The Albatross of MPT Thinking

By Michael Edesess
February 4, 2014
Go to page 234Next

The January/February issue of the Financial Analysts Journal includes an article titled “My Top 10 Peeves” by Clifford Asness, co-founder of the hedge fund AQR Capital Management and a University of Chicago finance PhD whose advisor was Eugene Fama.

Asness was trained in modern portfolio theory (MPT) and its underlying assumptions. The world of MPT is one of perpetual instantaneousness; the long arc of the future is collapsed into a few short-term measures of unspecified duration. According to MPT, market prices mediate so perfectly between the present and the future that there is no need to think ahead. The only meaningful task is to more completely perfect that mediation – thus standing theory on its head by making the satisfaction of its assumptions the goal of the theory.

Everybody has peeves. Often these derive from an orthodoxy that the peever holds dear, which they perceive as being undermined by the peevees. I’ve noticed, for example, that economist Paul Krugman’s blog reveres the Investment Saving–Liquidity Preference Money Supply (IS-LM) macroeconomic model, which equilibrates demand and supply of goods and money. He is annoyed at those who don’t grasp what he sees as its simple usefulness. As for me, I was trained in the mathematical orthodoxy of probability theory. My pet peeve is that financial mathematicians claim they have learned stochastic process mathematics, but they do not understand its mathematical underpinnings. As a result, they areconfused about simple things like the meaning of “random.”

Many of Asness’ peeves are directed at people who depart from the MPT worldview. In discussing his peeves, I will offer counter-arguments and explain why I think the MPT perspective is flawed.

Kvetching to the choir

Asness’ tone is that of someone delivering a roast to a gathering of an exclusive club. It projects his confidence in his thinking, with the implication that the members of his audience are equally self-assured. But that self-confidence shouldn’t be mistaken for accuracy.

From another viewpoint, Asness’ world is a looking-glass world, in which images are distorted and sense is often nonsense.

Let’s examine a few of Asness’ peeves as examples of why MPT is an insufficient theory,.

Is volatility the same as risk?

Asness’ peeve No. 1 is people who say “‘Volatility’ Is for Misguided Geeks; Risk Is Really the Chance of a ‘Permanent Loss of Capital.’”

I would say something similar: Financial risk is the chance of running out of money, permanently, sometime in the future.

But you can eliminate that risk completely without eliminating high volatility. Here is a simple example, though there are more realistic and more complicated variants.

Suppose you want, with certainty, to spend $500,000 in 30 years – say, to educate your grandchildren-to-be. At current yields, a Treasury bond strip purchased for $160,000 will do it. There’s no risk of not meeting your specific objective (at least not under the conventional assumption that U.S. Treasury bonds carry no default risk).

But if your overriding risk measure is volatility, the risk is the same as that of an equity portfolio. The bond strip’s market price (the price of a zero-coupon bond with a specific maturity date) will fluctuate as much as that of an equity portfolio, even though there’s no risk of failing to meet your objective. Volatility is an inappropriate measure of risk in this case.

Go to page 234Next

Display article as PDF for printing.

Would you like to send this article to a friend?