Introduction
Steve Edwards is a partner with the law firm Kilpatrick Stockton in Atlanta, Georgia, and received both his undergraduate and law degrees from the University of Michigan. He's practiced in the area of structured finance for over 30 years, a portion of which was spent practicing in Singapore. His expertise is structuring and documenting financial instruments, with an emphasis on their tax aspects, and his practice is currently focused on representing servicers in the commercial mortgage-backed securities market. He's written and spoken widely on these topics at conferences in both the United States and Asia.
Interview
Q: So, Steve, what do we mean by the term “Structured Finance”?
A: Well, like a lot of economic terms or terms used in the financial markets, it has different meanings for different people. There was even a survey of experts a few years ago to come up with a definition, but they weren’t able to come up with a clear consensus. At a minimum, most seemed to include securitizations, including asset-backed securities, residential mortgage-backed securities (RMBS), my specialty, commercial mortgage-backed securities (CMBS) and collateralized debt obligations (CDOs). Most included credit derivatives, such as credit default swaps (CDS) as well. I’ll describe all of these in a moment, but I’m going to be focusing on is mortgage-backed securities or securities in which the payments come from an underlying pool of mortgages – either mortgages on houses or mortgages on large commercial buildings.
Q: Can you give us a brief explanation of who the participants are in these transactions and how the deals are put together?
A: I think the easiest way to understand these transactions is by starting at the bottom and moving upward. At the very bottom are the borrowers – people who want to borrow funds and secure the repayment of the funds with an interest in real estate. These could be people like you or me buying or refinancing a house or people like Sam Zell buying or refinancing an office building. Unfortunately, as we all know by now, they also included people who had no business borrowing large amounts of money – the so-called subprime borrowers. Credit quality also supposedly declined on the commercial side in recent years, but nowhere near as sharply or deeply as on the residential side.
The borrower in turn goes to a bank or a mortgage broker, who agrees to give them money and take back a mortgage on the real estate. The bank or mortgage broker can be a big bank or a small bank or an independent mortgage broker – there are . . . or were . . . thousands of these operations around the country. You probably got spam from some of them.
The mortgage holder then sells the mortgage to a street bank . . . investment banks like Wachovia, Lehman Brothers, Merrill Lynch, or Bear Stearns, all very familiar names in the news these days. The investment bank would then structure an issue of bonds payable from payments on those mortgages. These were huge issues, literally billions of dollars apiece.
Q: Why would they do that? What was the financial incentive?
One simple answer is that there is security in diversity, so in order words, the if you only own a single one-hundred-thousand dollar mortgage, you’re in real trouble if it defaults, but if you own a one hundred thousand dollar piece of a billion dollar’s worth of mortgages, the default of one of the mortgages in that pool isn’t so much of a concern. That’s one reason. The other reason you’d do this is to allocate the repayment risk to various purchasers according to their appetite for risk. The way to do this is by putting all of those mortgages into a pool – a trust fund actually – and then allocating the payments that the pool receives on the mortgages among the various holders of interests in the pool – the certificate holders. The certificateholders – are divided up into classes, which are usually referred to as “tranches.” “Tranche” is just the French word for “slice” and it just means that you’ve divided up the repayment pie. I think a better analogy is one those pyramids of champagne glasses you sometimes see at fancy parties. Think of each tranche as a level of glasses and whether a tranche is paid depends on how much champagne you’ve got. The most secure tranche is the one on top. If any champagne gets poured out, it’s going to go into that glass first. Once that glass is full the champagne flows over into the next layer and once that layer is full it goes into the next layer, and so on down to the silver tray at the bottom. Tranches in mortgage-backed securities operate the same way. The top tranche gets paid before anyone else does, so it’s the most secure class and it gets the lowest interest rate, the next layer is a little less secure and the layer after that is still less secure. This continues right down to the bottom layer, which is called the “first loss piece.” In residential mortgage-backed securities (or “RMBS”), it’s sometimes called the “toxic waste.” If there’s any loss whatsoever in the underlying mortgages, that layer will experience a loss. In return, of course, they get a much higher interest rate.
Q. So once an issue of mortgage-backed securities is structured, what happens next?
A. The investment bank takes the structure to the rating agencies. The rating agencies rate all but the bottom tranches based on their analysis of how likely it is that there would be defaults on the underlying mortgages and how likely it is that such a default would affect each successive layer. Going back to my analogy to the pyramid of champagne glasses, if the champagne is the cash flow on the underlying mortgages, they look to see how much champagne they think will be in the bottle. If they are absolutely certain there will be enough to fill a layer of glasses, that layer (or tranche) gets a triple-A rating. If they are a little less certain, it gets a double A and so on.
Q: Who buys the various tranches?
A: After getting the rating, the investment bankers sell the highest-rated layers to institutional investors. Remember that these securities really exploded during a period when rates on traditional triple-A rated securities (such as Treasuries) had been pushed to a very low level, particularly after the Fed started lowering rates to forestall a recession following the dot-com crash. Institutions and their investors (such as bond fund investors) were very hungry for yield on highly secure obligations, and these securities helped fill that appetite.
The lowest-rated and unrated pieces go to much more sophisticated, highly specialized investors. On the residential side, these are shops that employ math or physics Ph.D.’s to do Monte Carlo simulations and other highly complex calculations of the likelihood of defaults based on certain assumptions. On the commercial side, they are shops with real estate experts who look at the rent rolls, the management of the building, and the strength of the tenants. In both cases, they’re very, very sophisticated investors, but like all investors, they have to make assumptions. Unfortunately, some of their assumptions – particularly on the residential side – were wrong.
Q: So, it sounds as if these were sophisticated investments backed by real assets, purchased by sophisticated investors. What went wrong?
A: The biggest problems have arisen even further up the chain from the original borrowers. After these structures had developed for a while, some financial wizards came up with the concept of collateralized deposit obligations (CDOs). Their great idea was to take some of the cash flows at the very bottom of the chain of a number of issues – the stuff that ends up in the silver tray in that pyramid of glasses – and bundling those together. You take a portion of the toxic waste from a variety of issues and put them into a single investment. Then you create a new issue that does the same things you had just done with the cash flows from the mortgages themselves. In other words, you have the least credit-worthy instruments in an issue, but you’d say to one investor that at least some of these really bad investments are going to pay, and when they do, you’ll be paid first. Voila! You’ve really made a silk purse out of a sow’s ear, because you’ve ended up with at least one class of investments that is rated AAA, when all of the underlying assets are, in a word, crap.
Q: How do credit default swaps fit into this?
A: Even this structure might not have produced the disastrous consequences we’ve seen accelerating over the last couple of years, but those same financial wizards had yet another idea to produce high-yield, low-risk investments. They took the lower-rated portions of these mortgage-backed securities and CDOs and “insured” them using credit default swaps (CDS). The investors would go to an insurance company, the largest and best known of which was AIG, and they would agree that, in exchange for an annual payment, the insurance company would pay the amounts due on the underlying securities in the event of a default.
Now, up until now, all of the risks of these various parties, whether they be investors in the mortgage backed securities, investors in CDOs, or even companies that issue the credit default swaps, are related to the underlying mortgages and secured by the mortgaged property. As a result, up until less than a year ago, everyone thought that the overall risk, the systemic risk, was a fairly manageable amount. The subprime borrowers could default on their mortgage loans, but, hey, the loans are secured by a house, which could be sold to help pay down the loan. These are real assets.
That’s a still fairly good logic. In fact, my understanding is that, on the residential side, if all the subprime mortgages defaulted, it would be a significant hit to the investors in RMBS, but it wouldn’t be catastrophic. In fact, less than eighteen months ago, Ben Bernanke said that the losses from subprime would be limited to 50 billion dollars and that the losses had been contained. On the CMBS side, defaults are historically low. I think the current default rate is one-half of one percent.
What increased the aggregate risk in the system exponentially is that institutions started buying credit default swaps on assets they didn’t own. It was a bit like a naked short sale of CMBS, RMBS, and a wide variety of other financial assets, including the debt of major corporations. In fairly short order, CDS and other derivatives based on mortgage-backed securities amounted to over ten times the aggregate amount of the securities themselves. I’ve heard that the total amount of CDS earlier this year was 60 trillion dollars. That is more than all of the stock sold anywhere in the world. Another way to put this into perspective is that there are 70 trillion dollars worth of money reserves in the world – the entire world.
Leverage produces magnified gains, but it also produces magnified losses. None of us on the commercial side worried when defaults rose in the residential mortgage market because, as I said, we thought the number of defaults would be relatively small and the owners of the defaulted mortgages would be able to cover most of those losses by selling the mortgaged homes. Unfortunately, what everybody missed was that so many institutions had made bets on those mortgages using CDS on CDOs and other derivatives that those losses would be magnified into huge losses for those institutions. The losses were further magnified because a couple of other things happened: The losses those institutions incurred (which were exaggerated by the mark-to-market rules imposed by the accountants) reduced the amount they had available to lend or invest in other mortgage-backed securities. That meant, in turn, that the free-and-easy mortgage market ended abruptly. That meant there were fewer qualified buyers for houses, which reduced demand at exactly the same time that more houses came on to the market as a result of foreclosures, which increased supply. That drove the entire real estate market down.
As more and more institutions experienced losses, less and less money was available to keep the credit markets flowing. No one wanted to lend to other institutions because they couldn’t be sure who else had significant exposure in the mortgage-backed securities markets, the CDO market, or – worst of all – the CDS market. Most CDS was “netted,” which means that an insurer of an obligation would usually turn around and buy its own insurance from another institution, which would turn around and do the same. Because the the Federal Reserve and Congress had affirmatively decided not to regulate this market, no one knew who the ultimate counterparties were and whether they were creditworthy. Credit froze . . . . completely.
Q: So, who’s the villain here, and what’s our path out of this mess?
A: There’s no one villain or group of villains. Everyone along the line was culpable. The borrowers were in many cases naive or dishonest and too many of them borrowed more than they could afford. The mortgage brokers were venal or, in the case of FANNIE or FREDDIE, arguably politically motivated to make aggressive loans. The investment bankers were perhaps too eager to put together financial instruments that even they didn’t understand. The buyers (including those of us who invested them) were so hungry for supposedly secure yield that they bought things they didn’t understand. The CDS counterparties were too eager to get the revenue from the payments on those swaps and didn’t do sufficient due diligence on the underlying obligation or on their own counterparties. The Fed and Congress were too confident in the ability of the free market to prevent important financial institutions from risking more than they could afford to lose. Even the accountants imposed rules that forced companies to report losses they might not incur if the related investments worked out over time.
Where we go from here is almost anyone’s guess. I’m confident that the credit markets will right themselves, but I have no idea of how long that will take. We’ll certainly have more regulation of home mortgages, CDS, and the special purpose vehicles that many institutions used to issue and invest in CDOs and MBS, but what form those regulations will take will be up to the politicians, so it may be too little, too late. The party will almost certainly start up again, but we’ll all probably drink less champagne.
