Resiliency of Canadian Covered Bonds
In the current world of sharply rising interest rates and a possible recession, questions have arisen about the resiliency of the cover pools for Canadian covered bonds. Additional focus is brought to this question by the declining housing values in Canada — Vancouver and Toronto in particular have seen reported 15 to 20 per cent declines in house prices from recent peaks.
Canadian covered bonds take their strength from several factors, not the least of which is the conservative nature of property investors in Canada, combined with fairly strict underwriting standards set by OSFI. This has resulted historically in a typical annual loss rate for residential mortgage pools for most banks in basis points in the single-digit or low double-digit range.
This low loss experience is supported by the full recourse nature of Canadian mortgage loans. A mortgagor under Canadian law is personally liable for full payment of the mortgage loan if the loan is foreclosed on and liquidated at a loss. Unlike the case in many U.S. States, a Canadian property owner cannot turn over the keys to house and walk away free of the debt.
Another protection for Canadian cover pools is the monthly Asset Coverage Test that each program must pass. If the value of eligible mortgage loans in the cover pool does not exceed the outstanding amount of covered bonds by the required overcollateralization amount, the test is failed. Defaulted mortgage loans are not included in test. If the test is failed, the issuing bank is required to transfer additional, non-defaulted eligible mortgage loans to the cover pool. Thus, the cover pool is constantly refreshed with performing mortgage loans protecting the value of collateral backing the covered bonds.
More protection is provided by the requirement that an eligible mortgage loan for Canadian cover pools must have a loan to value ratio not exceeding 80%, measured each month based on an index of current property values in the location of the property. If the loan to value ratio exceeds 80% at any time, only only the portion of the loan not exceeding 80% of the value of the property is included in the cover pool for the calculation. This means that the value of the cover pool is protected from declining property values.
Moreover, the typical average loan to value ratio of mortgage loans in cover pools for Canadian covered bonds is between 50% and 60%, which provides a substantial buffer before loan amounts are reduced in the cover pool because they fail the loan to value maximum for eligibility. Statistical information on cover pools is available in the monthly report provided to investors by each of the Canadian banks.
Lastly, in addition to strong cover pools, investors in Canadian covered bonds hold exposures to banks that operate in a conservative banking environment. Canadian banks came through the financial crisis of 2008 in excellent shape and continue to be highly regarded in international capital markets. Banking regulation in Canada contributes to the conservative environment with a with a regulatory approach that prioritizes stability. The recent tightening of mortgage loan underwriting evidences this caution.
This collection of protections is what supports the perception of quasi-sovereign risk for Canadian covered bonds.
Why is CB legislation tied to GSE reform?
Why is the adoption of covered bond legislation linked to housing finance reform? Housing finance reform is all about the role of the GSEs. While covered bonds certainly can be used to finance residential mortgage loans, they do not require any form of government support. The consideration of the proper role of the government in housing finance can occur independent of covered bonds. However, I hear from many sources that covered bond legislation would only be considered after GSE reform had been adopted or perhaps considered with GSE reform.
There is no apparent logic to this position. Covered bonds are a private sector financing technique that has proved very effective in other jurisdictions. There is nothing in GSE reform that would be a necessary predicate to the issuance of covered bonds by U.S. banks. Covered bond legislation would not touch the status of the GSEs. It is possible that covered bond issuance by U.S. banks could develop into an attractive alternative to financing through the GSE and thus reduce the tension in GSE reform, but that would be beneficial to GSE reform.
It seems as though both sides are determined to keep as much pressure on GSE reform as possible in order to achieve their objectives and not permit any private sector initiatives to sidetrack the discussion until the role of the government in housing finance has been solved. But this seems to put the cart before the horse. Shouldn’t the government intervene only where the private sector is not functioning properly? Wouldn’t it make sense to let private sector initiatives develop first before assigning the government a role? If we can agree that the answer to those two questions is yes, why not adopt covered bond legislation and see how the market develops while we debate how to wind down the GSEs and what would be the appropriate future structure for the government’s role in housing finance?
Certainly we can have a fulsome debate on how the government can support housing access for those who need assistance independent of how covered bond legislation is drafted. Certainly if covered bonds, RMBS and the federal home loan banks fail to provide adequate private sector funding for residential mortgage loans there may be a need to consider a larger government role.
It is not essential that covered bonds be enabled through legislation as it is possible to achieve covered bond issuance through securitization techniques, as has been done in other countries. See, e.g., Time for a US alternative. However, investors will have more confidence in a covered bond sector established through legislation and the market may be expected to develop quicker with legislation. Enacting legislation for covered bonds would be a low cost experiment that would have no harmful side effects. Covered bond legislation, therefor, should be enacted before GSE reform is attempted so that we have a better chance to assess what works in the private sector before designing the government’s role in housing finance.
Why do investors like covered bonds?
Why do investors like covered bonds?
What is it about covered bonds that investors like? Even U.S. investors who have plenty of other fixed income investment opportunities. In Europe there is more than €2 trillion in covered bonds outstanding. Even in the U.S., which has no statute to enable its banks to issue covered bonds, there is $150 billion of covered bonds outstanding.
What is it about these bonds and who is buying them?
You can find out who the investors are at The Cover or The Covered Bond Report. Both publications provide a breakdown of type and location of investors by offering. And what the data shows is that banks and central banks are between 50% and 80% of the investors, depending on the offering. The remainder goes to funds, asset managers and insurance companies. While not a distinct class of investors, the composition is quite different from the class of investors in bank senior debt or securitizations. Why? And why central banks? Aside from QE, of course.
The answer lies in the nature of covered bonds. Covered bonds are a different kind of investment. They are more than senior bank debt because there is recourse to the cover pool. They are more than securitization because there is recourse to the issuing bank. Covered bonds are a dual recourse instrument, which raises an investor’s confidence in their safety. Importantly, in Europe covered bonds are not subject to bail-in, while senior debt is.
And in Europe, covered bonds receive favorable capital treatment under the bank capital rules, attracting only half the capital that a senior bond from the same institution would attract. But that only makes sense given the dual recourse nature of covered bonds compared to senior debt. And of course central banks are not subject to the capital rules anyway.
So what appears to attract banks and central banks and other investors is the high level of safety with covered bonds combined with a yield that exceeds similarly rated sovereign debt. And covered bonds have a similar risk profile – no defaults in 250 years. Quite a record.
There are other details about covered bonds that are also considered important.
Covered bonds are issued by regulated financial institutions and the covered bond programs of the institutions are separately regulated.
The quality of the assets in a cover pool is high and subject to regulation. Any assets that default or become delinquent must be replaced on a monthly basis. The bank has 100% “skin-in-the-game.”
The bonds are simple, bullet pay instruments with either a fixed or floating rate. If the issuing bank were to become insolvent, the assets in the cover pool are intended to continue payments on the bonds through their maturity. No pre-payment risk.
Each series of covered bonds is a single class, so there is no complex class structure and complex payment waterfall to analyze. The credit analysis is primarily an analysis of the strength of the issuing bank and for this there is a huge community of analysts to assist an investor and a wealth of analytical experience covering more than 100 years of corporate credit analysis. The lack of experience and analytical talent was one of the prime failings of securitizations leading up to the crisis.
So what’s not to like? A risk profile like sovereign bonds and a better yield. No wonder there is a €2 trillion market. But will the United States Congress like them?
Why not covered bonds?
Why not covered bonds?
Why not pass legislation for covered bonds in the United States? It is easy to do and there is basically no cost to the Treasury. In 2011, covered bond legislation passed the House Financial Services Committee by a vote of 44 to 7, a very strong bi-partisan vote. The only dissenters were hoping to implement provisions requested by the FDIC that were rejected by the majority. The dissenters were unable to retain even members from their own party on the final vote. Probably more than any other development, this demonstrates that covered bonds are not a partisan concept, they are not divisive and they have broad support.
Covered bonds will bring private funding to residential mortgage loans, but there is no good reason that passage of covered bond legislation should be tied to GSE reform. There is nothing about covered bonds that would implicate GSE reform, except that by bringing private funding to the market they could reduce the dominate role of the GSEs. And there is no credible evidence that either party in Congress believes that a continued dominate role for the GSEs is a policy mandate.
An unlike securitization of residential mortgage loans, no concerted effort is needed to get investors to participate. Investors are wary of residential mortgage securitization as a result of the financial crisis, which was precipitated by mortgage securitization. There are many efforts underway to convince investors to return to the RMBS market. Covered bonds carry no such baggage. Covered bonds have been readily sold to U.S. markets since 2010 and investors remain eager for more.
Covered bonds are more transparent to regulators.
Covered bonds are a simpler investment analysis for investors.
Covered bonds are proven financing technology with worldwide acceptance.
Covered bonds are friendlier to mortgage borrowers because the originator retains the right to modify loans to assist borrowers in working out difficult loans.
What’s to lose? Maybe covered bonds could develop into a vibrant private sector funding alternative for residential mortgage loans. If not, no loss. And no cost. But if it works, it’s a plus all around.
Use CBs to Restart the PLS Market
In the United States there is a lot of discussion about how to reduce the taxpayers’ exposure to housing finance. The conservatorships of Fannie Mae and Freddie Mac were a pointed reminder of the extent of the government’s, and the taxpayers’, role in housing finance. Additionally, after being taken into conservatorship, the GSEs were intentionally used to prevent a feared collapse of the housing market, further expanding their role.
As a result, the GSEs, together with FHA and GNMA, at one point provided more than 95% of the financing for new residential mortgage loans, and today they still finance roughly nine out of every ten new mortgage loans. At the same time, the private label securities (PLS) market, which provided significant private funding for residential mortgage loans, essentially vanished. The issuance of new PLS declined nearly 99% in the years following the crisis.
Today there is general agreement across the political spectrum that it would be desirable to restore some level of private funding for residential mortgage loans and reduce the role of the GSEs. There are also plans to reform the GSEs and refine their mission in housing finance. There is less agreement on the reform proposals.
On June 27, 2014, Treasury Secretary Jacob Lew requested comment on what steps could be take to restore a functioning PLS market. Almost uniformly the commenters voiced the need to shrink the presence of Fannie and Freddie in the market in order to make room for a PLS market. This reflected the view that under the existing authority of the GSEs, the private market would not be price competitive and would only be able to finance a small number of loans that did not qualify for financing through the GSEs under their very high loan limits.
However, the presence of the GSEs is not the only factor impeding the development of the PLS market. The regulatory landscape for securitizing mortgage loans has been altered dramatically since the crisis. Basel III imposes heavy, in some cases almost punitive, increased capital requirements on banks and most particularly for securitization exposures held by banks. The risk weighting for securitization exposures can range up to 1250% and in some cases, such as non-cash gain on sale, the capital requirements will result in a deduction from Tier 1 equity capital. Rulemaking by the Consumer Financial Protection Board exposes any holder of a mortgage loans to severe penalties for those loans not underwritten in accordance with the CFPB underwriting criteria, which are designed to protect the borrower. SEC requirements for mortgage securitization require 270 data points of information to be disclosed for every mortgage loan in a securitization, both at the time of the offering and in periodic reporting for the life of the securitization. Then there is risk retention, prohibited conflicts of interest, margin requirements for swaps, the Volcker Rule, conflicting requirements in foreign securities markets and a litany of other burdens on the PLS market.
Comment Letter to Treasury on Restoring PLS Market
On August 8, Morrison & Foerster LLP filed a comment letter with the United States Department of the Treasury in response to a request for comments on the private sector development of a well-functioning private label securities (PLS) market for residential mortgage loans. The comment letter notes that perhaps the easiest way to restore private funding to residential mortgage loans in the United States is to implement a covered bond statute in the United States to enable U.S. banks to issue covered bonds. The letter notes that the conditions necessary to the establishment of a covered bod market in the United States are well advanced and that the market could be established quickly. Based on the development of the investor base in the United States, the SEC’s establishment of disclosure and reporting standards, and the well-developed legislation, creation of a domestic covered bond market would appear to be low hanging fruit that Treasury should take advantage of.
EBA report on favorable capital treatment
On July 1, 2014, the European Banking Authority published a favorable opinion on the preferential capital treatment of covered bonds. The EBA also delivered a companion report entitled EBA Report on EU Covered Bond Frameworks and Capital Treatment. See also the press release from the EBA. The EBA concluded that the favorable capital treatment under the capital requirements regulation (CRR) for bank investments in covered bonds was appropriate, but called for some “further qualifying criteria for their preferential treatment.” The EBA recommended against the use of aircraft liens, residential mortgage backed securities (RMBS) and commercial mortgage backed securities (CMBS) as cover pool assets after December 2017. The opinion and the report were delivered to the European Commission in response to a request for advice.
With respect to favorable capital treatment the EBA said:
“Due to the good historical default/loss performance of covered bonds in the EU, the dual recourse principle embedded in covered bond frameworks whereby the covered bond holder has a claim on the issuing institution and a priority claim on the cover assets, the special public supervision for the protection of the bondholders mandated by the UCITS Directive and the existence of qualifying criteria in Article 129 of the CRR, the EBA considers the preferential risk weight treatment laid down in Article 129 of the CRR to be, in principle, an appropriate prudential treatment. “
The Case for U.S. Covered Bonds
Updated: 5/18/2015
Two U.S. banks issued covered bonds in 2006 and 2007. No U.S. banks have issued covered bonds since, due in part to the complexity of issuing structure used by U.S. banks. Legislation to implement covered bonds in the U.S. would change this. The benefits of covered bonds to both the issuing bank and to investors is set out below. These benefits would seem to provide a strong rationale for passing covered bond legislation.
The covered bond market in Europe is about $3 trillion of outstanding bonds. This suggests that covered bonds could provide a significant funding instrument for banks in the U.S. While no one expects covered bonds to provide all of the funding for residential mortgage loans in the U.S., given the historic role of the federal government in the market, covered bonds could be a significant factor. The combination of covered bonds, RMBS, FHLB loans and GNMA/FNMA/FHLMC would provide banks with improved diversity of funding.
What are covered bonds?
Covered bonds are senior, secured debt of a regulated financial institution. As in typical secured debt, if the issuing bank defaults, the collateral is used to cover any shortfall in payments due by the bank on the covered bonds. With covered bonds there is the added feature that, so long as there is sufficient collateral, the covered bonds are not accelerated, but rather paid on their scheduled payment dates. If the collateral at any time is inadequate to make all scheduled payments on the covered bonds, all outstanding covered bonds are accelerated and paid pro rata from the proceeds of the collateral in the same manner as typical secured debt. For a detailed description of covered bonds, see Covered Bonds Handbook available from the Practicing Law Institute. See also the Covered Bond website at Mayer Brown LLP.
Covered bonds come in differed forms, depending on the jurisdiction of the issuer. In some jurisdictions, such as Germany, the collateral is held by the issuing bank and is isolated on its balance sheet and pledged to support the bonds. In other jurisdictions, such as the U.K., the collateral is transferred to a subsidiary of the issuing bank and the subsidiary guarantees bonds issued by the bank and secures the guarantee with the collateral. (See the accompanying map of the covered bond market, provided by Global Capital, for a list of jurisdictions with covered bonds.) The subsidiary, in effect, is only a security device to hold the collateral separate from the issuing bank in case of its insolvency. This two-tier arrangement is necessary in jurisdictions that do not have a special statute for covered bonds.
The collateral and any related interest rate and currency swaps are referred to as the “cover pool.” For the protection of investors, the adequacy of the assets in the cover pool to pay the covered bonds as scheduled is tested monthly. If the test is not passed, additional collateral must be added to the cover pool by the issuer.
Most often the collateral consists of residential mortgage loans, but some jurisdictions permit commercial mortgage loans, ship mortgage loans, and obligations of public sector entities.
In the single tier form, it is clear that the collateral continues to be owned by the issuing bank and is simply pledged to support the bonds. In the two-tier structure, the collateral is transferred to the subsidiary, but the bank continues to have an economic interest in the performance of the collateral. The collateral held by the subsidiary will be consolidated back onto the balance sheet of the issuing bank and losses on the collateral will be losses, on a consolidated basis, for the bank. Thus the bank continues to own the collateral as an indirect owner.
Covered bonds, then, are dual recourse instruments – the holder will look first to the bank for payment and, if the bank is unable to pay, the investor will look to the collateral. Thus the creditworthiness of the bank is of primary concern to investors and the collateral is of secondary concern. See the slides presented at ASF 2014 for more detail on structure.
Investor diversity.
The investors who purchase covered bonds typically purchase sovereign debt and agency debt. They do not typically purchase corporate bonds or ABS securities. These investors seek investments with low risk of principal acceleration due to issuer default or insolvency. The certainty of payment at, but not before, maturity is an important feature. Accordingly, this is a class of investors that an issuing bank does not reach except by issuing covered bonds, providing important diversity of funding for the issuing bank.
The investor base is comprised primarily of banks, central banks and investment funds. Banks and central banks together are often 50% to 70% of the investor base of an offering.
Excellent funding rates.
This conservative class of investors is seeking a return that exceeds that of sovereign or agency debt, but with a similar risk profile. This permits the issuing bank to fund itself with covered bonds at very attractive rates. In fact, other than in times of severe turmoil, covered bonds are viewed as a ‘rates’ product and not a ‘credit’ product, similar to sovereign and agency debt.
For example, TD Bank in 2012 issued $3 billion of 5 year covered bonds at mid-swaps plus 45. RBC in 2013 issued $2 billion of 5 year covered bonds at mid-swaps plus 43. This is very attractive funding. BNS in 2014 issued €1.250 billiion of three year covered bonds at mid-swaps minus 4.
Transparent to regulators.
From a regulator’s perspective, a bank’s exposure on its covered bonds is simpler to analyze than its securitization exposure. As the numerous recent litigation settlements by banks has shown, securitization exposes banks to the risk of significant cost that is not apparent in the financial statements or other disclosures of the banks. JPMorgan, for example, in a single settlement with the Justice department and various State attorneys general, recently settled one set of claims for $12 billion. The aggregate settlement amount to date in the U.S. for the banking industry in early 2014 was estimated to be around $100 billion and there is more to come. This exposure was not apparent to regulators pre-crisis. Securitization had been viewed by banks and regulators as a transfer of risk of the assets to investors, relieving the banks of exposure to the assets.
Covered bonds do not present this hidden risk to regulators because the assets in the cover pool remain on the balance sheet of the issuing institution. The nature and performance of the assets is a constant audit and financial reporting item, readily apparent to the supervising agencies. Because the issuing institution retains 100% of the risk on the assets, it has a strong incentive to monitor and maintain high origination standards. This incentive tends to align the interests of the banks and the regulators in a way that securitization never will.
Friendly to borrowers.
Covered bonds are friendlier to borrowers than securitization. In a securitization, loans are sold to a securitization vehicle; the selling bank ceases to be a contractual party to the loans. As a consequence, the securitization entity acquires the right to make any and all decisions with respect to granting payment rescheduling or other relief to a borrower in financial difficulty or foreclosing on property securing a loan. Securitization entities are bound typically by a trust agreement or pooling agreement that spells out all actions the entities may take with respect to the loans. These provisions generally do not provide for discretion to be exercised by the securitization entities to work with borrowers to avoid default and foreclosure.
The transfer of a loan to the securitization entity means that a borrower can no longer discuss with his lender alternatives that the lender might offer to avoid default. The lender no longer has the power to effect any relief for a deserving borrower. That power has been transferred to a securitization entity and the securitization entity is typically bound by agreements not to provide such relief.
Additionally the borrower often has difficulty determining who may have some ability to grant relief. The borrower’s loan may have been transferred through several entities before reaching the securitization entity. And securitization entities do not have employees. All actions of securitization entities are performed under contract by third parties including trustees and loan servicers. Consequently, even identifying the party that may have the power to grant relief can be hard. And if identified, locating the appropriate person at the third party can be yet another difficult task for a borrower.
Covered bonds do not create this obfuscation. Each loan continues to be owned by the originating lender, who retains all rights to amend the loan or the payment terms or otherwise accommodate a borrower to preserve a performing loan. It is possible that the loan may cease to be qualified to be included in a cover pool, but in that case, the lender simply substitutes another loan for the loan being worked out. And from the borrower’s perspective, the borrower always deals solely with the lender that the borrower originally chose to borrow from and should have little difficulty identifying who in the organization might be able to provide relief.
Analytically simpler for investors.
Covered bonds also provide simplicity for an investor. In contrast, a securitization presents a considerably more complex investment decision. Securitizations typically involve complex security class structures that result in complex payment provisions. An RMBS offering may involve 20 or 25 classes of securities supported from a single pool of loans. And the class structures and payment mechanisms and priorities are very seldom the same on consecutive offerings. Accordingly, on each offering an investor must commit significant resources to analyzing the precise terms of the proposed investment, in addition to analyzing the performance risks of the collateral pool, which will be unique for each offering. And finally, the risk of early or delayed principal repayment must be analyzed, which risk can be critical to assessing the value of each class of securities.
Covered bonds, on the other hand, provide a simpler challenge to the investor. A covered bond is primarily a senior debt offering of a regulated financial institution. The institution is usually reviewed by several investment industry credit analysts, files extensive financial statements with regulators and exchanges, and has listed common stock that has a long history of public pricing. The institution is audited regularly by independent auditors and supervised by regulators. Moreover the senior debt of the institution is rated by rating agencies and traded in the secondary market, providing important pricing transparency and indications of continuing credit evaluations.
Only if the issuing institution defaults on its covered bond debt does the strength of the cover pool become important. Until then the issuing institution is bound to continually refresh the cover pool with new loans to replace delinquent, defaulted or matured loans. At all times prior to the default of the institution, the cover pool should consist of fully performing loans.
And while the cover pool may vary over time as new loans are added to the pool, the investor’s burden of analysis is considerably simpler than in a securitization. First, all outstanding series of covered bonds of an issuing institution are secured by the same cover pool. New assets can be added to the cover pool only if they satisfy the eligibility criteria that are set when the covered bond program is first established. Thus an investor analyzes the strength of the cover pool once at its first purchase of a covered bond from an issuing institution based on the eligibility criteria and can purchase subsequent offerings by conducting a simplified ‘bring down’ analysis of continuing loan performance.
Moreover, covered bonds are not tranched, so there is no complex class structure and payment mechanism or priority. And, because covered bonds are bullet pay securities, there is no prepayment risk to analyze. Covered bonds are primarily a payment obligation of the issuing institution, so the credit analysis is primarily an analysis of the credit worthiness of the issuing institution. This is a risk that is publicly reviewed by other analysts and supervised by regulatory authorities.
And, finally, the risk attendant to senior debt of being written down, or possibly eliminated or converted to equity, in the insolvency of an issuing institution is not present with covered bonds because the cover pool will continue to pay the covered bonds through their maturity date independent of the outcome of the resolution of the failed issuing institution.