Risk Models and Portfolio Theory

Most investors probably have no idea what a risk model is but anyone who has ever worked with a stock broker probably has heard that smart investing requires a diversified portfolio. They've been told to invest in different segments of the market such as the U.S., Europe or Asian stock markets, maybe a small investment in commodities or REITS. Divide your money between growth and value stocks. Small-cap, mid-cap and large-cap stocks. Brokers probably brought out colorful charts showing how different markets have low correlation so that if one goes down the others don't.

Large financial institutions like banks and insurance companies developed complicated computer programs to slice and dice their investments to maximize their gains while minimizing their risks. And, as we now know it was all a house of cards that fell down - and here's why:

1. Correlation is based on statistics not cause and effect. In other words, just because two markets have historically behaved differently, there is no logic, science or force that will make them behave differently tomorrow.

2. When panic hits, everything is correlated. Proof: October, 2008.

3. Risk models are based on perfect information. There is an implied assumption that all information is known. Obviously, this is not true. In the current crisis, we still have no idea what the true magnitude of mortgage loss will be. On a much smaller scale, when evaluating the potential for investing in a company, do we ever really know all relevant financial detail of that company. Case in point - Enron. (I can hear someone saying: "Oh, but Enron was fraud. That doesn't count." I answer: Has there ever been a time where there has not been fraud in the financial markets?) Read this article. The father of modern portfolio theory, Harry Markowitz, says the only problem is transparency and proper oversight. Well maybe, but until that happens (which it won't) risk will never be mitigated in times of panic.

4. People are not rational. Read this article about AIG's risk models. A Yale professor named Gary Gordon developed these models. Here's the last paragraph of the article:

"On a rainy morning last week, Mr. Gordon briefly discussed with his Yale students how perplexing the struggles of the financial world have become. About 30 graduate students listened as Mr. Gordon lamented how problems in one sector caused investors to question value all across the board. Said Mr. Gordon: "There doesn't seem to be fundamental reason why.""

In other words, people are not making decisions in the way Mr. Gordon thinks they should. Risk models assume that investors make decisions strictly based on their own financial self-interest. However, people don't always behave rationally and for a risk model to even come close to working it has to model irrational human behavior and they don't.

Irrational human behavior always wins out both with upside bubbles and downside panics.