Wired has an article called Recipe for Disaster: The Formula That Killed Wall Street. It's about how a mathematician came up with an idea of how to easily quantify the coupling of risks in individual cases that investors want to know so they can make better decisions. For exmaple, you know that one investment has a given chance of going into default, and that one debt arrangement is bundled with others to make a bigger financial instrument in which you could put money. If one starts going badly, are others also more likely to head south, or are the individual risks really independent of each other? You need to know the answer if you're going to intelligently understand the risk.

The solution was a formula using a technique called a Gaussian copula function. All of Wall Street rejoiced because suddenly it was easy to calculate risk without waiting for historic data:

The effect on the securitization market was electric. Armed with Li's formula, Wall Street's quants saw a new world of possibilities. And the first thing they did was start creating a huge number of brand-new triple-A securities. Using Li's copula approach meant that ratings agencies like Moody's—or anybody wanting to model the risk of a tranche—no longer needed to puzzle over the underlying securities. All they needed was that correlation number, and out would come a rating telling them how safe or risky the tranche was.

And then, eventually, the market imploded. But when you look at the approach the mathematician took - examining the historic prices of credit default swaps - you start to see just how stupidly so many in finance acted:

The damage was foreseeable and, in fact, foreseen. In 1998, before Li had even invented his copula function, Paul Wilmott wrote that "the correlations between financial quantities are notoriously unstable." Wilmott, a quantitative-finance consultant and lecturer, argued that no theory should be built on such unpredictable parameters. And he wasn't alone. During the boom years, everybody could reel off reasons why the Gaussian copula function wasn't perfect. Li's approach made no allowance for unpredictability: It assumed that correlation was a constant rather than something mercurial. Investment banks would regularly phone Stanford's Duffie and ask him to come in and talk to them about exactly what Li's copula was. Every time, he would warn them that it was not suitable for use in risk management or valuation.

The article focuses on how you cannot count on the correlation of financial securities, because risks that seem out of sync one day can suddenly all manifest at the same time.

I'll put it a little differently than stated in the article becuase I think there's an even more fundamental point that financiers missed. Like any market, CDO purchases largely move on emotion - it's one of those indisputably human activities. When people think they are safe, they will do the most astoundingly stupid things because they simply don't perceive danger.

CDO prices are an accurate historic measurement of what people thought risk was, not of the actual inherent risk of the underlying investment on which the CDO was taken. So all of these investment decisions were made based on looking at people's perception of risk, whether right or wrong, and not the actual risk. No wonder everything blew up. It was like betting on the results of a card game when you weren't one of the players, didn't know their history of success, and couldn't see any of the cards. Here's the scary part: the people who didn't notice the difference are the ones still in charge.

The Next Big Financial Debacle?

Some experts on a risk management mailing list I'm on are guessing that there's a new finanical bugaboo on the horizon, credit default swaps derivatives, that could make the sub-prime problems look like small change. Two parties enter into an agreement. One pays the other a fixed sum periodically though a coupon bond - which means that the second has purchased the bond from the first. The second doesn't have to pay anything else, and keeps collecting interest over the life of the bond (and, presumably, recaptures the initial investment) unless some pre-determined credit event - such as a debt restructuring, bankruptcy, or a drop in credit rating - occurs. In such a situation, the second party pays a fixed sum to the first and then the entire arrangement is terminated.

The derivates can become a form of credit insurance, with the second party being the insurer. However, they can also allow third parties to speculate on an entity's ability to repay a debt - legalized gambling, I think it should be called. So long as you only have the occasional credit explosion, the system can work. But what happens when you have one comapny after another in trouble? Now you have to wonder whether the insurers actually have the money to cough up - and some of the big players in that market are the very financial institutions that have been pouring money out because of the sub-prime mess.

According to an article in Wikipedia (link above), these are the most widely traded credit derivative product. More so than mortgage-backed bonds. The outstanding swaps currently top $46 trillion - with a t. To add some perspective, the US stock market is only $22 trillion, with mortgate securites hitting a mere $7 trillion. And an article in Bloomberg notes that this is the fastest-growing form of derivative on the market.

This could bear out Warren Buffet's remarks from 2003, in which he called derivatives in general "financial weapons of mass destruction" that could hurt the entire global financial system, and not just the people involved with the specific transactions.

A Sudden Realization About the Sub-Prime Crisis

I've written a fair amount about this topic, so this will be short. (Having impending deadlines, like other forms of death, concentrates the mind wonderfully.) But when I started thinking about how this debacle came to being, I wrote the following:

By combining pools of mortgages with rising housing prices, lenders were able to wash off the risk because the failure of some percentage of borrowers still left the pool safely covered.

That was the explanation I had read, but something bothered me about it. It didn't quite make sense that some juggling could improve the credit rating that much. After all, to really cover the potential default rates, you'd probably need to add a significant number of "good" mortgages that would be unlikely to default so the return on the investment had a reasonable chance of occurring. But then the default rates shouldn't have had that kind of impact. And yet, it seemed that the derivative securities were largely based on poor credit lending. How did bundling them get better ratings than the individual loans would have?

After paying attention to more reporting on the subject, I think I now understand. The rating agencies abdicated any ethical or moral responsibility to give an honest opinion on the derivatives because they are paid by the very financial institutions that were issuing them. Unfortunately, investors pay significant heed to these ratings and often pass on doing further due diligence. I could understand that from individuals who are intimidated by understanding financial matters. But these derivatives could never have taken off without significant institutional investor participation. What happened here? Don't these large organizations that hire lots of brainy people actually do their own thinking? Actually, many rely on the opinions of others far more than you might thing, certainly in proxy voting issues, as I learned in writing an article for Corporate Secretary.

So many individual investors, putting their money into money markets, pension funds, and other aggregations of cash might be taking a larger risk than they realize, because they don't always know who's really making the decisions. Where has the financial media been through all this? This story has an Enron-like cast, with lots of people writing in awe of the clever financing and no one pointing out that the Emperor ruled in the buff.

Taking Bets on New Internet Bubble

John Dvorak, who pretty much came in with the PC and whose death will probably be a harbinger of Computing As We Know It, wrote an interesting article on the concept of Web 2.0 being the next bubble waiting to happen:

Every single person working in the media today who experienced the dot-com bubble in 1999 to 2000 believes that we are going through the exact same process and can expect the exact same results—a bust. It's déjà vu all over again. And since this moment in time is only the beginning of the cycle, the best nuttiness has yet to emerge. Nevertheless, this is not to say that a lot of nuttiness hasn't already happened.

He then suggests that there is a continuous series of bubbles in high tech, each one worse than the other. I don't know about the progressive deterioration, but successive bubbles? Absolutely - as in any part of industry. There was the 17th century tulip bubble. I remember the paperless office mania, the total quality hysteria (remember TQM?). There is the current housing bubble, along with the bubble in mortgage-based security derivatives. Every multi-level-marketing scheme is, to some extent at least, a bubble. Then there are the countless ones that no one ever names.

A bubble is a form of economic hysteria that comes about when a small group of people find a way to make a whole lot of money at something that seems to promise the same for many others. There are economists who argue that the term bubble is inexact and sloppy thinking.

So forget economics and focus on psychology. Call this phenomenon the free lunch, or something from nothing, syndrome. People become convinced that they're going to become fabulously wealthy, that a trend will continue ad infinitum, that no one will tire of buying a given something, that everyone will want it, and that enough available money is around to fund the whole rollercoaster ride. That's a human failing that I suspect is as old as the species. No wonder we keep falling into manias - we're too anxious for the free payoff to realize that we're two-legged lemmings and that there's a cliff looming ahead.