Quote:
Originally Posted by BYU71
This is truly a question I will have to trust you will give me an unbias answer too. I don't know the answer. When we go back to long term capital, even though those guys were not using portfolio theory, they were using some sort of mathmatical theory to come up with all there conclusions on how to invest weren't they? I thought they were Nobel Economics winners from Stanford. Someone told me they were actually MIT.
|
LTCM wasn't being run on a daily basis by academics, but it was, like most quant funds, relying on academic models and being run by academically trained people. There were a couple of famous academics involved with LTCM to some extent: Myron Scholes (Stanford) and Bob Merton (Harvard Business School). Both won the Nobel Prize for the development of the Black-Scholes option pricing formula (Black and Scholes were at Chicago when thet developed the formula).I don't know how heavy their day to day involvement was but I would certainly guess that they help developed some of the specific strategies.
Math wasn't those guys problem. It was the economics. They were using math but you can't say what they did was just math. They made some assumptions about the way the world should behave (all of which seemed pretty reasonable). In late 1998 the world didn't behave that way at all. Making assumptions about behavior and modeling how people behave is economics.
Quote:
|
While math isn't flawed, couldn't how you use the math and conclusions you come from in using the math be flawed?
|
Yes, people can use portfolio theory in a very flawed way potentially. However, what is usually happening is that they are overlaying additional economic assumptions on the basic theory. Its really those additional assumptions that can go wrong. Portfolio theory is a very general framework. There is no one "right" or "efficient" portfolio in portfolio theory (there are some dumb portfolios in portfolio theory but no one optimal portfolio or even optimal risky portfolio). When people get to one "right" or one "efficient" portfolio, they are adding addition assumptions that may or may not be good.
That said base portfolio theory can go wrong as well, but not because of the math. It would go wrong because of the economics. The economics of portfolio theory is that investors care about return and standard deviation is the right measure of risk. If those are bad assumptions then, while the math of portfolio theory is right, its not very useful. You have to buy into those basic economic assumptions. Once you do buy into those assumptions is just math: specifically its quadratic programming. Finally, there are implementation issues (I highlighted one of those early) that can cause problems.
Also, note that because its "just math" portfolio theory doesn't necessarily have strong implications. You can only get so far with relatively benign economic assumptions. At some point you have to take a stand to get strong predictions like, for example, beta being a good measure of risk. Portfolio theory would never by itself give that implication (portfolio theory implies the risk of a security is the following: cov(r_i,r_p) where i is some security and p is your portfolio, which implies the risk of a security may be different for every person because everyone may be holding different portfolios overall). Portfolio theory is just a framework and because it doesn't make strong assumptions it doesn't give precise implications. That's both a strength and a weakness.