By Thomas Adams, at Paykin Krieg and Adams, LLP, and a former managing director at Ambac and FGIC.
In my view, Goldman, and a host of other clever bankers, are deliberately obscuring one of the most important points about modeling, CDOs and sophisticated investors. One of their defenses against the tremendous losses these products delivered amounts to “caveat emptor”: everyone is a grown up and should have known what he was buying.
But that conveniently obscures a critically important fact: for so-called ABS CDOs (the kind made from asset backed securities, meaning tranches of either residential or commercial mortgage bonds), 75% to 90% of the deal was rated AAA. And these ratings did not depend on the insurance provided by AIG or monolines; those ratings were issued on the CDO as concocted by the packager/underwriter.
The argument that the CDO market blew up because it was so complex and speculative is fundamentally flawed. Believe it or not, the bonds that caused the damage to AIG, the bond insurers, and banks were not highly speculative, high risk bonds. They were AAA securities and were supposed to be virtually free of credit risk. In many cases, they were “super senior” bonds – meaning they had another layer of protection above the AAA level to make them even safer than regular AAA bonds. In return for this high level of safety and large levels of protection, the investors or insurers received a very low AAA level yield. In addition, because the bonds were heavily protected, AAA rated and presumably safe, the investors and insurers did not (and were not required to) allocate very much capital to them. The bonds were not considered risky, so there was little need to reserve capital against them. In contrast, the investors and insurers held much more capital against the BBB bonds in their portfolio. This was fundamental to their business model.
Historically, investors viewed super-safe, high quality investments, such as Treasury securities, as simple, non-sophisticated investments where their money could be placed safely without worry about complex models and detailed review. In most institutions, a novice portfolio manager starts out by investing in AAA bonds, and works his way up the sophistication curve over time. Over the past few years, many investors seem to have forgotten this. But anyone who thought they needed to be super-smart, sophisticated and using “cutting edge” models (the phrase my old boss used) in order to invest in AAA bonds was either a fool or a fraud. In retrospect, there were plenty of both types.
In other words, these highly complex products were targeted to buyers who in many cases were the very least sophisticated institutional investors. And fund managers, both freestanding ones, and ones within larger organizations (think portfolio managers at insurance companies) are subject to competitive pressures. If industry benchmarks for AAA returns start to include complex, manufactured AAA paper, these investors would damage their careers by sticking to what they were sure they understood and shunning the more complex instruments that offer a bit of a yield pickup. As Keynes observed,
A sound banker, alas, is not one who foresees danger and avoids it, but one who, when he is ruined, is ruined in a conventional way along with his fellows, so that no one can really blame him.
The whole notion that an investor must do complex, detailed, sophisticated analysis of AAA bonds in order to understand them and not get ripped off is, as a result, completely upside down. A high yield or a distressed bond requires a lot of analysis to determine if it is safe. This may involve elaborate models and multiple scenarios. Investors who pursue these types of risky bonds believe they have the tools to understand all of the considerations that go into the decision to purchase it or not. This type of investor is sophisticated and is aware that he is either taking risk or making a speculative bet. To compensate this investor for such a bet, the bonds pay a high yield and, if the bet is right, produce a high return. Even if the bet eventually goes bad, the yield may be enough to pay the investor so that the loss is offset. Often, investors in these bonds employ teams of quants and Phd’s to help them understand their bets.
Investments in AAA bonds are supposed the opposite of this type of high yield investment. It makes absolutely no sense that a AAA investor should need the same high tech models to make his investments or that if he fails to properly use his models he will be completely blown up and his investment will be worthless.
This “highly sophisticated investor” argument has been used by Goldman, other banks and a remarkably high number of journalists (in my opinion just repeating the crap they have been fed by their sources) as a way of getting the banks off the hook. But it is a fundamentally flawed argument. The CDO bonds that AIG insured were rated AAA. If you have to be a rocket scientist to understand the investment and if anything short of perfect analysis of the bonds means you will be blown up – then by definition the bonds are not AAA.
AAA = easy and low yield. BBB = complex, high yield. This is a pretty simple and bright light distinction. And yet it has been blurred and ignored throughout the discussion of this asset, as recently as two days ago by Goldman.
A “super senior” investment in any asset should, by definition, not require sophisticated analysis. it should be as safe as, or safer than, Treasury securities from a CREDIT perspective (ratings do not address market value or interest rate risk – which could be more complex).
Neither investors in, nor insurers of, AAA or super senior bonds were compensated with high yields – they were not yield hogs, as suggested by one comment on NC the other day. If they were getting huge yields they would have been on notice that their risk was high.
I don’t want excuse the insurers or investors of all of their responsibility. I was one of them. We got sucked into the game of “complex” AAA bonds and believed that we were sophisticated. It was part of the cultural experience of the era – even boring old muni people wanted to believe that they were “special” and “sophisticated”.
An investment that purports to be AAA or super AAA should be obviously of high quality. There should be “no ifs ands or buts” about it. The investment should not be binary, such that if you are “right” it is AAA and if you are wrong it is worthless. Model sophistication, extreme diligence, the need to have skepticism about potential lies or misrepresentations from sellers and issuers – all of this is completely incompatible with a AAA bond.
Goldman Sachs and dozens of other banks and captured journalists want the story to be about the failure of the investors who were sophisticated and assumed the risk. They ignore the crucial fact that the bonds that blew up were AAA and were sold as virtually risk free. The entire environment created by the banks, managers and rating agencies for the AAA CDO market was false and contrary to the definition of what was being sold.
The reason that is most frequently cited as why AAA CDO bonds collapsed in value is that they had extreme cliff risk, or tail risk. However, the notion of “cliff risk” should be incompatible with a AAA bond – by definition. Any model that obscures or ignores or adjusts this issue away, is a form of sophisticated lying, as it relates to AAA bonds.
The problem with the CDO market, and a good chunk of the financial crisis, is that the participants took complex, highly volatile, highly risky and highly leveraged assets and passed a magic wand over them to turn them into AAA. Unfortunately, this process did nothing to remove the volatility, risk, complexity or leverage (in fact, the CDO made all of these worse). From the very start, the market for AAA CDO bonds backed by ABS collateral was a fraud; the advocates of these bonds used fancy models, flashy Powerpoint presentations, expensive meals and a whole lot of flattery to convince people it wasn’t a fraud, but that didn’t change the truth.
The essence of the issue is that these AAAs that blew up and went to zero (and this is no exaggeration, many former AAA-rated CDOS are utterly worthless) were hopelessly badly designed and/or fraudulently sold. As Yves and I will be discussing in upcoming posts (and Yves covers at some length in her book), there is ample reason to believe a lot of the CDO packagers and underwriters knew exactly what they were doing
The rating agencies should have known that this degree of complexity and an AAA rating were fundamentally incompatible, but they were financially incented to ignore it (they got paid much more money for rating CDOs). The investors and insurers should have known it too. But I am pretty confident that the biggest responsibility lies with the sellers and the creators of the bonds – they were selling something that was supposed to be super safe but turned out to be worthless – and they knew this to be the case, one way or the other.