Editor's note: This is Part 1 of a special series. Here are Part 2 and Part 3 NEW YORK ( TheStreet) -- Today I want to get into the weeds a bit on the securitization market. All through the crisis we kept hearing how we needed to "get the securitization engine humming again" in order to get credit flowing, but I am not sure if people fully understood the role of securitization in the crisis. I think we need to understand how this market evolved over the years -- because if we don't understand how we got here, we will run the risk of jumping to conclusions based on faulty assumptions. In a way, a market based on assumptions that proved to be vastly incorrect eerily foreshadows the dangers of our policy response. I will warn up front that while I will try to simplify, this may be a little dense. This is not something that can be easily encapsulated into a sound bite or "Tweet," but only through understanding can we properly evaluate the situation. You will not need a PhD in calculus to get through this; simple multiplication is fine, but I will make a few options market allusions. Part of the reason people have trouble explaining the root of the financial crisis is that this took the last couple of decades to develop. We kept building on what we took for gospel. And as I mentioned in one of my previous articles which encompassed the Hayek quote, sometimes we need to look at the very bedrock of our assumptions -- those views which we took for granted, that couldn't possibly be wrong. When you are in the weeds, often you forget to look at how you got there. So sometimes it is difficult for a market-maker to describe his or her role because the market-maker is on the battlefield. This is why I think some of the traders at the Senate hearing couldn't say, "No my role isn't as fiduciary, I serve my clients by making markets for my clients so that they can best execute their role as fiduciary." The last systemic financial crisis we had was with the saving and loans in the late '80s. Of course this is subject to debate, because we have had the peso crisis, the Asian flu, the Russia crisis and the tech bubble since then. But the S&L crisis was the last real systemic crisis we had. If you think about the parallels, it seems like our recent crisis was a replay of the S&L crisis: Lax regulatory oversight led to excessive risk taking, we had real estate slumps both residential and commercial, certain key industries were in secular decline (then it was defense, now it is real estate tangential), large U.S. financial institutions failed and the financial plumbing underwent restructuring. Of course there are some differences (rapidity of response, interest rates, global coordination), but I am sticking to the similarities because they are so striking.
However, despite the similarities, there is one key difference that has dramatically increased the complexity of our current situation. In the S&L crisis, there were tangible assets at these financial institutions that were out of favor/under water: mortgage whole loans, commercial real estate loans and high-yield bonds. Securitization was used as the takeout strategy. Securitization was the answer; securitization was the savior. By packaging and securitizing these assets, we were able to monetize them (indeed the first CDOs ever issued were a takeout strategy for high yield portfolios at S&Ls and insurance companies in the late '80s). Securitization allowed the fragmentation of risk -- people could gravitate toward the piece of the transaction that best met their risk/reward preferences. Today, however, securitization may have contributed to the problem. How did securitization go from savior to goat? How in the heck did we come full circle? How did we get from promoting home ownership to the Great Recession? That is the story we'll explore, so grab a cup of coffee because here we go. I will be educational, and occasionally I will editorialize in order to point out what I view were the faulty assumptions/policy decisions. Let's start with some background. Securitization of home loans has generally been viewed as a very good thing. It enables a recirculation of capital by moving loans off banks' balance sheets, which enables them to lend to more homeowners, and moves those assets onto the balance sheets of institutional investors who are seeking fixed income-type instruments. Within mortgage-backed securities, there are two basic flavors: agency MBS, and nonagency MBS. Agency MBS are your basic FHA/VA "conforming" loans. They get insured by the government-sponsored enterprises, or GSEs ( Fannie Mae ( FNM), Freddie Mac ( FRE) and Ginnie Mae) who take a fee for their guarantee of timely payment of interest and ultimate repayment of principal. Investors in agency MBS love this stuff because they view it as a de facto piece of government paper with no real credit risk. But there is prepayment risk. In addition to scheduled principal payments, the borrower can prepay all or part of the loan at any time. The borrower may move, may refinance or may decide he or she doesn't want any debt anymore. This makes MBS very quantitative to analyze. Wall Street hired many rocket scientists to help quantify the prepayment risk. This "quant" phenomenon with mortgages dates back to the 1980s and helped set the stage for what was to come.
Agency MBS also brought about the experimentation of "tranching" securities. The first collateralized mortgage obligation, or CMO, was done in the early 1980s as a means of segmenting the prepayment risk of the pool. Some people wanted the shorter-maturity cash flows because they bought shorter-duration bonds and others (like pension funds or insurance companies with longer-dated liabilities) wanted longer-duration bonds. The CMO was a useful instrument because it allowed investors to tailor their exposure, and thus allowed for a deeper and more liquid market. But now we needed more quantitative analysis to not only model the prepayment risk, but also how the prepayment risk would flow through the structural cash-flow waterfall of the CMO. So that is agency MBS. What about the nonagency eligible MBS? Some mortgages do not "conform" to agency standards. Is there a way to securitize those? By the late '80s people began to look to ways to securitize these loans, but it didn't really take off until after the securitization boom following the S&L crisis. Nonagency MBS does not have the quasi-government guarantee of Fannie and Freddie securities. So how do we provide credit enhancement to get these securities the coveted triple-A rating? Because, after all, mortgages are about prepayment risk, not credit risk. The first batches of nonagency MBS had third-party guarantees, similar to what Fannie and Freddie were providing. But those guarantees (either from a bank or a monoline insurer) had issues. What happens if the guarantor gets downgraded? Using the "weakest link theory," if a guarantor gets downgraded, then the security gets downgraded. Plus there were capacity issues. If I buy a mortgage pool guaranteed by Citigroup ( C), that security could take away a credit line to Citi that I may have. That may impede Citi's ability to finance itself because now I won't buy its commercial paper or bonds because I have full Citi exposure. Citi may decide that it has enough exposure to mortgages, so the business line may not be sustainable as a guarantor, either. That led to a drive to provide credit enhancement for MBS from within the pool of mortgages themselves, so that securities wouldn't be reliant on a guarantee from a third part, be it government or corporate. This concept is really quite simple, because most mortgages themselves work this way. The buyer makes a down payment, and that equity in the home provides credit enhancement for the lender that provides the mortgage.
So the preferred credit enhancement method became what is known as the senior/subordinated structure. A portion of the pool would be subordinated in order to provide credit enhancement for the senior tranche. So we have now migrated from time tranching (the CMO) to credit tranching. Bring in more quants! How do we determine how much credit enhancement is necessary for a triple-A rating? Let's go through a simple example. Let's say I have a pool of mortgage assets that I wish to securitize. The rating agency will look at the pool characteristics and say that historically maybe 2.5% of this type of pool will default. (In reality they also will look at loss given default, which is how much the mortgage loses when the buyer sells the house, etc., but I don't want to get too weedy here.) So being conservative, the rating agency might say, if you want a triple-A rating, you will need to provide credit enhancement for six times that historical default rate. That means you need to provide 15% subordination. So 85% of that pool is the senior, triple-A-rated piece, and 15% is the subordinated piece. Sometimes this is referred to as the residual, or the sub piece, or the equity, or sometimes the "toxic waste" -- a characterization which I will dispute in a moment. But first let's look at our example. Let's say the mortgage pool yields 7%. The underlying mortgages are probably 7.5% coupons, but the servicer takes out a 50-basis-point fee for collecting payments, etc. Now, let's say the market clearing level for triple-A nonagency MBS is 6%. That may be 100 basis points higher than Treasuries (which, in our example, will yield 5%), with the extra yield provided to compensate for the prepayment risk and for the fact that they do not have the full faith and credit (or even quasi-faith and credit) of the government behind them. So our triple-A security gets 6%. The residual cash flows down to the residual piece. That triple-A senior piece has a "delevered" position in the portfolio. Someone else is absorbing the initial losses. The important thing here is that they can never be worse off than just owning the outright portfolio. They own the portfolio, but they also have protection from some amount of losses. So one can think of this as an investment in the "stable value" portion of the mortgage pool.
So what does that look like for the residual? Well that bottom 15% gets the 7% yield on the portfolio plus that extra 1% on 85% of the portfolio that is only getting a 6% coupon. Since 85% of the portfolio is 5.67 times the size of that 15% sub piece, that means that extra 1% across that piece is worth an extra 5.67% of yield to the "resid." So the resid yield is 7% plus 5.67%, or 12.67%, or 7.67 percentage points higher than Treasuries. Juicy, huh? But remember that the trade off is that they are in a first-loss position in the portfolio. What's more, they are in a levered loss position with respect to the portfolio. If the portfolio loses 1%, that is 1/15 (or 6.67%) of their investment. So a 1% loss in the portfolio brings their return in line with the triple-A piece. A loss great than that means they're worse off, and a loss less that means the resid is better off. This is why I don't understand why people deem this "toxic waste." Let me explain. I will start with the question as to why one would invest in a portfolio of mortgages (or any portfolio of risky assets, for that matter) in the first place. The reason is they feel there will be some excess return above the risk-free asset. If they didn't, they wouldn't invest in the risky portfolio. So if our portfolio yields 7% and the risk-free asset yields 5%, we can say that there is 2% excess return that can absorb losses before we break even with the risk-free yield. If losses are greater than 2%, we are worse off; therefore if we think losses will be greater than 2%, we wouldn't be enticed to buy the portfolio. (And I am ignoring the time aspect, to simplify, but keep in mind that is 2% a year). So one could argue that owning the residual in the portfolio is a pure play on the "excess return" of that portfolio. Let's say I want exposure to $1 billion of mortgages. I could put $1 billion at risk in the pool, or I could buy a much smaller piece of a resid -- one that gives me the excess return of that portfolio, but limits my losses. Yes, I can get wiped out, but I have limited my losses.
In essence I have created a call option on the portfolio (I can think of it as I am long a put on 85% of the portfolio while owning the portfolio -- put/call parity says if I own a portfolio and a put I am long a call). That may not be a bad way to think about it if you have an options-oriented mind. Or, you can think that this is just like owning a stock. At some point if the company gets into trouble, the equity investor is wiped out and he hands the keys to the bondholders. Where people get into trouble is when they say, "Hey I want $1 billion exposure to MBS, so I will buy $1 billion of these residuals. That is like saying I want $1 billion exposure to the S&P 500, so I will go spend $1 billion in option premium. Much different leverage! Continuing the options example, how should I think of the senior tranche? Because there is credit protection of six times the historical loss rate, one could say the senior piece has written a way out-of-the-money put and is collecting a little bit of premium for that. It is secured by the mortgage pool so it has overcollateralization, but if losses reach a certain point, the owner get stuck simply owning the pool. That is essentially what AIG ( AIG) was doing, by the way. It thought it was writing way out-of-the-money puts that would never get exercised. But the assumptions on what "normal" loss rates should be were simply wrong. This is a prelude to the danger of building assumptions on assumptions, which we will get to in a little bit. With credit tranching, like the time tranching of CMOs, investors could now more specifically tailor their investments. They could gravitate more toward the stable value piece of the pool (the senior tranches) or the excess return piece of the pool (the subs). I think most people would agree a portfolio of assets can be broken down into a stable value portion and an excess return portion. The $64,000 question is determining where that demarcation line is. Bonds either mature at par or at recovery. So if we knew recovery for certain, we could say that any price below that represented the stable value portion of return, and anything above that represented the excess return. But we don't know in advance, so we must estimate. It's the same with determining how much subordination is appropriate. Clearly the last dollar of loss in a mortgage pool should be triple-A-rated. I mean, even if everyone defaulted, there would be some value left over in the inventory of houses, even if for simply scrap value. So the role of analyzing these pools is to determine exactly how much is needed to properly enhance the portfolio.