Skip Header

Federal Deposit
Insurance Corporation

Each depositor insured to at least $250,000 per insured bank

Home > News & Events > Speeches & Testimony

Speeches & Testimony

Statement of Michael H. Krimminger Deputy to the Chairman Federal Deposit Insurance Corporation On Covered Bonds: Potential Uses And Regulatory Issues Committee on Banking, Housing, and Urban Affairs, U.S. Senate; 538 Dirksen Senate Office Building
September 15, 2010

Chairman Dodd, Ranking Member Shelby and members of the Committee, on behalf of the Federal Deposit Insurance Corporation (FDIC), thank you for the opportunity to testify on the regulatory and legislative issues posed by covered bonds. The FDIC has long worked with the financial industry to establish a sound foundation for a vibrant covered bond market that will provide U.S. banks with an additional source of liquidity. These efforts include working with the first U.S. banks to issue covered bonds in 2006 and the FDIC's adoption of a Statement of Policy in mid-2008 to clarify key issues related to deposit insurance and bank resolutions. With this background, we hope our views on the covered bond market may be helpful for the Committee.

The FDIC supports balanced legislation to create a sound foundation for covered bonds that also promotes market discipline and protects the Deposit Insurance Fund (DIF). In order to meet these goals, we believe that there are three key principles that should be followed. First, the rights and responsibilities of investors, issuers, and regulators should be clearly defined. Second, the investment risks to covered bond investors should not be transferred to the public sector or to the DIF. Third, the legislative framework should be consistent with long-standing U.S. law and policy, and not unduly impair the interests of depositors and other creditors.

While the FDIC's existing Statement of Policy provides a sound foundation, a properly designed legislative and regulatory framework could further facilitate development of a vibrant covered bond market. In doing so, however, it is important to not create a new class of investments that appears 'risk-free' by providing investors with protections unavailable for any other investment. We have already seen the consequences when risks are mispriced in the market. Most importantly, the risks should not be transferred, implicitly or explicitly, to the government or the DIF. While covered bonds can be a valuable tool to provide liquidity, they do carry risks that should be considered in fashioning any final legislation.

Our testimony will discuss the FDIC's July 28, 2008 "Policy Statement on Covered Bonds," provide background on covered bonds and their potential role in the financial marketplace, and address the proposed legislation recently adopted by the House Financial Services Committee, H.R. 5823, the "United States Covered Bond Act of 2010."

The FDIC's Existing Policy on Covered Bonds

Before the crisis, the FDIC worked closely with Washington Mutual Bank and Bank of America when they launched the first U.S. covered bond programs in 2006. As a result of our efforts, the banks were able to issue covered bonds at a competitive price. The 2008 Statement of Policy later adopted by the FDIC's Board of Directors addressed questions from the marketplace about how covered bonds would be treated in the receivership of an issuing bank. The market's reaction to this Statement was very positive and most commentators stated that it provided a solid foundation for the covered bond market. Shortly after the adoption of the Statement of Policy, the Department of the Treasury issued a companion document entitled "Best Practices for Residential Covered Bonds" to establish greater clarity and homogeneity for the market so that investors would have confidence in future issuances. The FDIC worked with the Treasury Department in developing the Best Practices to create a coordinated framework for the responsible and measured roll-out and further development of covered bonds in the U.S. Unfortunately, the financial crisis disrupted all forms of structured finance. Even during the crisis, however, the FDIC was able to sell Washington Mutual's covered bond program intact to JPMorgan Chase Bank in a failed bank resolution - demonstrating the effectiveness of the process outlined in our Statement of Policy.

Given the FDIC's existing Statement of Policy, the Treasury Department's companion Best Practices, and the prior successful covered bond programs developed in cooperation with the FDIC, it is unclear that legislation is necessary to re-launch the market. At a minimum, the FDIC suggests that its Statement of Policy should be considered as a framework for any legislation in order to provide a sound, balanced foundation for the market.

Covered Bonds in Context

Covered bonds are general obligation bonds of the issuer, normally an insured bank or thrift, with payment secured by a pledge of a pool of loans. During normal operations, like any general obligation corporate bond, investors are paid from the issuing bank's general cash flows, while the cover pool of loans serves simply as collateral for the bank's duty to pay the investors. As a result, both functionally and legally, the cover pool is not the source for repayment as in a securitization, but is simply collateral to secure payment if the issuing bank cannot make payment from its general cash flows.

Another distinction between covered bonds and most securitizations further demonstrates that the cover pools function as collateral and not as sources of payment when covered bonds are not in default. In a covered bond, any loans and other assets in the cover pool that become delinquent must be replaced with performing assets. As a result, the collateral for the covered bond is constantly refreshed - and imposes an ongoing obligation on the issuing bank to produce new loans or other qualifying collateral to replace delinquencies. Finally, the issuer must always maintain more collateral in the cover pool than the outstanding notional or 'face' balance of the outstanding bonds. If the issuing bank fails to pay on the covered bond, then the investors have recourse to the cover pool as secured creditors. This is precisely how normal collateral arrangements work in other secured transactions.

Under the long-standing U.S. law applied to all types of secured transactions, secured creditors have a claim to the collateral - here the loans or other assets pledged to secure payment on the covered bond - only to the full amount of their claim for payment at the time of any default. They do not have a claim to any part of the value of the collateral that exceeds their current claim for payment. Any collateral or proceeds in excess of that claim for payment are returned to the debtor or, if it has been placed into bankruptcy or receivership, are used to pay the claims of unsecured creditors. If, on the other hand, the secured creditor's claims are greater than the value of the collateral, the creditor will have a secured claim up to the value of the collateral and an unsecured, general claim for the remaining balance along with other unsecured creditors.

The same rules apply in FDIC receiverships. Secured creditors are fully protected under Section 11(e)(12) of the Federal Deposit Insurance Act (FDI Act) for the amount of their claim up to the value of the collateral. As a result, covered bonds provide two avenues for recovery - from the issuing bank and from the cover pool of collateral. What they do not have, under U.S. law, is a right to keep collateral in excess of their right to payment.

Legislation to Address Covered Bonds

As mentioned at the outset, the FDIC supports balanced covered bond legislation. We believe this legislation should embody three key principles. First, it should clarify the rights and responsibilities of investors, issuers, and regulators. Second, it should ensure that investment risks are not be transferred to the public sector or to the DIF. Third, it should remain consistent with long-standing U.S. law and policy for secured creditors. Unfortunately, H.R. 5823 would muddy the relationship between investors and regulators, transfer some of the investment risks to the public sector and the DIF, and provide covered bond investors with rights that no other creditors have in a bank receivership. As a result, this legislation could lead to increased losses in failed banks that have issued covered bonds.

Clarifying Rights and Responsibilities - To clarify the respective roles of investors, issuers and regulators, we suggest that any legislation establish a regulatory framework for the appropriate federal regulators to jointly establish standards for covered bond issuances by regulated institutions. One existing forum for setting such joint standards is the Federal Financial Institutions Examination Council, which includes the federal regulators and a representative from the Conference of State Bank Supervisors. H.R. 5823 provides an alternative approach - by making the Federal prudential regulators the covered bond regulators - which could also be workable.

The resulting standards, like the FDIC's Statement of Policy, should address the key elements in covered bond transactions and the safety and soundness issues that can be implicated by a bank's use of covered bonds. The standards should address the types of collateral, underwriting standards, required over-collateralization, frequency and content of reports on collateral and satisfaction of required over-collateralization, disclosure standards for performance of underlying loans or assets, and the rights of the investors in the event of default. As discussed in greater detail later, a particularly important element in clarification of investors' rights is the treatment of the covered bonds if the issuer defaults on its payments under the bonds. This is both critical to the investor and to the relative balance of risks retained by the investor or transferred to other parties.

The standards setters for covered bonds should have discretion in expanding the use of covered bonds and categories of cover pool assets as sustainable markets develop and the liquidity of the instruments increases. The gradual expansion of cover pool categories is essential to ensure the quality of covered bonds and of the assets in the cover pools.

Unfortunately, H.R. 5823 appears to go beyond setting standards to provide for detailed oversight of the covered bond program for the benefit of the investors. This shift of the focus of federal regulation towards protection of the investment interest of specific investors raises significant questions about the proper role of federal regulation for individual investment programs. It must be made clear that the federal regulators are not guarantors of performance by the issuing banks and are not responsible for ensuring that the banks do not breach any of the standards. The federal government should not determine the roles, responsibilities, or quality of performance of the issuers or be perceived as protecting the investment interests of specific investors. These are issues best resolved by private contracts based on transparent disclosures about the operations of covered bond programs. It is important that the federal government is not viewed in any way as a guarantor of performance under the covered bonds. Performance should be a matter of private contract.

In addition, H.R. 5823 would also make the Federal prudential regulators the appointing and supervising authority of trustees that would operate the separate estates of the covered bonds. This level of government entanglement in what are private contractual matters could lead to an implied guarantee of covered bonds. An implied guarantee of covered bonds would put covered bonds on a near par with the government sponsored enterprises - a status that should not be granted without strong policy reasons because of the risk that status represents for taxpayers.

Legislation Should not Increase the Potential Loss to the DIF - Intimately related to the foregoing principle is the key issue for the FDIC - new covered bond legislation should not limit the FDIC's ability to recover the losses the DIF incurs in resolving a failed bank. To protect the DIF, any covered bond legislation must preserve the flexibility that current law provides to the FDIC in resolving failed banks - including the options of continuing to perform under the covered bond program pending a sale of the program to another bank, turn-over of the collateral to the investors, and repudiation - a statutory termination of the contracts - of the covered bond obligation.

Because there is sometimes confusion concerning the FDIC's power to repudiate, it requires some explanation. Repudiation is the ability of the FDIC to terminate (or breach) a contract and then pay statutorily-defined damages to the other parties. In the case of covered bonds, repudiation allows the FDIC, as receiver for the failed issuer, to cut-off future claims and end the obligation to replenish the cover pool with new assets. Under the FDI Act, the FDIC will then pay damages to compensate the covered bond investors.

Covered bond investors, as noted above, are secured creditors of the bank. The amount of their claim is defined by the balance or par value of outstanding bonds plus interest. The FDIC would support covered bond legislation that clarifies the amount of repudiation damages to be the par value of outstanding bonds plus interest accrued through the date of payment. This provides a remedy that fully reimburses the covered bond investors. In return, as in any other repudiation, the FDIC as receiver would be entitled to reclaim the collateral in the cover pool after payment of those damages. The FDIC could then sell this collateral and use the proceeds to pay part of the claims of the DIF (which has a claim after meeting its insurance obligation for insured deposits), uninsured depositors, and other creditors of the failed bank.

If the FDIC does not repudiate a covered bond, it should have the authority to continue to perform under the covered bond until it can sell the program to another bank. This would not expose the investors to any loss, by definition, since the FDIC would meet all requirements of the covered bond program, including replenishment of the cover pool and meeting the over-collateralization requirement. As long as the FDIC is performing under a covered bond agreement, covered bond legislation should not limit the time in which the FDIC has to decide how best to proceed.

Any legislation that fails to preserve these important receivership authorities makes the FDIC the de facto guarantor of covered bonds and the de facto insurer of covered bond investors. Unfortunately, H.R. 5823 would expose the DIF to additional losses by restricting the FDIC's ability to maximize recoveries on failed bank operations and assets. This is contrary to a long-standing Congressional goal of preserving the DIF to help maintain confidence in the U.S. banking system.

Over the past several decades, Congress has revised the laws governing the resolution of failed banks on several occasions. However, two of those revisions are crucial to today's discussion. First, Congress required the FDIC to use the "least costly" transaction for resolving insured depository institutions. Second, Congress created depositor preference, which gives depositors a priority among unsecured creditors. Both reforms were designed to reduce losses to the DIF.

Unfortunately, H.R. 5823 would restrict the FDIC's current receivership authorities used to maximize the value of the failed bank's covered bonds. The bill leaves the FDIC with only two options: continue to perform until the covered bond program is transferred to another institution within a certain timeframe; or hand over the collateral to a separate trustee for the covered bond estate, in return for a residual certificate of questionable value. The FDIC would not have the authority - which it can use for any other asset class - to repudiate covered bonds, pay repudiation damages and take control of the collateral. This restriction would impair the FDIC's ability to accomplish the "least costly" resolution and could increase losses to the DIF by providing covered bond investors with a super-priority that exceeds that provided to other secured creditors. These increased losses to the DIF would be borne by all of the more than 8,000 FDIC-insured institutions, whether or not they issued covered bonds.

Limiting the time in which the FDIC could market a covered bond program to other banks will constrain the FDIC's ability to achieve maximum value for a program through such a transfer. Similarly, preventing the FDIC from using its normal repudiation power will prevent the FDIC from recapturing the over-collateralization in the covered bond program. The 'residual certificate' proposed in H.R. 5823 is likely to be virtually valueless. More importantly, the legislation would provide the investors with control over the collateral until the term of the program ends, even though the FDIC (and any party obligated on a secured debt) normally has the ability to recover over-collateralization by paying the amount of the claims and recovering the collateral free of all liens. Providing the FDIC a residual certificate instead of the ability to liquidate the collateral itself would reduce the value to the receivership estate and would not result in the least costly resolution.

So long as investors are paid the full principal amount of the covered bonds and interest to the date of payment, there is no policy reason to protect investment returns of covered bond investors through an indirect subsidy from the DIF. However, some market participants have argued that continuing to allow the receiver to exercise its statutory repudiation authority would reduce investors' interest in U.S. covered bonds due to the reinvestment risks. This argument misses the mark both from the perspectives of equitable risk allocations and real financial risk.

As discussed earlier, if there is reinvestment risk, it should be borne by private investors, not the public sector, other creditors, or the DIF. Covered bond investors should receive full payment for the face value of their bonds plus interest. However, they should not be guaranteed control of the cover pool where it vastly exceeds the actual amount of their claims. In addition, there is no real financial risk if the FDIC repudiates the covered bond transaction, pays the full value of the outstanding bonds, plus interest, and takes control of the cover pool. If that happens, it simply means that the investors' trustee has a pot of money to reinvest into a guaranteed investment contract - like an annuity - to continue to pay investors the steady stream of bond payments which they are seeking.

The financial returns for the investors will not be different, in any meaningful way, from the return they could expect if they had been able to seize control of the cover pool as H.R. 5823 allows. The reason is that, once seized, the cover pool becomes a static pool with no new loans entering, but with delinquent and paid-off loans exiting. Like a static securitization pool, it will be a diminishing pool of collateral as these loans exit. In addition, like other pass-through investment vehicles, the amount of cash generated in any period can be highly variable because of delinquent or missed payments, prepayments and payoffs. A mismatch will occur between the bond payment obligations and the remaining cash flows of the cover pool. This mismatch would result in early prepayment of the covered bonds to maintain parity. To the extent investors put in place contingent liquidity and/or credit support mechanisms to reduce the asset/liability mismatch, they also reduce the internal rate of return on the covered bonds or increase the cost of issuance to the financial institution. There would also be administrative or management fees associated with the management of the pool. Finally, investors of a static pool pass-through would be subject to default risk, which would be eliminated by the payment in full of the covered bonds. The net economic consequences of the early redemption of the covered bonds would be roughly equivalent to the cost of managing the assets to the covered bond's maturity. However, by giving the FDIC the option to redeem the covered bonds, this cost would not be subsidized by the DIF.

The protections to the insurance fund, depositors and the flexibility afforded the FDIC as receiver of a failed depository institution has become a standard that other countries want to emulate. The flexibility that Congress afforded the FDIC permits us to respond to market conditions at the time of insolvency and to achieve bank resolutions that protect insured depositors at the least cost to the DIF. This is an important public policy that we believe has served the nation well and should be maintained.

Legislation Should Not Create a "Super-Priority" for Covered Bond Investors - Under U.S. law, secured creditors are entitled to payment of their claims before unsecured creditors up to the lesser of the full amount of their claim or the value of the collateral. We should avoid upsetting this settled principle of law - which is enshrined both in state commercial law under the Uniform Commercial Code and in federal and state insolvency law in the Bankruptcy Code and the FDI Act, among other statutes.

Covered bonds do offer some advantages over securitization towards improved underwriting. The potential for improved alignment of the bank's incentives toward better quality underwriting is a consequence of the loans remaining on the bank's balance sheet, the duty to replace any delinquent loans in the cover pool, and holding capital for the loans in the pool. However, these advantages come at a cost. The obligation to replace delinquent loans means that there is a continuing demand for new originations, which can act as a liquidity drain if delinquencies increase. This also means that, as poorer loans are taken out of the cover pool, the remaining balance sheet will consist of more and more delinquent loans. In a receivership, this can lead to greater losses to the DIF - particularly if the FDIC's options to sell the covered bond transaction are restricted.

Clearly, strong origination standards will continue to be required. The potential stress on issuing banks is illustrated by Washington Mutual Bank, which had to increase the cover pool to almost 150 percent over-collateralization in a failed effort to maintain high ratings for the transaction. This further exacerbated Washington Mutual's asset and liquidity problems.

This example also illustrates another important consideration in covered bond legislation - investors should not be completely shielded from investment risk and their risk should not be transferred to the public sector or to the DIF. If, as under H.R. 5823, the investors can seize the entire cover pool for the duration of the covered bonds irrespective of the degree of over-collateralization, it will provide a strong incentive for investors to maximize the over-collateralization. Naturally, this will increase pressure on the issuing bank during periods of stress. The ability of investors to seize the entire cover pool will also further reduce the loan assets available for sale by the FDIC in any receivership. If creditors of covered bonds are shielded from all risks, there is a strong possibility that covered bonds could lead to a mispricing of risk and distortions in the market, imperiling banks in the future. On the other hand, if the long-standing treatment of secured creditors is maintained - which would allow the FDIC to pay the outstanding principal and interest on the bonds and recover the over-collateralization - there will be very limited incentive for the creditors to demand increasing levels of collateral as a bank becomes troubled.

The super-priority given covered bond investors by H.R. 5823 also runs against the policy direction established by Congress in recent legislation. In 2005, Congress enacted Section 11(e)(13)(C) of the FDI Act, which prohibits secured creditors from exercising any rights against any property of a failed insured depository institution (IDI) without the receiver's consent for the first 90 days of a bank receivership.1 This provision prevents secured creditors from taking and selling bank assets at fire sale prices to the detriment of the receiver and the DIF. More recently, section 215 of the Dodd-Frank Wall Street Reform and Consumer Protection Act mandates a study to evaluate whether a potential haircut on secured creditors could improve market discipline and reduce cost to the taxpayers. This study was prompted by the recognized roles that the run on secured credit and the insatiable demand for more collateral had in the financial crisis of 2008. In contrast, the unprecedented protection for one form of secured creditors - covered bond investors - in H.R. 5823 runs counter to the policies underlying these provisions.

A further concern created by H.R. 5823 is that it could encourage covered bond transactions that include "triggers" for early termination or default before a bank is closed by the regulators. Under H.R. 5823, a separate estate, which removes the entire cover pool from the bank's control, is created upon any event of default. Once created, the separate estate and all collateral in the cover pool would be outside the control of the FDIC, as receiver for the bank. The residual value of the pool, and all of the loans, would be outside the receivership and be lost for all other creditors of the failed bank. This additional special protection creates a strong incentive for covered bond transactions to include a trigger that acts before the bank is placed into receivership. Since such a trigger would deprive the bank of the cash flows from the cover pool and signal to the market its imminent demise, the bank would almost inevitably suffer a liquidity failure. As a result, these early triggers represent another source of increased loss to the DIF.

The FDIC has recommended that the receiver should have the authority to cure any defaults under the covered bond transaction within 30 days of the appointment of the FDIC as conservator or receiver of an issuer. This would reduce the incentive for covered bond investors to declare a default and take control of the cover pool in anticipation of an FDIC receivership. Providing the FDIC 30 days to cure a default would allow the FDIC to recapture the value of the overcollateralization in the program for receivership creditors, including uninsured depositors and the DIF. The FDIC would then have the same options to resolve the covered bond transaction and maximize the value of this asset in the receivership.


The FDIC supports a vibrant covered bond market that would increase liquidity to financial institutions and enable sustainable and robust asset origination. However, any legislation should avoid promoting development of a covered bond market by reducing market discipline and protection for the Deposit Insurance Fund (DIF). We believe the principles, described above, will ensure that covered bonds serve as a sustainable investment for bondholders and the financial system. We will continue to work with the Congress, other regulators and market participants on ways to create a sustainable covered bond market in the U.S.

Thank you for inviting me to appear at this hearing. I will be happy to answer any questions.

1 The only exception to the stay in 11(e)(13)(C) is for qualified financial contracts (QFCs). This exemption is based on the fact that performance of the derivatives markets requires prompt transfer or closeout of derivatives positions, thereby reducing potentially negative systemic effects of counterparty failures. Covered bonds do not meet the definitions as QFCs. Nonetheless, H.R. 5823 gives covered bond investors a right to retain all collateral that not even secured parties with QFCs receive.


Last Updated 9/15/2010

Skip Footer back to content