Apocalypse of Structured Finance

Around January 2007, I received a job offer from AIG Risk Management. I did not take the offer for various reasons, one being that I was still under a consulting contract with my employer at the time, but I vividly recall the meeting with Chief Risk Officer Bob Lewis, who told me that nobody can price CATs (catastrophe bonds) well.

Following the tsunami that hit Indonesia, many southern Asian countries were willing to buy natural disaster coverage from AIG, but due to the difficulty in liability pricing, AIG was not selling this coverage (Lewis was hoping new hires might help in this regard.) This gave me an impression of prudence on AIG’s part. Little did I know that at that time AIG was actually selling massive CAT bonds disguised in the form of CDS, CDOs and so on. I no longer wonder what would have happened if I had joined “AIG Risk Management”; it’s enough to recall the Chinese adage: “Insult is spared if one knows how much is enough, peril can be avoided if one knows when to stop.”

The term “financial crisis” has become rather banal these days. From the 1636 tulip mania in Holland to the subprime fiasco of the US in the twenty-first century, there is no lack of spectacular stories to tell. Instead of repeating the same old lecture along the lines of “History demonstrates that in nearly every collapse, greed and corruption transcend all social, political, and racial divides…” let’s not be afraid to insist on the cliche that this time is different. It is different in the sense that there are new things to understand and new technologies to master, and also this time, the impact is global.

Finance basically asks two questions: “how much?” and “when?” The latest development in answering these questions as precisely as possible is in structured finance. It is the study of value and risk in structured securities, which are debt securities backed by pooled cash flow receivables. These cash flows can take the form of loans, leases, bonds or other financial assets whose credit risk has been de‐linked from the originator through sale, swap or assignment. (When receivables are sold, the transaction is typically referred to as a “securitization.” When only the risk, but not the ownership, is transferred, the transaction is more commonly referred to as a “synthetic securitization,” or else the generic term “structured finance” is used.) The result is a transaction governed under a set of documents with covenants crafted to achieve a unique profile of risk and return for one or more classes of debt. If done correctly, it is the most efficient and transparent way of funding, providing the lowest cost to the issuer and the most comprehensive performance information to investors. Yet when it is not done correctly, deception and disaster may occur on a grand scale.

While working as a trustee for the world’s largest trust, I personally witnessed ABS deals that lost tens of millions of dollars month after month. There’s no shortage of culprits for the current state of the US financial market, because everyone is to blame, from investors to rating agencies to investment banks:

Investor to rating agency:
“You irresponsible crooks, how could you give investment-grade ratings to all this toxic junk? Isn’t part of your name “Investor Service”? And investment banks, how can you sell me this crap? Where is the due diligence you promised me?”

Rating agency to investor:
“I serviced you all right, but you didn’t pay me, the issuer did, and our rating is based on the credit committee’s superb experience in corporate rating, plus the sentiment that house prices never go down, hence some minor misspecifications in the model parameters. The real culprits are those investment bankers; they originate this stuff and ask us to rate it, and the data they give us don’t include all the information that’s needed, yet we’re still able to come up with a nice rating by suggesting extra credit enhancements such as insurance wraps, etc. I think we did a great job under the circumstances.”

Investment banks to rating agencies and investors:
“I paid you handsomely (44% of Moody’s revenue in 2006 came from rating structured products), and look at what you did — I bought a lot of this stuff myself, and insured some more (as a CDS investor). Now I’m a victim, too. You investors are just as responsible, because it’s your hunger for yields that has driven me to go out and acquire as many loans as possible. And the damned mortgage brokers didn’t do their jobs right.”

Mortgage broker to investment banks:
“How can I do my job right when Fannie, Freddie and others, particularly you investment bankers, take away my incentive for due diligence? If I’d been diligent, I couldn’t have originated all the loans you wanted. The ones who should be blamed are the lax regulatory bodies and those homeowners who shouldn’t buy a house they can’t afford to begin with.”

Regulatory body to Investors:
“I just want to promote home ownership. The fact that some people abuse the system is unfortunate.”

Homeowner to everyone:
“I don’t understand what you guys are talking about! All I remember is you told me I could afford this house, and if not, I could refinance it or sell it at a profit later, since the price always goes up. Now I can’t afford my payments and I’ve lost my house, and because of the economy, I’ve lost my job too! All of you should be shot!”

You had to be “told” whether you could afford something…?

All this might lead someone to believe that securitization (or structured finance) does not work, and that we should avoid it or even abandon it all together. This is obviously the wrong conclusion. When the Tacoma Bridge collapsed, no one blamed the concept of a bridge; no one said, “We’re never going to build a bridge again — it’s too costly and complicated, and we can never be sure if the structure is sound. And besides, all these fancy complicated bridges do is to help move toxic waste around the world. Let’s always take a boat, no matter how inefficient or unfeasible the circumstances.” What we should blame is our lack of understanding of nonlinear dynamics and our naive assumptions. The engineering discipline and mathematics both work, but when the assumption is constantly appreciating home prices, then whatever rating comes out of whatever model is inevitably going to be meaningless.

While the guilt trip game can be fun, it’s not very productive, so let’s look at some revelations and go from there:

Revelation 1: The most important risk involved in securitization is not the number of tranches or what kind of fancy prepayment modeling acrobatics is practiced. The effort should be directed toward the quality of the underlying assets, making certain that the structure has a unique claim on the assets (i.e., that they are not simultaneously owned by the bank), plus the correct assessment of the leverage or structure of the deal. None of this can be achieved without transparency.
Revelation 2: The current status quo of the rating agencies in the market is like the Medieval Papacy and needs a Reformation now. The interpretation of “ratings” needs to be fundamentally changed if not completely replaced.
Revelation 3: If government is a necessary evil in society, then who plays the role of an “almost but not quite necessary evil”? The quasi‐government body (or GSE, government-sponsored entity) is basically a moral hazard. It separates risk creation from risk-taking, and creates undue complacency in the market.
Revelation 4: The so-called protection-sellers (CDS investor) and/or monolines sometimes find themselves on the Titanic as well (AIG serving as a prime example). If a CDS is “like” an insurance contract, then it should certainly be regulated “like” an insurance company. While insurance products have statutory reserve requirements from the National Association of Insurance Commissioners, CDS players have a “gang of six,” a group dominated by CDS traders, who usually have problems focusing on regulations.
Revelation 5: CDOs are like CAT bonds. They are very sensitive to the macro environment. The higher yield one gets from CDOs compared to similarly rated products is a false comfort. So before we know what we’re doing, maybe we should spare ourselves follies such as CDO2. Sure, we can play with different copulas and default correlation models, but whether the process creates further economic value is unclear. What is clear is that it is highly leveraged and creates concentrated systematic risk.

Of course this is not an exclusive list, but it should at least make us think and guide us in the right direction. We will definitely need to continue building bridges, but first we need a new arbiter paradigm. The ratings business today is very similar to thermoscope research in the early seventeenth century. Many scientists experimented with various liquids and designs of thermometers, and all had figured out that certain materials contracted and expanded due to outside conditions, but the concept of temperature had yet to be discovered, so each inventor and each thermometer was unique—there was no standard scale. Likewise today, the ratings of different rating agencies are interpreted differently. Even within the same rating agency, there might be a different scale for different sectors and hence different meanings (i.e., the rating itself has no objective, precise standard).

Let’s rethink this whole idea — what does a rating mean? If expected cash flow is an inherently cardinal measurement, why does its rating have to be ordinal? Put another way, why is the rating scale with respect to an ordinary company the same as rating the probability of a batch of cash flows and/or the legal soundness of a stack of papers? (Remember, the SPV is a legal entity that is destined to die for the purpose of the transaction, whereas a company usually tries to last forever).

The dynamic nature of the creditworthiness of structured deals needs to be continuously monitored as time goes on. Continuing with the bridge analogy, if there is a “rating” for bridges, do we blindly believe in it when the geophysical environment around the bridge is constantly changing — say an earthquake occurs at one end of it, and the frequency of storms in the area increases significantly due to global warming? Do we need to reassess its structure on that basis? Sure we do! We would also like to know what kind of vehicles cross it, what are their speeds, what causes accidents and delays, etc. In structured deals, we have new information almost every day. Given the current advances in information technology, having a continued “pulse” revealing the true dynamic state of the deal is not impossible. The monthly trustee report does not provide loan level details or an indication of how solid the structure of the deal remains as things evolve. Yes, sometimes it shows whether certain triggers were hit or not, but usually by then it is already too late. It is definitely possible to come up with system that can give investors detailed information about the performance of underlying loans, and hence the soundness of the deal structure without compromising privacy issues. In fact, we already have a system that addresses some of these issues (see the rest of the R&R website), but due to the “old culture,” these measures have not yet been widely adopted. This is the time to utilize such technologies, because they promote transparency and provide sensible analysis.

We need to develop a new open system that can avoid most if not all of the palpable conflict of interest between the issuer (seller) and rating agencies. The market needs a truly independent value arbiter. It can be a for‐profit entity, but at least it should be on the right side of the deal. The current market trend indicates that the pronouncements of the existing rating agencies will matter less and less. As R&R’s Sylvain Raynes has said, “The battleground will shift away from ratings to a more mature and scientific process based on consulting or other bodies.” A major Exchange could be a good candidate for providing “better ratings” for its participants. It would do a better job at serving its members because it would have better incentives to be fair when assessing the true risks of a deal. It might have better access to the underlying data as well, because many players would be its members. An exchange would also serve as a natural catalyst for transparency among all parties. Any market‘s success depends on credibility and transparency, because that’s what strengthens investor confidence.

One major barrier to the acceptance of structured finance is that some people perceive structured products and securitization as complicated games played by Wall Street crooks (i.e., they do not believe there is value in it). Some even regard securitization as no more than a scam. This negative reputation is unfortunate yet understandable given the rating conflict snafu and “six degree of separation” between the investor and ultimate borrower. Also, when securitization techniques are first exported to countries such as China, they are used mainly to deal with bad bank debt.

Technology can certainly be effective in supporting transparency, as I mentioned earlier, but the most fundamental means of promoting confidence in the market is through education (i.e., equip everyone with the right knowledge). Ever wonder why the US Security and Exchange Commission was so successful and effective when it first opened its doors, yet more recently has missed Enron, Madoff and the subprime scandal? It’s because the first chairman of the SEC, Joseph Kennedy (president JFK’s father), who was hand-picked by President Roosevelt, was himself a master of all the tricks. What about today’s regulators? I can’t say how well versed the SEC is in structured finance. I do know that they don’t have a department of structured finance; they have a division call Corporate Finance, which includes structured finance — i.e., they have the same mentality as the rating agencies, treating structured finance the same as corporate finance even though they are very different animals. No wonder they missed Enron, in which structured finance was the major theme of the swindle.

The subprime problem grew into a financial crisis because there is no regulation in many key parts of the market; regulators don’t regulate these sectors because they don’t understand them. For example, insurance-like products such as CDSs need to be regulated somehow (as Andrew Davidson observed, “In the simplest terms, what went wrong in the subprime mortgage market is that the people responsible for making loans had too little financial interest in the performance of those loans and the people with financial interest in the loans had too little involvement in the how the loans were made.”), and again, an exchange or clearing house could make a substantial contribution in this regard. Regulators believe the market will always self-correct sooner or later. It surely did this time, in such a grand fashion that regulators decided to intervene and try to bail it out instead.

Regulation without understanding assures that efforts will be in vain. As early as two decade ago, in July 1988, the Basle Accord was established for the very purpose of protecting the banking system against the credit risk inherent in derivatives transactions, and in March 1995, the Derivatives Policy Group (consisting of Goldman Sachs, Merrill Lynch & Co., CS First Boston Corp., Morgan Stanley, Lehman Bothers, and Salomon Brothers, which together accounted for more than 90% of the derivatives business in the U.S. that was not handled by banks) came out with its “Framework for Voluntary Oversight.” Today four of the six original members of the policy group are gone, and I’ve lost track of how many amendments and extensions have been added to the Accord. A lot of these efforts have been related to disclosure, but when people don’t understand why, what, when, which, where and how the material is disclosed, it doesn’t help. For example, CDS, CDOs, and CDOs made out of CDS are synthesized assets (hence the name synthetic CDOs). One can think of CDS as a bet between two people about a reference entity. The reference entity need not be one of the two people in the bet; it could be a third party. CDOs can be thought of as something like a mini bank, yet they are not regulated as banks are. Since in many cases a CDS or CDO is bespoke and synthetic, how and what disclosure means becomes a very tricky game.

Given that we’re paying a high price for this lesson now, we’d better learn it well. Both regulators and we ourselves need to be educated about the truth of structured finance: for example, securitization is not equivalent to structured finance; it’s part of structured finance, just as Statistics is part of Mathematics without being all there is to mathematics.

Securitization is not a magic wand for bad assets, although it is often an effective means of dealing with them. Structured finance is a great venue for discovering value and creating efficiency in the capital market by reducing the cost of funding and diversifying certain risks — notice that it does not eliminate risk; what it does is manage, transfer and customize the risk quantitatively. Products such as certain types of CDOs are highly leveraged because in essence they are a securitization of an existing securitization.

It is shocking that such an important subject, which has a significant impact on the global economy, is completely opaque to all but a select few. This state of affairs must change. My brother, a Chicago University-educated PhD, told me that he learned a lot about economics from a cartoon on the subprime crisis. This situation is actually not funny, and I’m sure we can do better.

The option market started by the CBOT succeeded because the Black‐Scholes formula is universally accepted, despite some of its less realistic assumptions such as the lognormal return process, constant risk‐free interest rates, etc. It’s close enough, and it’s appropriate. Most importantly, it’s a formula that people know how to use. We don’t have this luxury in structured finance; there are so many more moving parts, very few analytical results, and a lot of statistics. Monte Carlo simulation is still the dominant method in the analytical processes.

It’s a matter of public record that many exchanges, primarily in the world’s financial centers, are gearing up to provide clearing and settlement services for CDS. Creating an exchange for securitization is feasible because most of the IT and risk management infrastructure is already in place to prevent future disasters like the one we’ve been dealing with. Also, the task of deciding the right margin to create an attractive market for bespoke CDS would not be trivial, and working closely with a regulator while fulfilling market participants’ needs is a delicate balance act.

In order to make sensible choices for parameters, we need a clear understanding of the underlying assets and structure of a deal. “A little knowledge is a dangerous thing” — this is particularly true in structured finance. Instead of taking comfort from some letters uttered by the rating agencies, we need to ask fundamental questions about the underlying assets and our assumptions about them. For example, if one asks where the credit enhancement of a deal comes from, and the answer is a large insurance company (AIG?), then one might want to know the assumptions they used in calculating the premium. If credit enhancement is generated internally, say through subordination, then one should ask how the amount (or the percentage, detachment point) was calculated, and what assumptions lay behind these calculations. Likewise, in a waterfall, the difference between “pro rata” and “pro rata pari passu” seems subtle but has an enormous effect on whether subordination actually occurs. If it is an overcollateralization deal, how does the figure for an “OC holiday” come about? The usual answer I get is: “That’s the way it is; that’s just how these deals work; that is the market convention, based on experience…because historically this is how we protect senior tranches…etc.” What experience? Based on the market experience prior to 2006, housing price were never supposed to drop! Why has it become a market convention? There must be a rationale behind the convention. This shows that even people in the field often lack a clear understanding of the process. A structured deal can be complicated and tortuous, but it need not to be so. And that certainly should not be an excuse for not inquiring into the details. With the right education, when an investment bank says “due diligence,” we can say “caveat emptor,” because we know what to beware of. The devil is indeed in the details.

I’m tired of hearing that every deal has its own master and supplementary glossaries. One great way to improve transparency and reduce initial disorientation for people who first encounter the subject is to go “green” on documentations and standardization. In terms of both material and language, most documents in structured finance are designed so that ordinary people (i.e., those who are not structured finance lawyers) are unable to read them. Why does everything have to be stated in words (particularly in multi‐page passive‐voice run‐on sentences)? To claim that the indecipherable language only serves the job security of lawyers might be too extreme, but if you tell me it serves to protect investor interest, that’s simply preposterous. There must be a better way. Based on my experience as a trustee, a simple mathematical formula is often more precise and unambiguous than three pages of legal tautology. If a significant portion of finance is quantitative in nature, why can’t mathematical formulas be written into crucial documents such as a Prospectus Supplement? Why not give a few numerical examples relating to the assumptions and their implications? In addition, many aspects of securitization can and should be standardized (for example, the way losses, recoveries and defaults are recognized). Standardization would create efficiency; the last thing we need is to complicate matters further through excessive legal mumbo jumbo. As with any subject, appreciation comes from understanding; in any financial innovation, barriers of entry erode as familiarity increases and uncertainty is reduced.

For both the best and the worst of reasons, structured finance is becoming an essential skillset and a critical element in corporate governance. Regulators, market participants, and the general investment public all need to acquire knowledge of the principles and methodologies of structured finance, and gain an understanding of its virtues and limits. As of now, there is no standard-setting body for structured finance, and China has a great opportunity to fill this void.

Like everywhere else, China ‘s economy has been affected by the subprime crisis in the US, but its does not yet have its own subprime problems, and no one wants to see China repeat the same episode. A Center of Excellence in Structured Finance sponsored by a major China exchange could greatly benefit everyone and contribute significantly to the healthy growth of the global economy. If Japan could beat the US in automotive innovation, why can’t China beat the US in financial innovation? This is an opportune moment as well, given China‘s 400 billion yuan economic rescue package. There is value to extract and efficiency to attain.

All of the suggested courses of actions — a new rating paradigm, appropriate regulation, clear documentation, standardization, and most importantly, pervasive education — are geared toward promoting transparency in structured finance, and with it investor confidence and market liquidity. These steps are simple and feasible, and while far from trivial, as US president Barack Obama has said, “If you’re walking down the right path and you’re willing to keep walking, eventually you’ll make progress.” So let’s hope we’re walking on the right path while enjoying the mandate of Chinese President Hu Jintao‘s brilliant motto, “Don’t mess around!”


Malcolm Moore, ” China’s 10 steps to boost its economy,” The Telegraph, 10 November 2008

Ann Rutledge and Sylvain Raynes, The Analysis of Structured Securities

Carrick Mollenkamp and Serena Ng, “Wall Street Wizardry Amplified Credit Crisis, The Wall Street Journal, 27 December 2007

Ann Rutledge, “ABS Analysis: a rigorous approach,” RMA Journal, December 2004

Paul Kix, “The Man Who Would Save the Economy, Boston Magazine, February 2008

Neil Roland, “SEC plan to lessen credit raters’ clout hemmed in by other regulators,” Financial Week, 3 July 2008

Andrew Davidson, “Six Degrees of Separation,” Securitization.net, August 2007

All rights reserved.