 132.3 0.7 5.2US Bank 5.7 17.2 238.4 5.1 7.2

Total/average 148.1 368.3 5,354.0 4.6 6.9

NOTE: Merrill Lynch = Merrill Lynch Bank USA and Merrill Lynch Bank & Trust Co., FSB.

SOURCE: Call and Thrift reports.

advances ticked up slightly in July 2007, but then exploded during August and Septem-ber of that year—moving from $659 to $824 billion (a 25% increase). FHLB advancesgrew to $875 billion by the end of 2007 and then peaked at $1,012 billion as of 2008:Q3before falling to $929 billion by the end of 2008. FHLB advances outstanding sub-sequently slid back to $739 billion by mid-year 2009.

During the second half of 2007, the 10 most active FHLB members accounted foralmost $150 billion of the $235 billion increase (63%). Table 3 shows that WashingtonMutual, Bank of America, and Countrywide borrowed the largest amounts from theFHLB System during this period; for Washington Mutual and Countrywide, theirratios of advances-to-total assets rose to 20% and 40%, respectively. Notably, 4 of the10 institutions listed have subsequently failed or been acquired, while 2 others haverequired “exceptional assistance” from the U.S. government.

As liquidity pressures developed during the fall of 2007, FHLB advances becamean attractive source of funding in terms of pricing. During this time, investors soughtthe protection of (explicitly or implicitly) federally guaranteed obligations and FHLBfunding costs moved well below other benchmarks, like LIBOR and AA-rated ABCP(see Figure 2). For example, the average spread between 1-month LIBOR and 4-weekFHLB discount notes increased from about 16 basis points prior to the turmoil to 44basis points in the subsequent 12 months. During the same time, the average spreadbetween a 30-day advance from the FHLB New York and 4-week FHLB Systemdiscount notes has remained unchanged at about 25 basis points (see Figure 2).

Much of the growth in FHLB advances in the second half of 2007 reflected longer-term lending, which the GSE financed primarily by issuing short-term liabilities. Ofthe $235 billion increase in FHLB advances during the second half of 2007, $205billion carried an original maturity of greater than 1 year (87.4%). By contrast, overthe same period, discount notes with maturities of less than 1 year increased by


0

50

100

150

200

250

300

350

400

450

Oct-08Jul-08Apr-08Jan-08Oct-07Jul-07Apr-07Jan-07

Bas

is P

oin

ts

LIBOR 1M ABCP AA 30 Day FHLB NY Advance 30 Day

FIG. 2. Spread of Selected Funding Rates to 4-Week FHLB Discount Note Daily Observations, January 2007 to

December 2008.

SOURCE: Office of Finance, Bloomberg.

$213 billion, comprising 94.2% of net new FHLB consolidated obligations.18 Webelieve that this mismatch reflected the market stress during this period. For example,anecdotal evidence suggests that many depository institutions sought term fundingfor loans originally intended to be securitized but that were unable to be moved offbalance sheet. On the other hand, investors shunned ABCP issuers and instead soughtthe safety of short-term U.S. Treasury and Federal Agency debt securities.

In order to better understand why financial institutions markedly increased theirborrowing from the FHLBs during the second half of 2007, we take two approaches.First, we analyze aggregate growth within the balance sheets of banks and thrifts bycomparing this period to the trend in the six quarters preceding the crisis. Second,we take a statistical approach, documenting how the relationship between changes inFHLB advances and changes in other balance sheet items varied for banks and thriftsduring the last two quarters of 2007 relative to the benchmark of the previous sixquarters.

18. Discount notes are generally sold in sizes ranging from $500 million to over $5 billion each, withtypical maturities being overnight, 4, 9, 13, and 26 weeks.


2.1 Aggregate Balance Sheets

We start by aggregating the Call Reports of both commercial banks and thriftsover three time periods: the six quarters before the recent financial crisis, 2006:Q1–2007:Q2 (our benchmark), and each of the two quarters following the onset of thecrisis 2007:Q3 and 2007:Q4.19 In order to capture differences across institutions ofdifferent sizes, these aggregates are broken into two groups using a threshold of $5billion in total assets measured at the top holder level.20 “Large institutions,” or thosewith greater than $5 billion in total assets, accounted for 80% of FHLB advancesoutstanding as of December 31, 2007.

Table 4 documents the aggregate behavior of large and small depository institu-tions over the three time periods (Panels A–C). For each line item, the table reportsthe aggregate percentage of the item relative to total assets in the last quarter, thepercentage change over the quarter, and the change in the ratio of the item to totalassets measured in percentage points.

We begin our discussion focusing on the behavior of large institutions duringthe third quarter of 2007 (Panel B). Most striking is the 31.7% increase in FHLBadvances compared to the average quarterly growth rate in this balance sheet item of0.4% over the previous six quarters reported in Panel A. The overall increase in “otherborrowing,” of which FHLB advances are a part, more than offset a decline in federalfunds and repo borrowing by large institutions. This suggests that FHLB advanceswere used, in part, to mitigate a funding shock. While deposit growth was slow (2.1%)relative to growth in total assets (4.0%), it was slightly higher than the average depositgrowth over the previous six quarters (1.8%). This suggests that funding pressuresfaced by large institutions were largely isolated to federal funds and repo borrowing.

On the asset side, large institutions also reduced their cash holdings (relative tototal assets) during the third quarter of 2007—consistent with funding stress. Assetgrowth during 2007:Q3 for these institutions largely came from federal funds andrepo lending as well as trading assets. Large institutions also experienced an increasein total loans (3.4%) above the baseline average quarterly growth rate of 1.5%. Thisacceleration largely came from nonmortgage loans, likely the result of draws onoutstanding lines of credit. The increase in trading assets is consistent with largeinstitutions using FHLB advances in order to fund mortgage loans in the securitizationpipeline that were unexpectedly retained on the balance sheet due to the breakdownof the originate-to-distribute model.21

One possible explanation for the increase in federal funds and repo lending during2007:Q3 is that a number of institutions were granted exemptions from Section 23Aof the Federal Reserve Act, which restricts lending to affiliates, in order to allow

19. The bank and thrift Call Reports do have some minor differences and we have worked to keepcategories comparable and thereby minimize distortions.

20. Measurement at the top holder level means that institutions with individual charters, but commonownership (affiliates), are combined.

21. The Call Reports of Income and Condition did not break out mortgage loans from other items inthe trading account until March 2008, making it difficult to document this hypothesis.


TABLE 4

AGGREGATE BALANCE SHEET ITEMS FOR BANKS AND THRIFTS (TOP HOLDER LEVEL; QUARTERLY DATA FOR 2006AND 2007)

Large Small

Change Change

Percentage of Percent in ratio Percentage Percent in ratio

assets growth to assets of assets growth to assets

Panel A. 2006:Q1 to 2007:Q2 (Average quarterly growth rate and change)

AssetsCash 4.0% 4.0% 0.1% 3.1% 0.5% −0.0%Securities 15.1% 0.1% −0.2% 17.8% −0.5% −0.4%Fed funds and repos 5.1% 1.8% 0.0% 1.9% 1.8% 0.0%Trading assets 7.2% 5.9% 0.3% 0.3% 55.8% 0.0%Total loans 58.8% 1.5% −0.1% 71.3% 2.5% 0.4%Mortgage loans 35.2% 1.3% −0.1% 56.2% 2.5% 0.3%Nonmortgage loans 24.2% 1.6% −0.0% 15.9% 2.3% 0.1%Total assets 1.7% 1.9%

LiabilitiesDeposits 62.0% 1.8% 0.1% 78.3% 1.9% 0.0%Fed funds and repos 7.9% 1.5% −0.0% 2.8% 3.1% 0.0%Trading liabilities 3.1% 1.6% −0.0% 0.0% 38.5% 0.0%Other borrowing 11.4% 1.2% −0.1% 7.8% 0.5% −0.1%Of which: FHLB advances 4.9% 0.4% −0.1% 7.5% 0.2% −0.1%

Panel B. 2007:Q3 (Actual quarterly growth rate and change)

AssetsCash 3.7% −2.4% −0.2% 2.8% −6.1% −0.2%Securities 14.2% 0.4% −0.5% 17.6% 2.4% 0.1%Fed funds and repos 5.8% 10.2% 0.3% 1.6% −7.0% −0.2%Trading assets 7.7% 10.2% 0.4% 0.1% −10.8% −0.0%Total loans 58.9% 3.4% −0.3% 71.8% 1.9% 0.2%Mortgage loans 34.8% 1.7% −0.8% 56.8% 2.0% 0.2%Nonmortgage loans 24.8% 6.1% 0.5% 15.9% 1.6% 0.0%Total assets 4.0% 1.7%

LiabilitiesDeposits 60.8% 2.1% −1.2% 77.4% 0.5% −0.9%Fed funds and repos 7.3% −4.6% −0.7% 2.7% 5.0% 0.1%Trading liabilities 3.4% 13.7% 0.3% 0.0% 11.6% 0.0%Other borrowing 13.1% 18.9% 1.6% 8.5% 9.2% 0.6%Of which: FHLB advances 6.3% 31.7% 1.3% 8.3% 9.1% 0.6%

Panel C. 2007:Q4 (Actual quarterly growth rate and change)

AssetsCash 3.8% 6.5% 0.1% 3.1% 11.7% 0.3%Securities 13.4% −2.9% −0.8% 17.4% 0.4% −0.3%Fed funds and repos 5.9% 12.4% 0.5% 1.6% −2.3% −0.1%Trading assets 8.3% 9.9% 0.6% 0.1% −11.0% −0.0%Total loans 58.8% 2.3% −0.2% 71.8% 2.3% 0.1%Mortgage loans 34.3% 1.4% −0.4% 56.8% 2.4% 0.1%Nonmortgage loans 25.2% 4.1% 0.4% 15.8% 2.1% −0.0%Total assets 2.6% 2.2%

LiabilitiesDeposits 61.1% 3.1% 0.3% 76.8% 1.3% −0.7%Fed funds and repos 7.2% 1.6% −0.1% 3.0% 11.8% 0.3%Trading liabilities 3.3% 2.0% −0.0% 0.0% 6.5% 0.0%Other borrowing 13.2% 3.1% 0.1% 9.0% 8.9% 0.6%Of which: FHLB advances 6.4% 4.0% 0.1% 8.8% 9.2% 0.6%


commercial banks to support their affiliated broker-dealers.22 We investigated thisexplanation by comparing the increase in repo lending on the bank and thrift CallReports with similar lending on the holding company’s Y-9C, for the subset of U.S.institutions where such information is available. As lending from a bank to its affiliatewould not appear on the consolidated balance sheet, this should provide indirectevidence on the importance of changes in interbank lending. The evidence suggeststhat this phenomenon only explained a small part of the increase in federal funds andrepo lending. Hence, it appears that large institutions were using FHLB advances tohelp fund assets more generally. Overall, the FHLBs appear to have been performingas a typical lender of last resort, providing liquidity to depository institutions that, inturn, provided liquidity more broadly to the rest of the economy.

The data in Panel B also document a significant increase in FHLB advances (9.1%)during 2007:Q3 for small financial institutions, or those with less than $5 billionin total assets. This increase appears to largely have been to offset slow depositgrowth (relative to the baseline period). The growth in the assets of small institutions(1.7%) was slightly below average over the previous six quarters. One interpretationof this fact is that funding pressure was constraining the balance sheets of smallinstitutions, but another is that they reduced their demand for funding as they tightenedunderwriting standards. The decline in cash and the acceleration in the growth offederal funds and repo borrowing would appear to support the former explanation.Moreover, it is interesting to note the significant decline in federal funds and repolending, suggesting that small institutions were part of the investor class exitingshort-term funding markets. As reducing the level of interbank lending is cheaperthan increasing the level of interbank borrowing, this is also consistent with smallinstitutions facing funding pressures.

Panel C documents the aggregate behavior of banks and thrifts during the fourthquarter of 2007. Growth in FHLB borrowing by large institutions was only modestlyfaster than that of total assets (4.0% versus 2.6%). This asset growth is largely ex-plained by the same sources from the third quarter: federal funds and repo lending,trading assets, and nonmortgage loans. Small institutions appeared to be under con-tinued pressure in the fourth quarter: deposit growth was slow relative to precrisisaverages, federal funds and repo borrowing and FHLB advances expanded quickly,and federal funds and repo lending continued to contract.

Overall, the aggregate data suggest that both large and small institutions usedFHLB advances during the second half of 2007 in order to smooth a liquidity shock.However, large institutions also used advances to fund increases in federal funds andrepo lending, the trading book, and nonmortgage lending.

22. Exemptions were granted on August 20, 2007 for Citigroup, Bank of America, and J.P. MorganChase. Later in the third quarter of 2007, similar exemptions were granted for the New York branches ofDeutsche Bank AG, Royal Bank of Scotland PLC, and Barclays Bank PLC. These exemptions were an-nounced on the public website of the Board of Governors of the Federal Reserve (www.federalreserve.gov).


2.2 Regression Analysis

So far, our approach has been descriptive, but now we turn to some statisticalanalysis. In particular, we examine the relationship between changes in FHLB ad-vances at the bank and thrift top holder level and changes in their other balancesheet categories during the second half of 2007. We estimate cross-sectional OLSregressions of the quarterly change in FHLB advances on similar changes in cashholdings, federal funds and repo lending, trading assets, funding (the sum of totaldeposits, federal funds borrowing, and repo borrowing), mortgage loans, and non-mortgage loans using a two-step procedure that eliminates the influence of outliers oncoefficient estimates.23 Each variable is scaled by the previous quarter’s total assets.Eight quarters of data are pooled (2006:Q1 to 2007:Q4) and indicator variables forthe last two quarters are included and interacted with the changes in balance sheetcategories to statistically examine changes at the outset of the crisis. Finally, giventhe previously documented heterogeneity, separate regressions are estimated for smalland large institutions.

The results in Table 5 suggest that during normal times, changes in FHLB advancesare positively related to a variety of asset categories (cash, federal funds and repolending, mortgage loans, and nonmortgage loans) and negatively related to funding.This is consistent with depository institutions using FHLB advances as a generalsource of liquidity (Frame, Hancock, and Passmore 2007). Comparing across sizecategories, it appears that changes in FHLB advances in the six quarters before theliquidity crisis are more strongly associated with changes in federal funds and repolending, changes in mortgage lending, and changes in funding for small institutions.The latter fact is consistent with financial constraints at small institutions, whichbenefit more from access to FHLB advance funding. Interestingly, the results suggestan increase in the link between FHLB advances and the various asset categories atthe outset of the crisis for large institutions (relative to small institutions).

The first column of Table 5 presents results for small banks and thrifts. As indicatedby the interaction terms, during 2007:Q3 increases in FHLB advances are foundto be increasingly positively related to four of the asset categories (cash, federalfunds and repo lending, mortgage loans, and nonmortgage loans) and increasinglynegatively related to funding. This indicates a stronger relationship between thesevariables during 2007:Q3 than during the baseline period (2006:Q1 to 2007:Q2).These relationships tended to revert during 2007:Q4 toward historical norms—and insome cases overshooting these norms—suggesting a transitory shock for small banksand thrifts.

The second column of Table 5 displays results for large banks and thrifts. Relativeto the baseline, it appears that during 2007:Q3 FHLB advances were increasinglyused by large banks to significantly bolster cash reserves and federal funds and repolending. At the same time, advances were increasingly related to declines in other typesof funding. For large institutions, it appears that the relationships between changes in

23. Specifically, each regression is estimated by OLS and the associated DFITS statistic is used toidentify and eliminate influential observations (see Welsch and Kuh 1977).


TABLE 5

REGRESSION RESULTS ESTIMATING THE RELATIONSHIP BETWEEN CHANGES IN FHLB ADVANCES AND CHANGES IN

VARIOUS BALANCE SHEET ITEMS FOR COMMERCIAL BANK AND THRIFTS (TOP HOLDER LEVEL; QUARTERLY DATA

FOR 2006 AND 2007)

(1) (2)

Small institutions Large institutions

(Change in cash)/assets 0.436 0.323(0.007)∗∗∗ (0.077)∗∗∗

(Change in cash)/assets ∗ 2007Q2Q3 dummy 0.106 0.409(0.021)∗∗∗ (0.218)∗

(Change in cash)/assets ∗ 2007Q3Q4 dummy −0.069 −0.463(0.018)∗∗∗ (0.190)∗∗

(Change in fed funds and repo assets)/assets 0.431 0.152(0.005)∗∗∗ (0.039)∗∗∗

(Change in fed funds and repo assets)/assets ∗ 2007Q2Q3 dummy 0.057 0.470(0.014)∗∗∗ (0.125)∗∗∗

(Change in fed funds and repo assets)/assets ∗ 2007Q3Q4 dummy −0.050 0.175(0.013)∗∗∗ (0.113)

(Change in trading assets)/assets 0.020 0.046(0.019) (0.085)

(Change in trading assets)/assets ∗ 2007Q2Q3 dummy 0.215 0.337(0.261) (0.283)

(Change in trading assets)/assets ∗ 2007Q3Q4 dummy 0.191 −0.332(0.151) (0.275)

(Change in mortgage loans)/assets 0.554 0.466(0.005)∗∗∗ (0.030)∗∗∗

(Change in mortgage loans)/assets ∗ 2007Q2Q3 dummy 0.055 0.064(0.013)∗∗∗ (0.138)

(Change in mortgage loans)/assets ∗ 2007Q3Q4 dummy −0.040 −0.085(0.012)∗∗∗ (0.087)

(Change in non-mortgage loans)/assets 0.532 0.203(0.006)∗∗∗ (0.038)∗∗∗

(Change in non-mortgage loans)/assets ∗ 2007Q2Q3 dummy 0.062 0.281(0.018)∗∗∗ (0.183)

(Change in non-mortgage loans)/assets ∗ 2007Q3Q4 dummy −0.026 0.201(0.016) (0.105)∗

(Change in funding)/assets −0.485 −0.308(0.004)∗∗∗ (0.025)∗∗∗

(Change in funding)/assets ∗ 2007Q2Q3 dummy −0.052 −0.244(0.011)∗∗∗ (0.092)∗∗∗

(Change in funding)/assets ∗ 2007Q3Q4 dummy 0.034 −0.014(0.010)∗∗∗ (0.070)

2007Q2Q3 dummy 0.002 0.006(0.000)∗∗∗ (0.002)∗∗∗

2007Q3Q4 dummy 0.003 0.007(0.000)∗∗∗ (0.002)∗∗∗

Constant −0.002 −0.001(0.000)∗∗∗ (0.001)∗

Observations 28708 865R2 0.47 0.35

NOTE: Standard errors in parentheses.∗ Significant at 10%; ∗∗ significant at 5%; ∗∗∗ significant at 1%.

advances and the changes in the other balance sheet items reverted to precrisis levelsduring 2007:Q4. However, of note is the overshooting in the relationship betweenchanges in advances and changes in cash reserves during that quarter as well as aheightened relationship between advances and nonmortgage lending.


0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,000

Oct-08Jul-08Apr-08Jan-08Oct-07Jul-07

$ B

illi

ons

FHLB Net Advances Relative to June 30, 2007 Federal Reserve Lender of Last Resort Lending

Fed LoLR lending + TSLF

FIG. 3. Net Liquidity Provided by the Federal Reserve and Federal Home Loan Bank System: Monthly Observations,

July 2007 to December 2008.

SOURCE: Federal Reserve Board H.4.1. Release; Federal Reserve Bank of New York; and Federal Home Loan Banks’

Office of Finance.

3. CRISIS-RELATED LENDING BY THE FEDERAL RESERVEAND THE FHLB SYSTEM

During the ongoing financial crisis, the liquidity facilities of the Federal Reserveand the FHLB System seem to have both competed with and complemented each other.Below, we analyze prices and quantities in order to gauge the relative magnitude andimportance of the crisis-related lending from the two institutions.24 We focus on fivedistinct parts of the crisis: the initial shock in August 2007, the introduction of theTerm Auction Facility (TAF) and swap lines with foreign central banks in December2007, the introduction of Primary Dealer Credit Facility (PDCF) and Term SecuritiesLending Facility (TSLF) in March 2008, the heightened concerns about the financialhealth of Fannie Mae and Freddie Mac in July and September 2008, and the falloutfrom the Lehman Brothers bankruptcy in September and October 2008.

Figure 3 presents the month-end crisis-related lending by the Federal Reserve andthe FHLB System between July 2007 and December 2008. For the Federal Reserve,

24. Usage alone might be a misleading measure of the impact of a liquidity backstop facility as theoption of being able to use such a facility is valuable in its own right.


0

200

400

600

800

1,000

1,200

1,400

1,600

1,800


$ B

illi

on

s

Discount Window TAF Foreign CB Swaps

28 Day MBS OMOs PDCF Maiden Lanes + AIG

AMLF CPFF TALF

FIG. 4. Total Liquidity Outstanding Provided by the Federal Reserve Programs: Weekly Observations, July 2007 to

December 2008.

SOURCE: Federal Reserve Board H.4.1 release and the Federal Reserve Bank of New York.

NOTE: TAF = Term Auction Facility; Foreign CB Swaps = central bank liquidity swaps; PDCF = Primary Dealer Credit

Facility; Maiden Lanes = net portfolio holdings of Maiden Lane LLC, Maiden Lane II LLC, Maiden Lane III LLC;

AIG = credit extended to American International Group, Inc.; AMLF = Asset-Backed Commercial Paper Money Market

Mutual Fund Liquidity Facility; CPFF = Commercial Paper Funding Facility; TALF = Term Asset-Backed Securities

Loan Facility.

we present total lending through all of the various credit and liquidity facilities and abreakdown by facility is provided in Figure 4. During the first 4 months of the liquiditycrisis, the FHLB was clearly the dominate source of government-sponsored liquidity(see Figure 3). It was not until December 2007 that the Federal Reserve began to lendsignificant amounts as a result of the introduction of the TAF and currency swap lineswith foreign central banks. The figure also documents that the Federal Reserve didnot eclipse the FHLB System in terms of crisis-related lending until May 2008.25

3.1 August 2007: The Initial Shock

The Federal Reserve initially responded to the turmoil in the interbank marketsin August 2007 with the introduction of the Term Discount Window Program and a

25. If the Federal Reserve’s lending of Treasury securities under the TSLF is included, the FederalReserve surpassed the FHLB System in March 2008.


reduction in the price of primary credit through the Discount Window.26 Specifically,the term of primary credit was extended from overnight to as long as 30 days (laterextended to 90 days) and the spread of the primary credit interest rate over the federalfunds target rate was lowered from 100 to 50 basis points (and eventually to 25 basispoints). At the same time, the Federal Reserve also openly encouraged the use of theDiscount Window by identifying such use as a sign of strength during a speciallyconvened teleconference with a group of large banks and major investment bankingfirms (The Clearing House 2007). Despite this initial activity, Discount Windowborrowing was negligible during the second half of 2007. By contrast, the FHLBSystem saw a brisk business: in August and September alone, the FHLB Systemlent out an additional $165 billion, and by the end of 2007 the level of outstandingadvances was up $235 billion for the second half of that year.

One explanation for the lack of Discount Window borrowing is the perception bypotential borrowers that markets will view such borrowing very unfavorably (e.g.,Peristiani 1998). In other words, that there is a “stigma” associated with borrowingfrom the Discount Window.27 Figure 5 illustrates this point by documenting thefraction of days in each month where the intraday high in the Federal Funds market(as reported by the Federal Reserve Bank of New York) is above the primary creditinterest rate between January 2003 (when Discount Window policies were altered)and December 2008. The fact that institutions are willing to pay more in the interbankmarket than the interest rate at which they could borrow directly from the FederalReserve suggests there is some stigma associated with the Discount Window.28

While stigma is a compelling explanation of the data, the unwillingness of insti-tutions to borrow from the Federal Reserve at the outset of the crisis can also beexplained by the simple fact that FHLB advances were a less expensive option fordomestic depository institutions. The relative attractiveness of the Federal Reserve’sDiscount Window vis-a-vis the FHLB System is, for the most part, driven by thespread between the primary credit rate and the FHLB advance rate of comparablematurity. However, differences in the haircuts applied across types of collateral, stockpurchase requirements (and the associated dividends), and interest rate expectationsall influence the cost of borrowing. The FHLB System may also be attractive fordepository institutions seeking longer-term financing that is unavailable through theDiscount Window.

26. Primary credit is available to depository institutions in sound overall condition to meet short-term,backup funding needs at a price above the federal funds rate target. Normally, primary credit will be grantedon a “no-questions-asked,” minimally administered basis. There are no restrictions on borrowers’ use ofprimary credit.

27. As mentioned previously, Furfine (2003) documents a continued reluctance of banks to borrowfrom the Discount Window following the introduction of changes made to the facility in 2003 in order toreduce stigma.

28. As discussed below, the Federal Reserve’s experiences with the Term Auction Facility provideadditional evidence of stigma.


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Jul-08Jan-08Jul-07Jan-07Jul-06Jan-06Jul-05Jan-05Jul-04Jan-04Jul-03Jan-03

Fra

ctio

n of

Day

s in

Mon

th

FIG. 5. Fraction of Days within a Month Where the Federal Funds Intraday High Exceeded the Primary Credit Rate:

Monthly Observations, January 2003 to December 2008.

SOURCE: Federal Reserve Bank of New York.

Figure 6 provides the average weekly borrowing of primary credit from the DiscountWindow together with an estimated all-in cost spread between a 30-day DiscountWindow loan and a 30-day advance from the New York FHLB collateralized by aAAA-rated Federal Agency mortgage-backed security. The method of deriving therespective all-in cost measures is provided in the Appendix. According to this measure,the relative attractiveness of a 30-day advance from the New York FHLB averagedabout 80 basis points between January 2003 and August 2007.29 The attractiveness ofthe FHLB advance then fell to somewhere in the 20- to 40-basis-point range followingthe Federal Reserve’s 50-basis-point reduction in the spread of the primary credit rateover the federal funds rate target in August 2007.

The FHLB System clearly played an important role as a government-sponsoredsource of liquidity during the first months of the financial crisis. While the stigmaassociated with borrowing from the Federal Reserve’s Discount Window is one expla-nation for this fact, we document that FHLB members were also choosing to borrowfrom the cheapest provider of wholesale funds.

29. Prior to January 2003, the interest rate charged at the Discount Window was typically 25–50 basispoints below the federal funds rate. While the below-market rate for Discount Window credit createdincentives for an institution to borrow, regulation required institutions to first exhaust other availablesources of funds and explain their need for credit.


0

20

40

60

80

100

120

-120

-80

-40

0

40

80

120


$ B

illions

Bas

is P

oin

ts

Wednesday level / 5 business day moving average

Primary Credit (right axis) Spread in All-in-Cost of Discount Window and FHLB NY

FIG. 6. Discount Window Borrowings and the Estimated Spread between the All-in Costs between the Federal Reserve

Primary Credit and Comparable FHLB Advance: Weekly Observations, July 2007 to December 2008.

SOURCE: Federal Reserve Board H.4.1. release, the Federal Reserve Bank of New York and the Office of Finance.

3.2 December 2007: The TAF and Swap Lines with Foreign Central Banks

Despite the unprecedented increase in FHLB advances during the fall of 2007,funding markets remained stressed. Figure 7 illustrates this by plotting the dailyspread of 1-month LIBOR to the similar duration overnight indexed swap rate (OIS)between January 2007 and December 2008.30 This spread widened substantially inlate 2007 and became a widely followed indicator of financial market stress.

In order to better understand the limited effect of FHLB advances on the LIBOR-OIS interest rate spread, it is insightful to look at the panel of 16 banking organizationsthat are surveyed to measure U.S. term dollar LIBOR.31 Table 6 lists these bankingorganizations and indicates whether they have access to FHLB advances and/or the

30. An OIS is an interest rate swap with the floating rate tied to an index of daily overnight rates, suchas the effective federal funds rate. At maturity, the two parties exchange (on the basis of the agreed notionalamount) the difference between interest accrued at the fixed rate and that accrued through geometricaveraging of the index (floating) rate.

31. The LIBOR panel is surveyed each morning (11:00 a.m. local time) by the British Bankers Associa-tion about the rate at which it could borrow funds were it to do so by asking for and then accepting interbankoffers in reasonable market size just prior to 11:00 a.m. Contributor panels comprise at least eight banks,and are intended to broadly reflect the balance of activity in the interbank deposit market. Contributed ratesare ranked in order and only the middle two quartiles averaged arithmetically. This average rate becomesthe “LIBOR fixing” for that particular currency, maturity, and date.


0

50

100

150

200

250

300

350

400


Basi

s P

oin

ts

FIG. 7. One-Month LIBOR–Overnight Index Swaps (OIS) Spread: Daily Observations, January 2007 to December

2008.

SOURCE: Bloomberg.

TABLE 6

LIBOR PANEL BANK ACCESS TO FHLB ADVANCES AND THE FEDERAL RESERVE DISCOUNT WINDOW

LIBOR panel member FHLB member? Discount window?

Bank of America Yes YesBank of Tokyo – Mitsubishi UFJ Union Bank of CA Union Bank of CABarclays Bank Plc No YesCitibank NA Yes YesCredit Suisse No YesDeutsche Bank AG No YesHBOS No YesHSBC HSBC Bank USA, NA HSBC Bank USA, NAJ.P. Morgan Chase Yes YesLloyds TSB Bank Plc No YesRabobank Rabobank, NA YesRoyal Bank of Canada RBC Centura YesThe Norinchukin Bank No YesThe Royal Bank of Scotland Group Citizens Bank of Pennsylvania Yes

and RBS Citizens, NAUBS AG No YesWest LB AG No Yes

SOURCES: British Banker’s Association, Federal Home Loan Bank System’s Office of Finance, bank call reports, and thrift call reports. FHLBmembership measured over the period July 2007 through March 2008.

Federal Reserve’s Discount Window. In the table, a banking organization is definedto have “direct access” to the FHLB System if it controls a U.S. depository institutionthat is a FHLB member and that member is large relative to the banking organization.A banking organization is defined to have “indirect access” if it controls a U.S.


depository institution that is a FHLB member, but that member is small relative tothe banking organization.32 Very few of the banking organizations participating inthe LIBOR panel have direct access to the FHLB System.

Although access to FHLB advances is limited among the LIBOR panel contribu-tors, Table 6 also illustrates that access to the Federal Reserve’s Discount Window isuniversal among this group of institutions. This fact suggests that the Federal Reservemight have an important role to play in moderating liquidity pressures by lendingto institutions without access to the FHLB System, such as foreign banks. This wasachieved in December 2007 through the introduction of the TAF and reciprocal cur-rency swaps with the European Central Bank and Swiss National Bank.

The TAF provides credit to institutions with access to the Discount Window andagainst the same range of collateral, but at an auction-determined interest rate ratherthan the administered primary credit rate. It was believed that the use of a market in-terest rate instead of an administered penalty rate would reduce any stigma associatedwith borrowing from the facility (Armantier, Krieger, and McAndrews 2008). Theswap lines with the foreign central banks allow U.S. dollars to be lent, on term, to for-eign banks without access to the Discount Window or TAF. The ability of these centralbanks to lend dollars institutions would presumably reduce the rates on unsecureddollar funding in their jurisdictions.

The purpose of the TAF and swap lines was to increase U.S. dollar liquidity, al-though the early evidence concerning the effectiveness is conflicting.33 Taylor andWilliams (2008) find that the daily LIBOR-OIS spreads (1-month and 3-month)are unaffected on TAF bid submission dates. By contrast, Wu (2008) finds thatthese spreads have been significantly lower since the original announcement ofthe TAF program. Finally, McAndrews, Sarkar, and Wang (2008) find that bothTAF-related announcements and operations have significant effects on changes inthe 3-month LIBOR-OIS spread. The same authors also conclude that announce-ments with regards to the European swap lines have a larger effect, consistent withour conjecture that non-FHLB members were under greater stress in the LIBORmarket.

The introduction of the TAF and swap lines also had important implications forthe relative attractiveness of FHLB advances since most FHLB members are eligiblefor the TAF. Figure 8 displays the all-in cost of borrowing via the TAF for each ofthe 28-day auctions since its inception in December 2007 through 2008 relative toa 30-day advance from the New York FHLB (see the Appendix for our all-in costmeasure for the TAF). The data suggest that the TAF all-in cost was roughly on parwith the FHLB as a source of funding according to our measure through February2008. However, this changed in a dramatic way in following the demise of BearStearns when the all-in cost spread between the TAF and the FHLB rose to 50 basis

32. In practice, this excludes all foreign institutions that control a U.S.-based FHLB member banksince the size of their U.S. operations is small relative to the consolidated institution by any reasonablemeasure of size.

33. The divergent results arise in part due the differences in the TAF-related announcements andoperations examined and whether the level or changes in the LIBOR-OIS spread are affected.


-120

-80

-40

0

40

80

120

0.0

1.0

2.0

3.0

4.0

5.0

6.0

Dec-08Sep-08Jun-08Mar-08Dec-07

Basis P

oin

tsPer

cent

Primary Credit Rate TAF 28 Days Spread in All-in-Cost bwt. TAF and FHLB NY (Right Axis)

FIG. 8. Primary Credit Rate, TAF Stop-Out Rate, and All-in Cost Spread between 28-Day TAF Funding and 30-Day

FHLB Advance.

SOURCE: Federal Reserve Bank of New York, Bloomberg, Authors calculations.

points. In fact, during this period the stop-out rate at the TAF auctions printed wellabove the primary credit rate from the Discount Window. In other words, despite thefact that banks could get funds from the Discount Window on the same terms, theychose to seek funds from the TAF at a higher price. This is an indication of continuedstigma for the Discount Window.

While the FHLB System was the lowest-cost source of secured term funding forU.S. depository institutions during fall 2007, the new liquidity facilities createdby the Federal Reserve in December 2007 complemented FHLB advances by ex-tending “stigma-free” term dollar credit to non-FHLB members, especially foreigninstitutions.

3.3 March 2008: Single-Tranche OMO, TSLF, and PDCF

As 2008 began, liquidity pressures manifested themselves through an unwillingnessof counterparties to extend term unsecured credit. But soon liquidity pressures beganto infect the secured funding markets. While precrisis Federal Agency repo interestrates averaged only 9 basis points above Treasury repo spreads, these same spreadsblew out into the 70- to 90-basis-point range in early March. As highly leveraged usersof repo credit, broker-dealers were hit especially hard by these liquidity pressures.While some of these institutions were owned by financial holding companies that are


affiliated with commercial banks, others were not.34 Note that, as of March 31, 2008,4 of the 19 primary dealers did not have commercial banking affiliates (includingBear Sterns and Lehman Brothers). This limited the ability of these primary dealersto address these liquidity problems internally.

The initial response of the Federal Reserve to alleviate these pressures was theintroduction of the Single-Tranche Open Market Operation (Single-Tranche OMO)Program and the expansion of the TSLF to allow primary dealers to swap (for afee) a range of securities (including AAA/Aaa-rated asset-backed securities) for U.S.Treasuries. However, liquidity pressures in secured funding markets soon worsened,and the most highly leveraged primary dealer, Bear Stearns, had to accept a buyoutfrom J.P. Morgan Chase in order to avoid bankruptcy.35 After the Bear Stearns episode,the Federal Reserve expanded access to the Discount Window by creating the PDCF.

The scale and scope of Federal Reserve lending increased significantly in March2008 with the introduction of the single-tranche OMO, the TSLF, and the PDCF. Eachof these facilities was designed to provide secured funding to primary dealers that didnot have direct access to either FHLB advances or the Federal Reserve’s DiscountWindow.

3.4 July and Early September 2008: Concerns about Fannie Mae and Freddie Mac

The reduction of the discount rate to 25 basis points over the federal funds target inMarch, 2008 almost established parity in terms of the all-in cost of Discount Windowloans and FHLB advances (according to our measure). During the following months,the Discount Window became more attractive from a pricing perspective (see Figure6). An important reason for this was a negative change in investor attitudes towardFederal Agency debt issues during the summer of 2008. Specifically, in July 2008concerns about the financial health for Fannie Mae and Freddie Mac led Congress togrant temporary authority to increase the lines of credit for the housing GSEs to anyamount deemed appropriate by the secretary of the U.S. Treasury. The FHLB Systemin part found itself “guilty by association” and saw additional pressures in terms offunding (see Figure 9).

In September 2008, the Treasury then announced a three-part plan aimed at sta-bilizing the residential mortgage finance market.36 First, the regulator of all threehousing GSEs, the FHFA, placed Fannie Mae and Freddie Mac into conservator-ship. Second, the Treasury established a Federal Agency MBS purchase program.Finally, the Treasury announced the creation of a GSE credit facility for Fannie Mae,Freddie Mac, and the FHLB System to be operated by the Federal Reserve Bank of

34. As discussed above, during the fall of 2007, the Federal Reserve granted a number of waivers ofSection 23A restrictions on lending by banks to affiliates in order to permit banks to support the liquidityneeds of their broker-dealer affiliates.

35. In order to facilitate the takeover, the Federal Reserve Bank of New York in consultation with theU.S. Treasury financed approximately $30 billion of assets from Bear Stearns via the Maiden Lane LLC.

36. See Frame (2009) for a detailed discussion.


FIG. 9. Three-Month FHLB Discount Note–Overnight Index Swaps (OIS) Periodic Auction, January 2007 to

December 2008.

New York.37 The direct effect of these three announcements was a marked tighteningin the spreads for Federal Agency debt and mortgage-backed securities.

3.5 September and October 2008: Lehman Brothers Bankruptcy Fallout

Following the Lehman Brothers bankruptcy filing in mid-September 2008, investorconfidence plummeted and market liquidity markedly contracted. Enormous financialpressures subsequently forced several large financial institutions to restructure, andthe federal government responded to the turmoil through an array of liquidity- andsolvency-enhancing measures.

During this time, market pressures forced some institutions into the arms of others,including the federal government. The Federal Reserve provided a liquidity backstopto insurer American International Group (AIG). The central bank then subsequentlyapproved applications by Goldman Sachs and Morgan Stanley to become bank holdingcompanies. Bank of America, which previously acquired Countrywide Financial,then announced its intention to acquire Merrill Lynch. Shortly thereafter, Washington

37. Any such loans will be fully collateralized (using either Federal Agency MBS or FHLB advances)and priced at LIBOR plus 50 basis points.


Mutual became the largest U.S. bank ever to fail—with most of its assets and liabilitiespurchased from the FDIC by J.P. Morgan Chase. Finally, Wachovia agreed to mergewith Wells Fargo in October 2008.

In an attempt to restore liquidity and stability to the U.S. financial system ingeneral—and the banking system in particular—public authorities took a numberof bold actions in September and October of 2008. First, the Federal Reserve signifi-cantly expanded its reciprocal currency swap arrangements—both in terms of size andnumber of eligible central banks. The Federal Reserve and Treasury then took severalsteps to ease investor concerns about the money market mutual fund industry and tosupport the functioning of the commercial paper market.38 In October, the Congresspassed the Emergency Economic Stabilization Act (EESA), a move that created the$700 billion Troubled Asset Relief Program (TARP), which was subsequently usedto inject about $200 billion of capital into U.S. banking organizations. The EESAalso temporarily increased FDIC insurance coverage to $250,000 per account. TheFDIC also announced the establishment of the Temporary Liquidity Guarantee Pro-gram (TLGP), under which the agency provides guarantees of non-interest-bearingtransaction deposits and selected newly issued senior unsecured obligations of partic-ipating banks for a fee. The TGLP provided banks with a close substitute for FHLBadvances and likely contributed to the decline in advances during the fourth quarterof 2008.

During the third quarter of 2008, the 12 FHLBs reported a combined net increasein advance lending of $100 billion. Based on their debt issuance activity, we esti-mate that virtually the entire amount of the increase occurred during the month ofSeptember (see Figure 1). This is consistent with a significant spread between TAFfinancing and FHLB advances during this month (see Figure 9). Moreover, banks withFHLB access appear to have received much more favorable LIBOR funding after theLehman Brothers bankruptcy. During the remainder of 2008, FHLB members paidon average 20 basis points less for LIBOR funding than nonmembers—with a dailyhigh of 38 basis points (see Figure 10). However, of note is that FHLB System earn-ings declined significantly during this period owing to losses related to the LehmanBrothers bankruptcy as well as other than temporary impairment charges on FHLBholdings of privately issued mortgage-backed securities. This latter issue continuedto be a challenge for the FHLBs into 2009.

3.6 The Balance Sheets of the FHLB System and the Federal Reserve

While both the FHLB System and the Federal Reserve have lent significant sums(with interest and against collateral) during the financial crisis, there are importantdifferences in terms of balance sheet mechanics. As the FHLB System has increasedlending, its balance sheet has expanded proportionately. FHLB advances are funded

38. In particular, the Treasury introduced an insurance program for money market mutual fund investorsand the Federal Reserve announced the creation of: (i) the Asset-Backed Commercial Paper Money MarketMutual Fund Liquidity Facility to extend nonrecourse loans to banks to finance their purchases of high-quality asset-backed commercial paper from money market mutual funds and (ii) the Commercial PaperFunding Facility to provide a liquidity backstop to U.S. issuers of commercial paper.


-5

0

5

10

15

20

25

30

35

40

45

Jan-07 Apr-07 Jul-07 Oct-07 Jan-08 Apr-08 Jul-08 Oct-08

Bas

is P

oin

ts

FIG. 10. Non-FHLB Member Less FHLB Full Member 1-Month Dollar LIBOR Bids, Daily Observations, January

2007 to December 2008.

SOURCE: LIBOR bids from Reuters; FHLB membership from Call and Thrift Reports, FHLB websites, National Information

Center.

with consolidated obligations plus capital stock that members are required to purchaseas a condition of their advance borrowings. Hence, the leverage of the FHLB Systemremained unchanged in the face of its tremendous growth during the second half of2007.

By contrast, the size of the Federal Reserve’s balance sheet remained virtually un-changed through mid-year 2008, but the composition of assets was altered markedly.The Federal Reserve lent out cash by either selling from its holdings of U.S. Trea-suries or not replenishing its portfolio as issues matured. Consequently, the FederalReserve’s holdings of Treasury securities fell substantially over time (Figure 11).However, following the Lehman Brothers bankruptcy, the magnitude of Federal Re-serve liquidity injections exceeded the central banks’ ability to sterilize the resultingincrease in reserve balances. In response, the Treasury implemented the Supplemen-tary Financing Program (which drains reserves) and the Federal Reserve receivedauthorization from Congress to begin to pay interest on reserves.

Without the FHLB System, the Federal Reserve would likely have faced signifi-cantly greater demand for borrowing at the Discount Window at on onset of the crisis.This would likely have influenced the Federal Reserve’s design of the various newliquidity facilities. Further, funding innovations like the Supplementary Financing


0

500

1,000

1,500

2,000

2,500


$ B

illi

on

s

Traditional Assets Liquidity Programs for Financial Firms

Support for Specific Institutions Direct Lending to Investors and Borrowers

Temporary Purchases of High Quality Assets

FIG. 11. Federal Reserve Domestic Financial Assets by Type: Weekly Observations, July 2007 to December 2008.

SOURCE: Federal Reserve Board H.4.1 release and the Federal Reserve Bank of New York.

NOTE: Unencumbered U.S. Treasuries = holdings of U.S. Treasuries less securities lent under TSLF. After March 25,

increases in the U.S. Treasuries holdings are attributed to “Temporary Purchases of High Quality Assets.” Liquidity

Programs for Financial Firms = Discount Window, TAF, central bank liquidity swaps, PDCF, and AMLF. Support for

Specific Institutions = net portfolio holdings of Maiden Lane LLC, Maiden Lane II LLC, Maiden Lane III LLC and

Credit extended to AIG, Inc. Direct Lending to Investors and Borrowers = CPFF and TALF. Temporary Purchases of

High-Quality Assets = purchases of U.S. Treasuries, Agencies and Agency MBS as authorized by the Federal Open

Market Committee.

Program and the authority to pay interest on reserves would likely have been enactedearlier.

While the FHLB System played an important role as a government-sponsoredliquidity provider for depository institutions during the second half of 2007, events in2008 revealed an important limitation. In particular, the Federal Agency debt marketssaw spreads widen appreciably during the summer and early fall, with little investorappetite for longer-dated issues. Hence, relying on market funding using an implicitgovernment guarantee is unlikely to be sufficient for a lender of last resort to beentirely effective during a financial crisis—exactly when you need one most.

4. CONCLUSION

The ongoing global financial crisis has provided an opportunity to learn about theroles of many often-overlooked financial institutions and financial markets. The FHLB


System is one such institution that emerged as an important provider of government-sponsored liquidity. Indeed, it was about 8 months into the crisis before the FederalReserve eclipsed the FHLB System in terms of crisis-related lending to the financialsystem. Nevertheless, the FHLB System remained the largest lender to U.S. depositoryinstitutions until the fall of 2008, as most of the Federal Reserve’s liquidity operationsprior to that had benefitted nondepository and foreign financial institutions. Withoutthe FHLB System, the Federal Reserve likely would have faced significant demandfor borrowing at the Discount Window at a much earlier stage of the crisis.

It is commonly noted that the structure of the current supervisory framework andsafety net for U.S. financial institutions is antiquated and fragmented. The lender oflast resort framework is no exception in that regard. Besides the Federal Reserve’sDiscount Window (and related liquidity facilities) and the FHLB System, there alsoexists the Central Liquidity Facility for credit unions (managed by the federal creditunion regulator, the National Credit Union Administration) and the credit facilitiesprovided by the U.S. Treasury to each of the three housing GSEs. Nevertheless,despite the institutional complexity of the existing lender of last resort framework,the ultimate lender of last resort is the U.S. Treasury and, by extension, the Americantaxpayers.

The tremendous upheaval in global financial markets and the perceived ineffective-ness of U.S. financial regulation, suggests to us that it would be helpful for Congressto revisit the roles of the respective lender of last resort facilities. In particular, the factthat different lender of last resort facilities have at the same time complemented andcompeted with each other raises important policy questions about which agencies ofgovernment should act as the lender of last resort and under what terms and condi-tions. The Federal Reserve’s and Treasury’s new, but ostensibly temporary, lendingfacilities of coupled with recent proposals to anoint the Federal Reserve as a “systemicrisk regulator” are clear indications of changes to come.

APPENDIX: ALL-IN COST MEASURES

A direct comparison of lending rates between the FHLB System and the FederalReserve is complicated by several factors. First, unlike within the Federal ReserveSystem, advance rates vary across individual FHLBs. We choose the 30-day advancerate from the FHLB New York. Second, on the margin, any borrowing from an FHLBrequires a simultaneous investment in capital stock. Consistent with the FHLB NewYork, we assume that this activity requirement is 4.5% of the borrowed amount andthat this investment is financed at the 1-month LIBOR rate. The stock investment,however, also earns dividends that we assume are 6%—consistent with the FHLBNew York’s experience during 2006. Third, borrowings from the Discount Windowcan be prepaid (and hence the rate reset), while the rate on 30-day advances is fixedfor the duration. We use the spread between the 1-month overnight index swap rateand the federal funds target to control for expected changes discount rate. Fourth,


TABLE A1

ALL-IN COST MEASURE COMPUTATION 30-DAY ADVANCE FROM FHLB NEW YORK VERSUS 30-DAY LOAN FROM

FEDERAL RESERVE DISCOUNT WINDOW DATA FOR SEPTEMBER 11, 2007

30-day loan/advance FHLB New York Discount Window

AAA-rated agency MBS collateral pledged 1,000,000 1,000,000Haircut 10% 7%Advance (subtotal) 900,000 930,000Activity stock purchase requirement 4.50% NAStock purchase 40,500 NARemaining left to fund 140,500 70,000Lending rate 5.01 5.751-month LIBOR 5.80 5.80Dividend rate 6.00 NA1-month OIS—federal funds target NA −0.34

All-in cost 5.08 5.41

SOURCE: Bloomberg and authors’ calculations.

while borrowings at both institutions have to be collateralized, the haircuts differacross collateral classes. The Discount Window haircuts have remained unchangedsince September 2006 whereas the FHLB New York adjusted its haircuts upward inApril of 2008. In order to account for these factors, we computed the all-in costs forthe Discount Window and the FHLB New York as

All-in cost at Discount Window = (1 − αDW ) × (PCR + OIS − FFR)

+ αDW × LIBOR (A1)

All-in cost at TAF = (1 − αDW ) × (TAF − FFR) + αDW × LIBOR (A2)

All-in cost at FHLB New York = (1 − αFHLB) × FHLBNY + (αFHLB + θ )

× LIBOR − θ × DIV (A3)

where αDW = Discount Window haircut (7%), αFHLB = FHLB New York hair-cut (10% (12% after April 9, 2008)), PCR = primary credit rate (discount rate),TAF = stop-out rate from TAF, LIBOR = London interbank offer rate (1 month,USD), FHLBNY = FHLB of New York 30-day advance rate, OIS = overnight indexswap rate (1-month USD), FFR = federal funds rate target, θ = capital stock purchaseactivity requirement (4.5%), and DIV = dividend rate (6.0%).

For illustrative purposes, in Table A1, we compute the all-in cost measures us-ing data from September 11, 2007. We focus on an AAA-rated Federal Agencymortgage-backed security with a market value of $1 million that needs to be fi-nanced for 1 month. The all-in cost at the FHLB New York was 5.08% comparedto the lending rate at 5.01%—a difference of 7 basis points. At the time the primarycredit rate stood at 5.75% but less than a week later the federal funds rate target was


lowered by 50 basis points. The OIS indicated that the federal funds rate, on aver-age, would be 34 basis points lower over the next month. Hence, the all-in expectedcost of the discount window was 5.41%, implying a spread in all-in costs betweenFHLB New York and the discount window of 33 basis points in favor of FHLBNew York.

LITERATURE CITED

Ambrose, Brent W., and Arthur Warga. (1996) “Implications of Privatization: The Costs toFannie Mae and Freddie Mac.” In Studies on Privatizing Fannie Mae and Freddie Mac,edited by U.S. Department of Housing and Urban Development, pp. 169–204. Washington,DC: Government Printing Office.

Ambrose, Brent W., and Arthur Warga. (2002) “Measuring Potential GSE Funding Advan-tages.” Journal of Real Estate Finance and Economics, 25, 129–50.

Armantier, Oliver, Sandy Krieger, and James McAndrews. (2008) “The Federal Reserve’sTerm Auction Facility.” Federal Reserve Bank of New York Current Issues in Economicsand Finance, 14, 1–11.

The Clearing House. (2007) “Major Banks Endorse Action by Federal Reserve.” Press Release,August 17.

Flannery, Mark J., and W. Scott Frame. (2006) “The Federal Home Loan Bank System: The‘Other’ Housing GSE.” Federal Reserve Bank of Atlanta Economic Review, 91, 33–54.

Frame, W. Scott, Diana Hancock, and Wayne Passmore. (2007) “Federal Home Loan BankAdvances and Commercial Bank Portfolio Composition.” Federal Reserve Bank of AtlantaWorking Paper No. 2007-17.

Frame, W. Scott. (2009) “The 2008 Federal Intervention to Stabilize Fannie Mae and FreddieMac.” Journal of Applied Finance, 18, 124–36.

Freixas, Xavier, Curzio Giannini, Glenn Hoggarth, and Farouck Soussa. (1999) “Lender ofLast Resort: A Review of the Literature.” Bank of England Financial Stability Review, 7,151–67.

Furfine, Craig. (2003) “Standing Facilities and Interbank Borrowing: Evidence from the FederalReserve’s New Discount Window.” International Finance, 6, 329–47.

Gatev, Evan, Til Schuermann, and Philip E. Strahan. (2005) “How Do Banks Manage LiquidityRisk? Evidence from Equity and Deposit Markets in the Fall of 1998.” In The Risks ofFinancial Institutions, edited by Mark S. Carey and Rene M. Stulz. Chicago: University ofChicago Press.

Madigan, Brian F., and William R. Nelson. (2002) “Proposed Revision to the Federal Reserve’sDiscount Window Lending Programs.” Federal Reserve Bulletin, 88, 313–19.

McAndrews, James, Asani Sarkar, and Zhenyu Wang. (2008) “The Effect of the Term AuctionFacility on the London Inter-Bank Offered Rate.” Federal Reserve Bank of New York StaffReport No. 335.

Nothaft, Frank E., James E. Pearce, and Stevan Stevanovic. (2002) “Debt Spreads betweenGSEs and Other Corporations.” Journal of Real Estate Finance and Economics, 25, 151–72.

Passmore, Wayne, Shane Sherlund, and Gillian Burgess. (2005) “The Effect of HousingGovernment-Sponsored Enterprises on Mortgage Rates.” Real Estate Economics, 33, 427–63.


Peristiani, Stavros. (1998) “The Growing Reluctance to Borrow at the Discount Window: AnEmpirical Investigation.” Review of Economics and Statistics, 80, 611–20.

Stojanovic, Dusan, Mark D. Vaughan, and Timothy J. Yeager. (2008) “Do Federal Home LoanBank Membership and Advances Increase Bank Risk-Taking?” Journal of Banking andFinance, 32, 680–98.

Taylor, John B., and John C. Williams. (2008) “A Black Swan in the Money Market.” NBERWorking Paper No. 13943.

Welsch, Roy, and Edwin Kuh. (1977) “Linear Regression Diagnostics.” MIT Sloan School ofManagement Technical Report No. 923-77.

Wu, Tao. (2008) “On the Effectiveness of the Federal Reserve’s New Liquidity Facilities.”Federal Reserve Bank of Dallas Working Paper No. 2008-08.

Copyright of Journal of Money, Credit & Banking (Blackwell) is the property of Wiley-Blackwell and its

content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's

express written permission. However, users may print, download, or email articles for individual use.

Documents

The Federal Home Loan Bank System: The Lender of Next-to-Last Resort?