Tech Should Have Seen a Meltdown Coming
The credit crisis looms as the greatest single financial meltdown that any of us are likely to see -- a lot bigger than the subprime debacle, given that the credit default swap derivative market is somewhere close to $70 trillion (yes, with a t). And as the world wakes up to a conclusion that started to be obvious a few years ago, when it was clear that we all couldn't keep generating profits from thin air, we've started to see the predictable concern over the high tech industry.
Well, sorry to be blunt, but anyone in this business should have known that risk exists, even when people who are trying to sell you specialized software and services say that they've found a way around it.
Headlines are starting to show the nervousness about the tech market:
- Will Tech Survive the Wall Street Meltdown?
- Tech Stocks: No Safe Haven
- Tech Stocks Tumble as Turmoil Continues
The big news? "[M]ost Wall Street computer models radically underestimated the risk of the complex mortgage securities." Not only is the current crisis supposed to be the equivalent of the hundred year flood, as many take up act of God metaphors, but firms "chose to program their risk-management systems with overly optimistic assumptions and to feed them oversimplified data" because executives had incentives to ignore risk and didn't want to respond to automated warnings not to set fire to large piles of money.
However, if you've got a background in any engineering or science field, and if your business involves computers, you should have known all this, and known it better, for years. Let's start with the hundred year metaphor. That may be true for a single event, and would roughly translate into a one percent chance of something occurring. But the problem is that there isn't just one possible nightmare scenario. There may be dozens: the S&L crisis, the dot com crisis, the oil crisis (take your choice of which one). Many others as well; I just mentioned a few that have already happened.
Say there are only ten potential world-shaking financial situations of the once-in-a-hundred-years variety. Because the events are independent (otherwise they would form one crisis), the probability of any one of them happening is the sum of their independent probabilities. Suddenly we're facing a ten percent chance of something really big going really wrong. To put it differently, our quick back-of-the-envelope analysis suggests that the world's economy is going to get severely whacked about once a decade.
As far as the overly optimistic assumptions and simplified data, maybe it was all part of a Wall Street executive plot to keep the good times rolling. But as any engineer or scientist could tell you, there are few real world problems so neat and tidy that anyone can easily solve them. Instead, we use models and simplifications, tossing data and factors that are tiny in comparison to the total picture. Engineering is largely the art of sophisticated simplification. So those once-a-century issues get tossed because they're so small and can so complicate potential solutions.
For example, the post author, Saul Hansell, mentions that the computer models would look at several years of trading history instead of focusing on the previous few months.
But since the markets were placid for several years (as mortgage bankers busily lent money to anyone with a pulse), the computers were slow to say that risk had increased as defaults started to rise.It was like a weather forecaster in Houston last weekend talking about the onset of Hurricane Ike by giving the average wind speed for the previous month.That sounds reasonable until you consider the converse. If you look primarily at recent history, then you're jerking this way and that, wasting a lot of money and getting tricked by transient values into doing things that don't make sense with any perspective. There are reasons that those who create financial models simplify. The problem is that the modelers buy their own press, sometimes given in the form of Nobel prizes, and people use the results without any understanding of the dangers and implications.
Last October, Nassim Nicholas Taleb, author of The Black Swan: The Impact of the Highly Improbable, wrote about how the popular modern portfolio theory focuses on factors that "are incompatible with the possibility of those consequential rare events" he calls black swans. This approach to modeling assumed so discontinuities, or sudden jumps, in what they tried to model. That's great so long as the ground remains stable. The stability may last a long time, but eventually there will be tremors.
Most people in the technical arena should have a sense of this. The problem is that they don't invest, or manage companies, based on what they know. Instead, they have been making decisions based on what they wished. Mark Cuban recently replayed some blog posts about his experience in stock trading. He did well because he knew from personal experience the space in which he primarily invested, technology, and he kept trying to learn more about what has actually happening, not what he wished would happen.
People who run and invest in technology businesses generally have the same type of information. They know how things actually work. If they've been putting their knowledge to use, then they've been socking it away for the bad times, they're working to find some advantage in the current climate, and they're not blaming their problems on a bad economy. The present is rarely a shock when you realize that it's the sum of everyone's choices up until now.
Of course there's a more prosaic view: when the entire financial structure of the world is being battered, how could anyone think that there would be a business haven?
Political rally on Wall Street via Flickr site of the Library of Congress, public domain image.