Established in 1932, the FHLB System was created to enhance and facilitate the extension of mortgage credit to homebuyers. Since that time, its primary role as a GSE has remained largely unchanged. However, the FIRREA of 1989 added two additional responsibilities to the FHLB objectives, namely, funding for affordable housing to complement its basic housing finance function plus requirements to pay for a portion of the cost of the savings and loan debacle. The 1989 Act also provided opportunities to commercial banks and credit unions to become members of the FHLB System on a voluntary basis.
The combination of the FIRREA plus radical changes in the financial environment in which FHLBs and private financial institutions must operate have created several issues which should be addressed, not all of which are discussed in this presentation.(1) The primary issues to be discussed here are what could be called "inefficiencies" in the current system which have arisen over time. These three items are:
Now, these are not new issues. Indeed, these problems have been raised in several earlier studies which are discussed later in this paper.
In this paper, these topics, along with the issue of declining bank deposits in rural markets, are presented in a question and answer format. All of the topics discussed are elements of the S. 1423 currently under consideration by the Senate Banking Committee.
1. Is there a decline in the availability of funds in rural markets?
One of the arguments offered in favor of S. 1423 is that the expansion of eligible collateral and the provision to allow member institutions to use advances for funding small business, agricultural, rural development and low-income community development lending in addition to residential housing would offset the reduction in funds availability in rural communities precipitated by the reduction in bank deposits in those markets.
The issue at hand is therefore whether or not bank deposits, and the sources of funds for bank
loans, are declining or growing less rapidly in rural markets than in urban or MSA markets.
Clearly, this is an empirical question.
There are several pieces of evidence suggesting that this phenomenon is occurring. The first is anecdotal, and is based on the rapid increase in the growth of non-bank competitors in all markets, including non-urban. Mutual funds and brokerage firms have made substantial inroads into the traditional bank markets, and customers have moved substantial portions of their funds from bank deposits to mutual fund shares and brokerage firms. This is also true in rural markets where some brokerage firms such as Edward D. Jones concentrate their efforts. However, there are only data at the economy-wide level to indicate the extent of the growth of funds in non-bank firms. There are no available data on the change in relative shares of banks and non-banks in non-urban markets.
Second, there is clear evidence that the absolute level as well as the proportion of deposits in smaller financial institutions, banks and thrifts, are declining. For example, in data on bank deposits reported on the FDIC website and summarized in Table 1, we see that the actual level of deposits in banks and savings institutions with total assets below $100 million fell from $290 billion as of June 30, 1996 to $269 billion as of June 30, 1997, a decrease of 7.2% in one year. As a percentage of total deposits, deposits in smaller institutions declined from 8.7% of the total of $3.3 trillion in 1996 to 7.7% of the total deposit base of $3.5 trillion in 1997, a decline of one full percentage point in one year. Deposits in institutions under $500 million in total assets declined to 23.8% of total deposits in 1997 from 25.7% in 1996.
While this is only one "snapshot", and an imperfect one, it does indicate that deposits in smaller institutions are shrinking. Since most of the smaller institutions are in non-urban markets, the clear implication is that deposits in these markets are declining absolutely and relatively.
Third, additional evidence on the behavior of bank deposits in non-urban markets is reported in Table 2 which presents deposits in community financial institutions with less than $500 million in total assets relative to total domestic deposits in all insured institutions over the 1990-1997 period. From a base of $312 billion in 1990, deposits in community financial institutions increased by only 2% over the 1990-97 period to $319 billion. Total domestic deposits in all markets in all institutions increased by nearly 18% over this period. The slower growth rate led to a reduction of 12.3% of domestic deposits residing in community financial institutions in non-urban markets in 1990 to 10.6% by 1997. Clearly, the evidence suggests that relative shares have fallen in rural markets in smaller institutions, and that the bulk of the reduction has occurred since 1995.
Fourth, there have been significant numbers of bank acquisitions and mergers in the past several years, resulting in numerous branch closings in both rural and urban markets. In some cases, branches, along with their deposits, in rural markets are sold to local banks. However, at this point the effects of bank acquisitions and mergers on deposit structure and credit availability are unclear. Several studies are underway by scholars in the Federal Reserve System addressing these and other aspects of bank mergers.
In summary, there does appear to be evidence of a declining, certainly on a relative basis, deposit base in rural markets in smaller institutions. As such, ease of access to the advance window of the FHLB, and the expanded use of those funds from advances would add to the availability of funds in non-urban markets.
2. What has happened to the mix of member institutions of the FHLB System,
especially as it pertains to mandatory members? Are these good reasons to remove
the mandatory membership requirement for federally chartered savings institutions
and make membership in the FHLB System completely voluntary?
Before FIRREA, all FSLIC-insured institutions were required to be members of the FHLB System and FDIC-insured savings banks were allowed to join on a voluntary basis. In 1995, state-chartered SAIF-insured institutions were given the option of becoming voluntary members. That is, only federally chartered savings institutions now face mandatory membership in the system. Since they have the option of converting their charter to either state savings institutions (in some states) or to a commercial bank charter, the option to be a voluntary member basically exists for all members.
As indicated in Table 3, the mix of voluntary and mandatory members has changed quite dramatically in recent years. Since 1990, the number of mandatory members has fallen from 1509 to 1002, a decrease of 34%. As shown in Table 4, assets held by these mandatory members decreased from $855 billion in 1990 to $683 billion in 1997, a decrease of 20%. At the same time, the number of voluntary members increased from 1339 in 1990 to 5254 in 1997, an increase of 292%. Most of this increase in membership came from commercial banks voluntarily joining the system. By 1997, the voluntary membership comprised 84% of the total FHLB membership and 75% of total assets of member institutions. In terms of total assets held by member institutions, the proportion accounted for by the mandatory group fell from 67% in 1990 to 25% in 1997. This was the result of the significant reduction in the number and total assets of mandatory members and the dramatic expansion of both numbers and total assets of the voluntary membership base.
An analysis of FHLB membership trends and a chart showing the decline of thrift membership and the increase in commercial bank membership was presented in a recent analysis of the FHLB System by Moody's Investors Service.(2) An excerpt from that report is shown below along with the "membership makeup" chart attached as Figure 1.
"Commercial banks continue to alter the traditional mix of the System's membership base by entering at a rate that exceeds all other types of institutions. By June 1997, System membership reached a record level by climbing to 6,300, up 2.5% since December 1996, and by 196% since the passage of FIRREA in 1989. In 1994, the System was made up of 2,067 thrifts, 3,133 commercial banks, and 106 credit unions and insurance companies. By the middle of 1997, the number of thrifts had dropped to 1,801 while the number of commercial banks had increased to 4,274. Credit unions and insurance companies increased, but still retain a small presence with just 225 members. The pace of membership growth is expected to slow, considering the finite and consolidating membership pool. However, proposed legislation that may loosen eligibility requirements for smaller institutions, and the recent permission granted to nationally chartered banks by the Office of the Comptroller of the Currency to become shareholding members, should allow for additional membership growth in the near term."
There are several good reasons for making membership in the FHLB System completely voluntary and totally eliminating the mandatory membership requirement. First, some of the current mandatory members do not use any of the services of the FHLB System. If they demonstrate by their non-use they do not want or need such services, there is little reason to impose such a membership on them.
Second, converting membership to a completely voluntary basis would force the FHLB managers to provide services that add value for the members. Essentially, it would force the managers in the system to compete for members both by providing value-added services and by meeting the changing needs of these members and their markets. Since more than 5000 new voluntary members have joined the FHLB System since FIRREA, there is a significant amount of evidence that the system is able to provide services with value. As such, there appears to be little, if any, risk that removing the mandatory requirement would precipitate an outflow of members from the system and thereby impair the capital of the FHLBs as they leave the system.
Third, providing for an all-voluntary membership introduces an interesting market test for a GSE, which is very unique. Should FHLBs fail to offer services with value, resulting in members exiting the system, it would represent a clear market signal that the FHLB System has failed to provide useful service and thus has outlived it usefulness. The requirement to meet a "market test" on an on-going basis is much more likely to lead to a viable FHLB System, which will then be better equipped to meet its requirements for providing affordable housing programs as well as meet its fixed obligations on the REFCORP bonds.
Fourth, as shown in Table 3, the number of institutions required by statute to be members of the FHLB System is shrinking as the savings and loan industry consolidates and/or is acquired. Therefore, mandatory membership is already disappearing due to market forces. If the FHLB System were forced to rely on mandatory members to meet its obligations, the system might actually have the perverse incentive to take on greater risk to meet its obligations, and thereby make it more risky overall.
In summary, there appears to be a strong case to remove the mandatory membership requirement
currently in place and place all membership on a voluntary basis.
3. Are there any current impediments to community financial institutions (non-mandatory members) joining the system on a voluntary basis?
At the present time, the law requires the insured institution to have at least ten percent of total assets in residential mortgage loans in order to be eligible to become a member of the FHLB System. This would be altered so that those institutions which quality as "community financial institutions"(3) would no longer be required to meet the ten percent residential mortgage loan requirement. They would be required, as before, to demonstrate financial soundness so that advances can be safely made by the FHLBs.
In short, easing the membership eligibility requirements serves to fulfill the purpose of the 1989 FIRREA which allowed and encouraged community banks to become members of the FHLB System.
4. Would the basic functions of the FHLB System be changed as a result of S. 1423?
The primary provisions of the proposed bill include the following items:
The current mission of the FHLB as a GSE is to facilitate the extension of mortgage credit. Initially, the FHLB activities in this mission were limited to extending advances to and establishing various programs with savings and loan associations. As outlined above, some but not all of these S&Ls were mandatory members of the system and were obligated to own stock in the FHLB in their district.
To a certain extent, there was a change in the role of the FHLB system in 1989 when, following passage of FIRREA, commercial banks were allowed to become members of the FHLB System. Since that time as shown in Table 3, the number of commercial bank members has increased to 4514 as of the end of 1997, or 72% of total membership and 66% of assets of FHLB System members.
Also as a result of FIRREA, FHLBs were given responsibility to contribute a portion of their earnings ($300 million annually) toward payment of the obligations on the REFCORP bonds issued to pay the cost of the bankrupt savings and loan resolution and to become more proactive in working with the private sector in providing and developing low income housing.
As constituted, the bill would do nothing to alter the mission and functions that are already in place by existing legislation and regulation. Rather, the bill could best be interpreted as a modernization of the system to better reflect the economic and financial environment within which both the FHLB system and member institutions must operate.
5. Is there an expansion of activities of the FHLBs contemplated?
The list of activities in which the FHLBs can engage is restricted to the following:
Under the proposed legislation, the list of activities would not be changed. The legislation offers no provision whereby the FHLBs themselves would be able to engage in additional activities such as direct lending for housing, community development or economic development. Neither is there any provision for the FHLBs to engage in securitization of assets on their own balance sheet. In contrast, some GSEs do engage in direct lending and securitization activities and in some cases become actual competitors of private institutions such as banks, savings institutions, credit unions and others.
6. Would the activities of member institutions be changed under the proposed bill?
The overall powers of member institutions would be unaffected by the provisions of S. 1423. However, member institutions would now be allowed to use advance proceeds from the FHLBs for small business, agricultural, rural development, or low--income community development lending. At the present time, long-term advances can be used only for the purpose of funding residential housing finance.
While this appears to be an expansion in how funds from advances may be used in the member institution, in reality funds have always been fungible. That is, deposit and/or fund flows are seldom, if ever, traceable to a specific asset use in any financial institutions. Advances, both short and long-term, have always been used as tools whereby institutions could achieve certain desired asset-liability combinations.
There may be, however, a benefit from the specific listing of certain additional allowable uses of funds. Financial institutions are always searching for ways by which to improve their service levels to traditionally underserved areas, both for CRA purposes and general banking purposes. Having access to funds for particular types of economic development purposes, especially low-income development lending, should enhance the flow of funds to the underserved portion of the community.
7. Is the QTL test a logical measure by which to determine the amount of borrowing a
member can have from the FHLB? Put another way, should membership rules be
identical for all members?
Currently, member institutions which meet the QTL-test (qualified thrift lender) are able to obtain and use advances on more favorable terms than those which do not meet the QTL-test. As one might expect, the specific rules governing membership requirements and advance rules for non-QTL members and QTL members are complex and are not detailed in this paper.(4)
At the time the advance mechanism was established for savings and loans in the 1930's, there was some logic for the requirement since the thrift industry was the dominant, if not the only, funding source for residential housing finance. Since that time, however, financial markets and industry conditions have changed dramatically such that savings institutions now play only a minor role in providing residential housing finance.
These are several good reasons for eliminating the QTL-test as a means of determining the amount of advances by a member institution and the amount of capital required to support those advances. First, simple logic suggests that all members should be subject to the same membership rules. This enables institutions to estimate the benefits and costs of membership on a more equitable basis, and more closely align the interests of all parties.
Second, as mentioned above, the QTL-test is a relic from the financial system and mortgage market of the 1960's and earlier when specialized institutions were commonplace, especially residential finance lenders. Economic and regulatory events have reshaped the financial scene so that the mortgage market is now spread among many providers. Indeed, if one were to attempt to channel advances to the largest "residential mortgage providers", the most likely recipients would be free-standing, pure-play mortgage banking companies and mortgage-banking subsidiaries of large commercial banks and non-banks. There is little justification for or support for such a proposal. Hence, the QTL-test has no relevance to the structure of the current financial system.
Third, one of the objectives of the advance window and the FHLB System has been to increase the flow of funds for housing and for economic development into smaller markets, based on the extension of membership to community financial institutions, including banks, by the 1989 FIRREA. Limiting the access to funds to only those institutions who have met some past allocation-of-assets test appears to be inconsistent with that objective.
In summary, by treating all members, and potential members, on an equal basis, the current intent
of the FHLB System as amended in 1989 would be better served.
8. What are the costs and benefits of expanding the list of acceptable collateral as
securities for advances for FHLBs?
At the present time, eligible collateral to provide security for advances consists of the following items.(5)
S. 1423 would modestly expand the above list to include the following:
The benefits of expanding the list of eligible collateral include:
The costs of expanding the list include:
However, there appears to be little, if any, risk in expanding the list of eligible securities, for the following reasons:
First, the FHLBs have experienced no credit losses in their 60 plus year history. This impeccable loss record reflects the conservative "underwriting" and credit standards used by the FHLBs in extending credit to members. Basically, they have accomplished this zero-loss record by credit-enhancing using over-collateralization; i.e. requiring collateral with a current market value of at least 110 percent of the amount of the advance. There is no reason to believe that continued credit enhancement by over-collateralization would be unsuccessful in the future, since the same methods could be utilized for the expanded list of eligible collateral.
Second, the FHLBs may choose to utilize additional credit enhancement methods to ensure the continued safety of lending to members. Examples include varying the amount of over collateralization depending on the credit risk exposure of the asset pledged; for example, even higher levels for small business loans. In addition, security could be provided by the host of credit enhancement techniques that have been developed in financial markets in recent years, such as letters of credit, standby letters of credit, requiring member institutions to provide first loss protection of the collateral via insurance, etc.
In short, there is absolutely no doubt that FHLBs have experienced outstanding (zero-loss) performance in underwriting credit for advances to members. Given their success in credit enhancement, the evidence suggests that expanding the list of eligible collateral would not endanger the continued success of credit risk management.
9. Do the current capital requirements for the FHLB system and the individual FHLBs reflect relevant economic and financial risk exposure of the system and the banks? Are they consistent with other GSEs and with banks?
The role of capital in a financial institution is to provide a cushion for any unanticipated or unexpected losses that might occur. The capital held by the institution should reflect the risk exposure to unanticipated changes in underlying economic, interest rate, credit, liquidity or other conditions in the environment.
However, for the FHLB System and the banks therein, the amount of capital held is determined by the FHLB stock purchase requirements imposed on member institutions by their membership status and their use of the advance mechanism. That is, a formula determines how much FHLB stock a member must purchase upon joining the system and an additional formula (or formulas) determine whether additional stock must be purchased when they borrow funds from FHLBs at the advance window.
In short, FHLBs are recipients on a passive basis of the amount of capital they hold, rather than managers of their capital base. The current capital base therefore changes as membership in the system changes. Adding new members adds to the capital base while losing members reduces the capital base.
Basically, there are two capital issues which should be addressed; (1) the current lack of permanence in the capital base, (2) the fact that the capital does not reflect the risk exposure of the activities of the FHLBs. The permanence issue arises because as membership mix is altered, those leaving the system can elect to redeem their stock investment in the FHLB. Thus, the capital base is not like the usual common equity which can be traded among investors but not redeemed. Unlike most common equity, the capital base of the FHLB System is much less capable of dealing with the loss absorption function normally required of equity capital. To ensure the viability of the FHLB System, this problem should be addressed. Provisions included in the proposed legislation address this issue.
The second issue arises because the capital required in the FHLB System is unrelated to the risks undertaken by and risk exposure of the FHLBs. This, of course, arises and follows from the fact that capital requirements are based on membership mix and advance levels rather than the risk exposure of the system. Clearly, this is a problem which should be addressed.
Basically, there are three choices available to deal with the issue. The first is to leave the capital requirement in its present form. This is not wise from a purely financial and economic perspective. The second is to adopt the capital rules that apply to Fannie Mae and Freddie Mac. The third is to adopt a variation of the risk-based capital rules applicable to insured deposit institutions.
Each of the second and third suffer from their own unique criticisms as applied to the FHLBs. For example, a stress-test model applied to the FHLBs for both credit and interest rate risk would be less appropriate than for Freddie Mac and Fannie Mac. This arises from the fact that in a stress test on FHLBs, one would be required to make certain assumptions about the membership pattern, and hence capital, in the system as voluntary members might leave and mandatory members might fail. Thus, a pure-stress test approach is an imperfect test of the financial strength of the FHLB System. At the same time, a pure risk-based capital (RBC) approach as utilized for insured deposit institutions suffers from the lack of an interest rate risk test explicitly accounted for in the capital requirement.
Realistically, some blended approach appears to be the appropriate solution. One approach is to impose a RBC requirement on System banks based on their on and off-balance sheet credit risk exposures, similar to that imposed on commercial banks and thrifts based on BIS standards, complemented by basic leverage requirement. The second part would be an additional capital requirement based on the interest rate risk exposure of the FHLBs as measured by a "stress test" which tests for the volatility of the FHLBs portfolio (and hence equity) under alternative scenarios of rate level changes, rate volatility changes and yield curve changes.(6)
None of these ideas about altering FHLB capital requirements are new. Numerous studies within and outside the government have been conducted on this capital question and other issues (see more detail in Section 10 below). Studies have been conducted by GAO, HUD, FHFB, and Capital Study Task Forces in the FHLB system. These studies have been supported by private sector reports by J.P. Morgan (investment banking), Price Waterhouse (accounting firm), and Gibson, Dunn and Crutcher (law firm). In addition, the capital issue is discussed regularly by investment ratings services such as Moody's in their periodic reviews of the debt issues of the FHLB System and of individual FHLBs.
Despite the oddity of the capital requirements and capital structure of the FHLB System, the obligations of the FHLBs receive an Aaa rating, based on the strong financial fundamentals of the banks, their sound management, and their GSE status.
In summary, it is a misnomer to take the current FHLB capital rules as "capital requirements" because the term capital requirements usually refers to the amount of capital an institution is required to hold based on some measures of the riskiness of the institution's activities. As mere "recipients" of capital from member purchases of FHLB stock, the FHLB system is unique among GSEs and other private financial institutions.
The capital issue should be resolved by making the capital base of the FHLB System less subject to withdrawal risks and more reflective of the business risk of the FHLBs. This can be accomplished by imposing both capital standards for credit risk similar to those for banks and also "stress-testing" the adequacy of the FHLB capital against exogenous shocks in interest rate levels and structure.
Like many issues dealing with government and GSEs, this capital situation in the FHLB System has not been addressed because there has not been a serious problem to date. The FHLBs have never lost money, they have very high asset quality (low credit risk) and their interest rate risk exposure is small. Thus, the issue has been treated with benign neglect. The problem, however, remains and it is clearly easier to deal with it in a non-crisis environment than in a problem environment. One need only recall the disaster which resulted from the delay in dealing with the savings and loan crisis to recognize the benefits of dealing with problems and issues before they turn into serious problems.
10. Has S.1423 been the only suggested reform of the FHLB system?
As a result of previous legislation, including the Housing and Community Development Act (HCDA) of 1992, the Federal Housing Enterprises Financial Safety and Soundness Act of 1992, several groups have been charged with the responsibility of studying and offering recommendations regarding the operation, structure and functions of the FHLB System. Over the past several years, there have been (at least) the following studies:
In addition, there have been at least two studies conducted on the capital structure and capital adequacy of the FHLB System, with the latest published in November 1994 by the FHLBs Capital Study Task Force.
Providing details on the recommendations of each of these studies and reports is beyond the scope of this paper, but it is clear that the recommendations of these reports are consistent with most of the provisions suggested in S.1423 regarding the issues discussed in this paper, namely
In short, one does wonder how many times the obvious points must be restated and repeated. There is general agreement that reforms along the lines suggested by S.1423 and the previous studies cited above should be implemented.
Congress has determined by previous legislation that the FHLB System should play a role in housing finance. Congress has also determined that commercial banks should be allowed to be members of the FHLB System. Housing and economic development are inexorably intertwined in local markets, perhaps especially in rural markets. The bill under consideration would clearly facilitate and enhance the flow of funds to local markets, especially by removing obstacles to membership and using the advance window.
Furthermore, numerous recent studies by various government agencies concur that reform of the FHLB System along the lines suggested by S.1423 is warranted. Since this proposed legislation both facilitates the interest of Congressional wishes and incorporates the most important features of the recommendations of several recent studies, serious consideration of the proposals therein seems warranted.
Finally, although the credit market continues to rate the obligations of the FHLB System and
FHLBs at the Aaa level, mention is continually made of the lack of rationality that prevails in the
capital requirements imposed on the system, specifically regarding the issue of permanence in the
capital base and the non-relationship to actual risk exposure of the system. Prudence and wise
policy-making suggest that these capital problems, and the other issues discussed in the paper, be
addressed in a non-crisis environment when the solutions can be achieved at the lowest possible
1. The major issue not discussed because it is not relevant to the fundamental points raised here is the earnings pressure on FHLBs resulting from the requirement to pay $300M per year for the next several decades toward the cost of paying the obligations on the REFCORP bonds issued to pay off claims generated at insolvent S&Ls.
2. Moody's Investors Service, Global Credit Research, Federal Home Loan Banks, November 1997.
3. An institution whose deposits are insured by the FDIC and has $500 million or less in total assets, with the $500M adjusted annually by the CPI.
4. The reader is referred to the extensive treatment of this issue in U.S. Department of HUD, Office of Policy Development and Research, Report to Congress on the FHLB, Vol. I, Summary Analysis and Recommendations, April 1994. Chapter IV particular provides considerable detail on this question.
5. Summarized from Title 12, Banks and Banking, Chapter 11, Federal Home Loan Banks in Section 1430.
6. There is a rapidly growing body of research which addresses the issue of capital requirements
in financial institutions based on the exposure of those institutions to movements in interest rate
levels and structures. In banking circles, considerable discussion is underway on establishing
capital requirements for market risk based on the use of a bank's internal measurement model,
subject to review by appropriate bank regulators. See, for example, D. Hendricks and B. Hirtle,
"Bank Capital Requirements for Market Risk: The Internal Models Approach," Economic Policy
Review, FRB of New York, December 1997, 1-12.
|Asset Size||Total Deposits (millions)||Total Deposits (millions)|
|Less than $100||$289,814||8.7%||$269,407||7.7%|
Source: FDIC, Summary of Deposits Data Book, 1996, 1997. Accessed at FDIC website:
|Domestic Deposits in
Institutions in Nonurban
Deposits in All
Markets in All
|Percent of Total
a Community financial institution is defined as one having less than $500 million in total assets.
Source: FDIC, Summary of Deposits, 1990-1997.
|Number of Voluntary Members|
|End of||Number of
Source: FHLB of Atlanta.
|Total Assets Held By*
Voluntary Members (millions)
Held by Mandatory Members
(millions of $)
Source: FHLB of Atlanta.
* Totals do not include credit unions or insurance companies.
** OTS Factbook, 1990-1994.
Home | Menu | Links | Info | Chairman's Page