Be Wary of Serial Correlation

MIT’s Andrew Lo:

The key concept here, developed by MIT professor and noted hedge-fund theorist Andrew Lo, is “serial correlation.” Simply put, serial correlation is the degree to which each month’s returns in a fund mirror the results of the month before. A fund that returns the exact same amount every month is perfectly serially correlated. Madoff’s returns were strikingly consistent month after month, year in and year out. That kind of performance—a nice, smooth line going up no matter what the market does—is a really good sign that you should look more closely.

The extraordinary thing that Lo does in the third chapter of his book Hedge Funds, published earlier this year, is to demonstrate mathematically that an excessive degree of serial correlation is a powerful indicator that the holdings of a fund aren’t being reported realistically. What Lo shows from the pattern of historical returns in hedge-fund databases is that when funds’ returns grow too consistent, it is a sign that the investments are either very hard to value accurately and the returns are just guesses, or, worse, that they’ve been manipulated in a way that smoothes them artificially. What Lo creates is a mathematical model for judging what “looks too good to be true.” Lo’s work turns a lot of the conventional thinking about what’s safe on its head. It shows that the evenness that investors have traditionally been taught indicates safety and reliability can actually be the best sign risk is being hidden or that the data are unreliable.


Madoff Dilemma: How Can You Spot A Wall Street Crook?
Mark Gimein
Big Money, Friday, December 12, 2008 – 3:57pm

Hedge Funds
Andrew W. Lo
Princeton University Press (May 18, 2008)

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:

Posted Under