Last month Asset Securitization Report met with dealmakers to talk about the future of covered bonds in the U.S.
Covered bonds have their roots in late 18th-century Germany, and they've become an important finance tool for institutions in Europe and Canada. In the U.S., however, only two U.S. issuers have tested the waters: Washington Mutual and Bank of America.
But the sector is gaining more ground as regulators and dealmakers in various pockets of the financial services industry look to encourage the use of this financing tool.
The credit crisis has also spurred more conversations about covered bonds on Wall Street and in Washington. And some observers believe they can be a handy when securitization of nonconforming mortgage debt, such as jumbo loans, is still tough to pull off.
So is the U.S. ready to embrace a European import?
It might have to, if credit is to flow freely throughout the housing industry. As what happened with mortgage passthroughs in the 1980s, the introduction of covered bonds will require changes in legislation and regulatory framework, as well as investor education.
Investment Dealers' Digest Editor-in-Chief Aleksandrs Rozens moderated the roundtable.
ROZENS:Mr. Arnholz, we have legislation in place that may support the creation of a covered bond market. What can you tell us about the proposed legislation and what are the prospects for actually getting it passed?
ARNHOLZ: The bill has yet to be voted on by Congress and is sponsored by Republican House member Scott Garrett. It is an extremely comprehensive piece of legislation and is designed to provide a complete legal framework for the issuance of covered bonds.
The bill replaces earlier versions that were considered too narrow and follows the European model in most respects. I think it's fair to say that it does three things. First, it contemplates the issuance of bonds that are secured by a covered pool and issued by an eligible issuer. Second, if an issuer defaults, or becomes insolvent, the cover pool is severed from the estate of the issuer through the scheduled maturity date of the covered bonds. And third, the bill establishes a covered bond regulator, which was a feature that the market had expressed a real interest in seeing.
The bill provides a broad definition of an eligible asset. It includes residential mortgage loans, commercial mortgage loans, municipal securities, SBA loans. It's a very liberal definition. But mixing assets in one deal is not permitted.
Interestingly, the definition of an eligible issuer is broad as well and it includes not only banks, but nonbank, financial companies.
The key to the bill is that it provides clarity to investors in the event of a bank failure. The bill's approach is extremely simple in this regard. If a bank fails and the program is not transferred to another solvent issuer within a specified time, bondholders are effectively given control of the cover pool, including any overcollateralization that exists at the time of default.
It's very clear, and I think that's the merit of the legislation. What are its prospects? There is no Senate version yet and obviously there's going to be nothing done during the lame-duck Congress. But the prospects are reasonable that something will get done in the new year when Congress reconvenes.
ROZENS:In terms of anybody putting together covered bond transactions, are you hearing anything being put together, either publicly or privately?
ARNHOLZ: I don't think any U.S. issuer is about to initiate a program and jump in the market any time soon. That's true for a combination of reasons, but principally because the market is waiting to see what happens in Congress.
Now, Mike Krimminger may disagree with me, but I don't think that investors were fully satisfied with the [Federal Deposit Insurance Corp.'s (FDIC)] policy statement in 2008. I would point out that there clearly is investor appetite - more than $18 billion in dollar denominated covered bonds have been sold in the U.S. this year, mostly Canadian product, and I'm told there was very strong demand.
ROZENS:Do you think that this financial tool will ever replace securitization?
ARNHOLZ: It's hard to say. The securitization structures held together very well in the meltdown. What failed was the asset quality. I think the question is whether or not covered bonds will be able to generate investor confidence or whether residential mortgage backed securities will again generate confidence that brings investors back to the markets.
ROZENS:Given that we have this foreclosure problem that's affecting a lot of lenders and everybody on the underwriting side, will this hasten the development of a covered bond market or will it make covered bonds a viable alternative to securitizations?
ARNHOLZ: It could. The foreclosure mess has, I think, eroded investor confidence further and if covered bond legislation is successful in restoring confidence, I think it might.
I would point out that the same servicers that are involved in the current foreclosure situation are also most likely going to be the servicers of the covered pools, on covered bond offerings.
ROZENS:Before we go into the nitty gritty of a covered bond, I was wondering if the issuer files for bankruptcy or is seized by regulators, who is actually monitoring the collateral? Are there servicers who manage the loan pool?
ARNHOLZ: Under the Garrett bill, a covered bond regulator is appointed and the covered bond regulator, either by itself or through a trustee, would manage the collateral at that point.
ROZENS:Mr. Krimminger, can you tell us what has been done legislation wise to facilitate covered bonds? Can you offer a perspective from your seat as to what else needs to be done to get this market going?
KRIMMINGER: The first thing I will do is try to give a little bit of context to the FDIC's involvement with covered bonds in the past.
One of things that we did back in early 2006 was we worked very closely with the two issuers who had been issuing covered bonds in the U.S., Washington Mutual and Bank of America, to try to work through how the FDIC process works in terms of risk management when a bank fails.
The second thing that we did in 2008 is we worked very closely with the market place and also with the U.S. Treasury to develop a covered bond policy statement that was issued in July of 2008, and the Treasury followed that with their best practices or gold standards for covered bonds in August of 2008.
I'll defer to John on the complete market satisfaction, but I think there was a feel in the market at that point, we were advised, at any rate, that this was a structure under which issuers could issue covered bonds. Of course, September of 2008 happened shortly thereafter, so again there were problems and we couldn't issue anything into the structured finance market.
So we haven't seen any issuance since then. I think one thing we also need to remember with regard to covered bonds in the context of the overall U.S. mortgage and structured finance markets is that there are a lot of sources for funding in the U.S., such as Federal Home Loan Bank advances.
Of course, we don't know how that process will work out going forward. So we need to make sure that as we begin to fashion a legislative process for covered bonds, we take it step by step and take into consideration how the U.S. financing and structured finance market has worked in the past.
I think it's quite different from Europe and always has been. There are very different aspects to European structures that we need to consider if we're fashioning a covered bond legislation.
For example, let's be quite honest, in Europe typically banks don't fail. That's not the approach that we want to have here in the U.S.
Typically, there's also a variety of different structures under which covered bonds are regulated in Europe, some of which are very close to the proposed legislation here in the U.S., and some that differ quite a bit. And I think there's not a one-size-fits-all, which is my main point on that.
Second, another thing on the covered bond market in Europe is that mortgages are quite different in Europe. There's much more in terms of required prepayment penalties and a shorter rate in terms of the somewhat different structures for the mortgage market.
Generally, as we approach the overall housing finance structure in the U.S., and covered bonds are part of that, we just need to make sure to see how it fits.
As far as legislation, the FDIC does support covered bond legislation. We think that the FDIC statement of policy framework could be sufficient but we do not have any objection at all to legislation that really accomplishes three key principles.
First, it should clarify the rights and responsibilities of investors, issuers and regulators. Second, we think it should not transfer investment risk from covered bond investors to the Deposit Insurance Fund. As an aside, one of the messages from the Dodd-Frank Act is that we want to make sure there's not a perception or an actuality of government support for financial institutions, and the private sector should be the primary driver of the financial markets.
Third, we think covered bond legislation should be consistent with long standing U.S. law and policy for creditors. I think that's an area that we've had some concern about in the proposal from Rep. Garrett.
I think there's a lot of merit to the proposal; it's certainly evolved over time to address some concerns that we've had, but we want to make sure that we maintain the protections of the Deposit Insurance Fund that's been enshrined under the Federal Deposit Insurance Act for many years and felt protected and secure under the Federal Deposit Insurance Act for banks and others.
In particular, I would note that when a bank closes, we have an obligation under the statute to get the maximum value for the bank's assets if possible, in a sale, and what we need to do is utilize that flexibility to achieve that.
Primarily, our goal would be to transfer the covered bond program over to another bank; that's what we did when Washington Mutual failed in the fall of 2008. The covered bond program was transferred very quickly and it continued to operate.
The second thing that we need to have is the ability to - and this is where I think the questions have arisen with regard to legislation - repudiate the obligations that the bank would have under the covered bond program; make sure the covered bond investors are fully protected accordingly to their rights as secured creditors under that program, but also give the FDIC the ability to recapture the over collateralization that's inherent in the deal. That's particularly important in Washington Mutual's situation, where it failed. They were reporting over collateralization of 149% plus. In some European covered bond structures I've seen effective overcollateralization of those structures of upwards of 300%.
So basically under the current draft legislation, the overcollateralization value would only be paid through a 'residual certificate' at the end of the entire term of the covered bond structure. That does not really give value back to the receivership estate. That's important because we don't want it to have an implication that there's some special protective rules for a class of creditors that is not given to any other secured creditor in the past under U.S. law.
In fact, U.S. law has basically told secured creditors that the value of their money when there is a default is set to the amount of the collateral that they can have claim to. That's all we're saying with our suggestions on how to modify the legislation. Take the full value to the full par value of the covered bond, plus interest to the date of the payment, which would be quickly reinvested and allow there to be a monthly repayment to the investors over the terms of the covered bond structure. That's one of the biggest issues for us, in a proposed legislation, and we want to also make sure we take the balanced approach to regulator involvement in protecting the investors.
We don't want to leave the impression that has truly been indicated in some European jurisdictions, that the regulators are virtually guaranteed to cover covered bond investors' returns because of the role of the regulators in being responsible for making sure that they're protected.
ARNHOLZ: Mike, I didn't mean to imply that the 2008 policy statement that you issued was met with investor dissatisfaction.
I think given the events that occurred after you came out with the statement, there became a much stronger movement toward looking for some legislation.
I know the FDIC has always been very supportive of a covered bond framework and I'm glad to hear the FDIC is supportive of legislation.
I think one thing worth mentioning is that under the legislation covered bonds would generally not be accelerated - which might not be the case under the policy statement. When you talk to investors that is something that appears to be important. Is it essential? Is it essential that investors get the severed covered pool upon default? Hard to say.
KRIMMINGER: I appreciate it, John. I think the acceleration remedy is really designed to put the investors on par and to make sure that the Deposit Insurance Fund is not suffering additional losses by actually being tied up in a secured claim far in excess of what the actual claim is.
In fact, I think there are ways of investing the acceleration, terms that investors have, simply something where there was an investment contract or some type of forward option that would come into effect upon the receipt of this large volume of cash that we would be paying out to terminate the covered bond program.
To make sure that investor interests are preserved, you need to make a monthly payment, I think that's one issue the investors of covered bonds have been concerned about. Nonetheless, I think there are structures that need to be put in place to solve their problems.
ROZENS:The perfect test case in the U.S. was how WaMu's covered bond program was handled when you took over the bank.
KRIMMINGER:I think that's true. It illustrates what I believe to be the likely result in the vast majority of cases. The world I'm dealing with is credit agencies and others; I sometimes tend to look at the worst-case scenario, which might be repudiation.
But nonetheless, the transfer of the covered program over to another institution operating is certainly our preference because that would maximize the value to the Deposit Insurance Fund.
ROZENS:That's effectively what happened with JPMorgan and WaMu?
KRIMMINGER: Yes, absolutely. The covered bonds were fully taken over by JPMorgan and were fully enforced after that.
ROZENS:Let us turn over to Mr. Forster of Moody's. What is the state of the U.S. covered bond market and who are the issuers?
FORSTER: OK, thanks Alex. I think it's important to distinguish between covered bonds issued by domestic U.S. banks and covered bonds that are issued by foreign institutions, but are denominated in U.S. dollars.
As you mentioned in your introduction, there have been only two public, domestic U.S. covered bond programs. The first was done by Washington Mutual in 2006, and Bank of America followed suit in 2007. One thing that is interesting to note is that four out of the five series of covered bonds that they issued were denominated in euros. So at least when these programs were initiated, those banks were targeting a European investor base rather than a domestic investor base.
However, in 2010, we've actually seen a tremendous amount of U.S. dollar-denominated covered bonds issued in the U.S. market by foreign institutions, especially Canadian banks. All five of the Canadian banks with covered bond programs have issued in U.S. dollars, and are targeting a U.S. domestic investor base. Issuers from France, U.K. Sweden and Norway have issued U.S.-dollar denominated covered bonds this year as well.
So clearly there's demand for covered bonds from U.S. investors, and foreign banks are taking advantage of that market. We've seen almost $28 billion dollars issued in the U.S. by foreign issuers this year alone, whereas there has been no issuance by domestic banks since 2007.
ROZENS:I can't help but wonder why someone would issue in U.S. dollars if they were an overseas firm?
FORSTER: Covered bonds are very flexible instruments in the sense that they give the issuer an ability to issue in a variety of currencies. Once the covered bond program is set up, the issuer is able to evaluate its funding needs, and fulfill its funding needs by issuing in whatever currency provides the best pricing.
ROZENS:What makes a covered bond a strong credit?
FORSTER: Most importantly covered bonds are dual-recourse instruments. In that sense, they have advantages over both ABS and unsecured debt. Unlike ABS, the institution that's issuing the covered bonds is the primary obligor of the covered bonds and so long as that entity is solvent, it will continue to pay the covered bonds to maturity. If the issuer fails, unlike unsecured debt, the investors have recourse to a pool of assets. So that's the dual recourse nature of covered bonds.
Second, covered bonds programs typically require the issuer to maintain a required level of overcollateralization, and that overcollateralization needs to be maintained with performing assets. In other words, if an asset becomes seriously delinquent, it's no longer given credit for the purpose of maintaining the required over collateralization, and the issuer would be obligated to segregate new performing assets for the covered bondholders.
And third, the asset quality tends to be high. Sometimes this is required by the covered bond program. For example, in four out of five Canadian programs, the assets are required to be insured by the [Canada Mortgage and Housing Corp. (CMHC)], which is a 'Aaa'-rated government entity. And sometimes the high credit quality is mandated by the covered bond law. For example, the German covered bond legislation mandates a maximum LTV of 60% for residential and commercial mortgage loans included in covered bond pools. And finally, if there is covered bond legislation, typically this provides a level of oversight that investors benefit from as well.
ROZENS:So how does the domestic U.S. covered bond program measure up when it comes to credit quality and just in general?
FORSTER: The credit quality of the mortgage loans in the domestic covered bond programs so far has been very high.
Where I think the U.S. domestic programs are at a disadvantage compared to foreign programs is with respect to what happens if the issuer becomes insolvent and defaults. In that case, we believe that for U.S. programs, investors are exposed to a comparatively higher degree of market value risk than foreign programs, at least under the current U.S. framework.
And to explain that I'd like to take a step back and look at the factors that determine the strength of a covered bond program. Basically, the covered bond program, starts with the strength of the bank or the issuer, and as long as the issuer is solvent, the covered bond investors are getting paid. If the issuer fails, then the covered bond investors rely on the value of the cover pool of assets. We believe you can evaluate the value of the cover pool based on three different risks of loss on the cover pool assets.
First is the collateral risk, which measures the credit quality of the cover pool assets. If the assets were held to maturity, what would be the loss on those assets.
The second risk is probably the most important risk, and this is what I'd like to focus on for the U.S. domestic programs, which is the market value risk (sometimes referred to as refinancing risk). This is the risk of loss resulting from having to sell assets in the market. You don't have this in ABS. ABS are often passthrough bonds but covered bonds are typically bullet bonds and the maturity of the covered bonds will likely not match the maturity of the assets.
So you could have, for example, a five-year covered bond backed by a pool of 30-year mortgages. So the risk there is if the issuer becomes insolvent, assets may need to be sold to make the bullet payment on the covered bonds.
And finally we look at interest rate and currency risk, which measures the risk of loss on the assets due to mismatches between the assets and the covered bonds in terms of the interest rate (for example, floating assets and fixed covered bonds) or in terms of currency, (for example, assets denominated in dollars and covered bonds in euros and so forth).
In addition to the value of the asset pool, we also look at the particular covered bond legislation or, in jurisdictions with no specific covered bond legislation, the applicable legal and conractual famework to determine how likely is it when the bank fails that the covered bonds investors are going to get paid on a timely basis.
ARNHOLZ: Yehudah, can I just ask very quickly: What do you think about the Garrett legislation? Is that helpful in your credit analysis?
FORSTER:Yes, as I'll discuss in a minute, the Garrett legislation does mitigate some of the exacerbated market value risk that exists in the current U.S. framework since it would eliminate the credit-negative scenarios following a bank default that exist under the current framework and replace them with scenarios that are more positive for covered bondholders.
Under the current framework, two of the three scenarios following a bank default expose covered bondholders to exacerbated market value risk. According to the FDIC's covered bond policy statement, there are three scenarios when a bank fails and the FDIC becomes its receiver. The first scenario is that the FDIC can affirm the covered bonds and transfer the covered bond program to a solvent bank. This is what happened with Washington Mutual's covered bond program, which the FDIC transferred to JPMorgan Chase Bank on the same day that Washington Mutual went into receivership. This outcome was positive for covered bondholders..
The second scenario is repudiation, where the FDIC can pay damages to the covered bondholders in exchange for the assets, which it keeps in the bank's estate. The policy statement says the FDIC will pay damages equal to the par value of the cover bonds, but only up the value of the collateral.
And the third scenario is a liquidation of the entire cover pool. This third scenario is where we have the most concerns.If you have to liquidate the entire cover pool in a short period of time, the price for those assets may be steeply discounted, particularly in times when the secondary market lacks liquidity (as we've seen over the past couple of years). This is especially true since the cover pool can be huge. For example, Bank of America's cover pool contains over $10 billion in mortgage loans. If you have to liquidate the entire thing in a short period of time, investors can be exposed to steep losses.
Furthermore, we also have concerns about the repudiation scenario. The way it's written in the FDIC policy statement, the amount of damages the FDIC would pay is limited to the value of the collateral. So if the FDIC measures that by the market value of the assets, really that's a similar risk as having to liquidate the assets, since the entire pool would be subject to the market's valuation.
KRIMMINGER: I think that is accurate in regards to the statement of policy. However, what we have proposed for legislation going forward would be that we would simply basically terminate the program, pay the full value of the bonds, plus interest, not liquidate the collateral at that point, to determine what the value is but pay the full value of the bond plus interest and reclaim the cover pool. Obviously without making a judgment about whether to simply repay the collateral or repudiate the program and pay the full value of the bonds plus interest. We'll be making our own judgment about the value of the collateral and whether there's over collateralization to recapture.
However, we would not be, at that point, liquidating the collateral on the market that presumably at the time one of the issuers would fail or would likely be rather distressed. So that's not one of the options we're proposing under the legislation.
FORSTER: And that's very helpful, Michael. Let me ask you a question on that. I noticed that this was your position in the Senate hearings. We actually wrote an article saying that was a positive credit development. Would that same stance (i.e. paying the full amount of the covered bonds plus interest) apply to the covered bond policy statement or is this just something you're willing to give in for the proposed legislation?
KRIMMINGER: Well, I think it's much easier to suggest this would be a legislative approach, simply because we are constrained, of course by existing law. I say it so many times I sometimes forget when I say it, but we certainly support covered bonds as being an alternate source of liquidity for the U.S. markets, We would support legislation that would adopt what I've just stated, which is obviously superseded by the policy statement.
FORSTER: OK, that's helpful. And, John, just to answer your question, the Garrett bill would mitigate this exacerbated market value risk. Under the Garrett bill, there would be only two options following a bank failure: either transfer the entire covered bond program to a solvent bank or, like the European programs, segregate the entire pool and have that pool be administered as an ongoing entity until the maturity of the covered bonds. Currently, in the U.S. framework, there is no mechanism for the pool to remain outstanding and be serviced and to pay down the covered bonds through a natural amortization of the assets.
ROZENS:How does the U.S. market measure up or compare with Canadian and European covered markets?
FORSTER: The structure of the current U.S. programs are similar to the Canadian and European programs in the sense that when the issuer is solvent, the issuer is paying the covered bond investors.
However, the non-U.S. structures are different when there is an issuer default. In the typical European and Canadian structure, following an issuer default, the cover pool is segregated, administered and serviced on an ongoing basis, and wound down over time through the natural amortization of the assets, selling assets only when necessary to pay off maturing series of covered bonds. In the U.S., this gradual wind down process really doesn't exist.
Either there's a transfer to a solvent bank under the first scenario, which is good for investors; but under the second and third scenarios either the FDIC pays damages or the entire pool is liquidated at once, and then you have just a lump of cash. If you receive the par value of the cover bonds, investors would get paid in full. But then again there's the risk that you will receive less than the full amount since the entire pool will be subject to the market's valuation within a short period of time following the bank's failure. Under such circumstances, prices could be highly discounted.
In terms of the Canadian covered bond programs, some observations of how they compare to the U.S. programs in terms of the assets: First, in four out of five Canadian programs, the assets are insured by the CMHC, a Canadian government entity. We consider these assets to be very strong.
Even in the Royal Bank of Canada program, the one program where the assets are not required to be insured, there is a high degree of regulation as to what kind of mortgage loans a Canadian Schedule I bank can make. For example, unless the loan is insured, the maximum LTV is 80%. Also, there are limits on debt to income ratios.
ROZENS:What were your experiences with - let's say the BofA collateral programs or WaMu programs, what was the collateral, was it ARM paper, fixed-rate, 15-year, 30-year paper? What was the LTV on average, in those pools?
FORSTER:Both of them were very high quality paper. In the WaMu program, at the time it got transferred to JPMorgan, almost all the loans were 30-year loans, with a weighted average FICO of around 750 and a weighted average LTV of around 60%. There was a significant amount of option ARMS in the pool but they generally had high FICO scores and low LTVs.
One of the big risks that we noted before the WaMu program was transferred was that, at that time, the secondary market for mortgage loans was completely closed down especially to option ARMS. We thought that if option ARMS needed to be sold in the market, they may get sold at a very, very steep discount, even if they had good credit characteristics. The Bank of America program contains very prime jumbo, fixed-rate and ARM collateral. The weighted average FICO is around 750 and the weighted average LTV is around 60%.
ROZENS:Could issuers or underwriters resell paper that would normally end up in conforming programs or regular passthrough security issues with a guarantee from the GSEs?
FORSTER: The domestic U.S. programs mainly contain jumbo loans that don't confirm to GSE standards since they are larger than the GSE-limits. Banks can easily sell conforming loans to the GSEs rather than keep them on balance sheet for covered bond programs. This is one of the issues why there may not be demand among U.S. issuers to issue covered bonds. Issuers have favorable execution for mortgages elsewhere, such as the GSEs which are buying tons of mortgages. The Federal Home Loan Banks also provide efficient funding for the banks. So until we see a funding void, I think there may not be a huge demand for U.S. issuers to issue covered bonds.
ARNHOLZ:Part of the answer to that question will be what GSE reform will look like.
ROZENS:One last question for Mr. Forster. How does a triple-A residential mortgage bond compare with a triple-A covered bond in terms of credit quality?
FORSTER: One key difference is that the covered bond ratings are linked to the rating of the issuer, whereas MBS are not.
When you're dealing with an MBS, you're dealing with a static pool of assets that are held to maturity. So the main risk that we look at is the collateral risk, which is the risk of credit loss on those assets. For covered bonds, the issuer stands behind the product and is the primary obligor of the covered bonds. The issuer often performs other key functions as well, such as servicing and hedging. In addition, while the issuer is solvent, the cover pool is not static. The issuer can add and substitute assets into the cover pool and change its composition. Because the issuer has such a primary function for covered bonds, we link the rating to the issuer.
If the issuer's credit quality is deteriorating, we're concerned, among other things, that the originations may be of lower quality and the quality of the asset pool could deteriorate. We generally cap the ratings on a covered bond based on the issuer's rating and also on our assessment as to how likely covered bonds investors would be to receive timely payments following an issuer's default.
For covered bonds, we also analyze the asset quality as we do in RMBS transactions, but we also factor in some additional risks that are not present in typical RMBS transactions, such as the market value risks and the interest rate and currency mis-matches that I mentioned earlier.
ROZENS: So it's more akin to a triple-A corporate as opposed to a residential bond which follows it's own patterns.
FORSTER: You can say that, although the ratings are not always 'Aaa'. It's a corporate rating, plus. We start with the corporate rating of the issuer and then we can notch up based on the value of the assets and our assessment of the likelihood of timely payment following an issuer default.
ROZENS:And Ms. Kanes, given that you're a dealmaker putting these transactions together, who is buying covered bonds and what kind of investors are out there really looking at this product?
KANES: In terms of the nearly $30 billion issued this year from Canadians and Europeans, it's been sold to a wide range of investors who come primarily from the rates space but also from a credit perspective. The interest on the part of the rates investor base is that historically these are investors that perhaps bought government-g uaranteed paper, agency paper, government paper, etc. and are looking for a way to diversify into new 'AAA'-rated investments. Given the backdrop of net down agency issuance, expiration of the government-guaranteed programs and extremely low rates, a positive Libor spread 'AAA' asset is of interest to the traditional rates investor base. And for more credit focused funds, the product offers a way to uptier in the bank credit.
ROZENS:How do the yields compare to let's say agency paper or regular mortgage paper?
KANES: Covered bond spreads range but they certainly offer a pickup over U.S. agencies, Canadian agencies and, of course, versus government debt.
ROZENS:And when you look at the investor base, what are the investor types? Are you getting mutual funds, any sovereigns, any central banks, especially with European issues or Canadian issues?
KANES: It's a real range. Asset managers and insurance companies have been big participants. We're seeing some degree of bank participation as bank treasury funds that invest their own money in liquid high- quality investments (traditionally treasuries or agencies) look for new investment opportunities and diversification. That has been a growing trend. There has been some degree of central bank participation, but that's been name and jurisdiction dependent.
ROZENS:In terms of the type of collateral that's being resold, what are you seeing most, let's say in the Canadian or European programs.
KANES: Well, it's been mainly residential and public sector loans. Of the Canadian deals, four out of the five have CMHC-insured loans backing the pools, the European deals are a range of mainly residential loans although one program has public sector assets as well.
In terms of investors' attitude towards assets, my sense is that they are looking at both the bank and the assets but looking closer at the underlying assets than do European investors. That said, it is a balance between the bank and the pool and each investor has a slightly different weighting. What has been interesting about this year is that we have seen a tremendous amount of education. Investors have had to learn about a variety of different banks and their approach to real estate lending as well as learning about the various covered bond structures, legal frameworks and even sovereign credits. Given the range of issuers who have come to market, investor education has been a big focus this year.
ROZENS:The loan type then could vary from an ARM in one nation to a fixed-rate product in another?
KANES: Each mortgage market is unique and has its own characteristics. For that reason, I think the education element has been key for investors to understand why the metrics for obtaining a loan in France are different from Sweden, are different in Canada.
ROZENS:How do you set up a covered bond issuance program and what do investors want when you start market a covered bond product?
KANES: The first step for the bank is determining what they're trying to achieve and accessing their funding alternatives.
U.S. banks right now are flush with deposits and have other efficient forms of financing their mortgages.
For a lot of nonâ€'U.S. banks we looked at and that we worked with, it's what are my options? What are my alternatives? What can I do in the European covered bond market? If I am looking to export my program to the U.S,, it's a question of taking the documentation and bringing it up to the 144A standard.
For the most part, what we've seen is European and Canadians taking their outstanding programs and just layering on the U.S. specific requirements to sell the same structure and cover pool to U.S. investors. We're not seeing variance in what they are fundamentally selling, rather just ammending the program to conform the the unique 144A requirements.
ARNHOLZ: So Sarah, other than the dollar denomination, it's the same program?
KANES: Well, the documentation is a little bit different, as you know.
ARNHOLZ: But other than that, they're the same, right?
KANES: Yes, I'll say that's true.
FORSTER:I think the important point there is that the legislation of the country of origin is what governs the covered bond program rather than U.S. laws. So, even though it's a dollar-denominated bond, it's not under U.S. legislation.
KANES: Right. Or covered bond structure. For example, the Canadian banks do not yet have legislation but investors are extremely comfortable with the structure given their view of the Canadian banks and legal system.
ARNHOLZ:It's interesting to me, that the Canadian deals that have been so popular and saw such a robust demand, that there is no Canadian statute or framework.
FORSTER:Right now there is no Canadian covered bond statute. Deals have been structured using securitization technology to create a sale of loans at closing to a special purpose entity. Our understanding is that the bankruptcy laws in Canada are very creditor friendly and those structures are able to work there, without a law. Even so, we've heard that the Canadian government supports the adoption of a covered bond law and they're working toward that goal.
But one question that I wanted to ask Sarah. Michael Krimminger mentioned earlier the idea of the FDIC wanting to maintain the right to repudiate the covered bonds and retain that extra over-collateralization in any legislation going forward, which makes perfect sense from the perspective of the FDIC.
But I guess what we've heard from various parties is that covered bond investors just won't want that. Do you think that's true? Do you think that that prepayment risk is something that is going to turn off investors or do you think it's viable to have such a structure.
KANES: My personal view probably varies a bit from some of my counterparts . I do not necessarily believe that a U.S. bank covered bond needs to look just like what a covered bond is in Europe. As John mentioned in his opening remarks, the U.S. market is its own deep and unique market, and I think we should have the flexibility to create something for U.S. banks independent from what has been done in Europe.
So does a covered bond have to survive insolvency? I do not necessarily think it must. Clearly getting the FDIC on board, Mike and his team, comfortable with the product is of critical importance. And while I think investors need to be comfortable that in the event of an insolvency certain mechanisms are in place, I am not sure that getting par plus accrued in an accelerated manner (which was one of the potential outcomes that Mike outlined) is a non-starter for investors. But again, I cannot speak for investors or the dealer community. This is just my own view.
FORSTER: I was just thinking, if you were an investor, and the bank sponsoring the covered bond failed, would you rather have your full principal back plus all accrued interest up to that date or rely on a pool of assets that were originated by that failed bank? I mean, you might be in a much better position to get the cash.
KANES: My sense is that this would be acceptable option but certainly others have a different view.
KRIMMINGER: My thinking is that one of the advantages they offer to investors is that they're a constantly refreshing pool of collateral that backs an effectively corporate bond obligation. The fact that they issue a corporate bond obligation, when the bank fails, inevitably you've now been given a static pool. So it's not going to be refreshing; it's going to be depleting in value, as you have to pay off, etc.
So from our perspective there are two alternatives; you can turn over the collateral pool to investors or the static pool, or at least be provided with the cash that could be structured in such a way that it can be a constant repayment or response, so that they don't really have to bear the investment risk.
In some ways, there are some options that have the same result either way you go.
ARNHOLZ: I think Mike, that's a fair point, a lot of it has to do with who you're selling to and what the market dynamics are at the time, also.
KRIMMINGER: What do you mean?
ARNHOLZ: Well, I think certain investors might have more of an interest in the credit aspect of the bond and others may be more focused on the timing.
KRIMMINGER:That's probably true. I would think there are one or two instances of overcollateralized obligations. Investors should be pretty comfortable, assuming the covered bond program is being administered in a proper way, to keep the OC as they're required; otherwise they would receive the par value of their bonds.
So yes, there always should be consideration as a matter of fact. I'd be hesitant to try to let investors feel that they were completely at the risk of the bank.
ARNHOLZ: Fair enough.
ROZENS:In terms of various assets that could be resold into the covered bond package, I noticed in Europe there could be a note on a cargo vessel. I would imagine that you can apply commercial real estate to this process as well; is that something we can expect to see or do we have to see the residential side take off first?
FORSTER: In Europe we've seen plenty of programs that are backed by commercial assets or mixed commercial assets and residential assets. The asset types are commercial mortgage loans, residential mortgage loans, public sector debt, and a very small amount of shipping loans in Europe. But it's interesting that the legislation proposed by Rep. Garrett, includes a wider variety of assets.
ARNHOLZ: If I can just add two quick thoughts to that. The securitization markets are functioning pretty well with respect to auto loans and other assets. The issue really is what kind of relief can be provided by covered bond programs in the residential mortgage sector.
The other point is that the Garrett legislation was initially written to include auto loans and leases and I believe student loans and home equity loans. I believe on amendment those asset classes were removed from the bill.
FORSTER:If I can jump in here and just give one comment on the Garrett legislation which I do think is a concern, at least from my perspective. Originally the proposal appointed the Treasury Department as the covered bond regulator and then it got switched to the [Office of the Comptroller of the Currency (OCC)], but the latest draft that passed the committee appointed the primary regulator of the bank as the covered bond regulator.
One concern that I have is that there is a natural conflict of interest between covered bond investors and the primary regulator of a bank. I think the primary regulator is going to be looking to maximize the value of the bank's estate upon bank failure; whereas covered bond investors are going to be looking to maximize the amount of assets that are segregated from the bank's estate.
And the legislation actually charges the covered bond regulator to set the minimum collateralization levels. So whereas in Europe, when we've seen issuers deteriorate, we've typically seen them increase the amount of required overâ€'collateralization in their programs to either maintain ratings or maintain investors' expectations.
But I would find it hard to believe that if a U.S. bank were to be deteriorating, that the primary regulator would start putting more overcollateralization into the program only to be taken away from the estate at insolvency. I think there would be a real tension there and could produce an outcome different from what European investors would expect.
KRIMMINGER:I think there's some cause for concern or potential conflict of interest.
I would note, however, that definitely the German bank regulator is also the covered bond regulator. I'm concerned about a conflict of interest in some ways from a slightly different line: if the bank regulator is responsible for making sure that the covered bond investors are protected, would the bank regulator take significantly aggressive action to make sure that institutions close at a time that would reduce losses to the system or to the deposit insurance fund?
And I think there is a potential conflict of interest there which is why we would prefer that the primary regulators not be in the game of being the overseer of the overcollateralization, but that regulators should be setting standards in the industry.
