Judging investor performance

One of the hardest things for investors to do is judge the performance of an investment manager.

Most financial periodicals emphasize quarterly, 1, 3 and 5 year records, but is that enough time to look at?

The problem is that there's a lot of noise present in investment returns. What do I mean by noise? Noise is what you hear in between AM radio stations, it's the static you see on TV channels where no content is broadcast. Investors need signal to make good decisions, so paying attention to noise can be a real problem.

Why is there so much noise in investment data? Part of the reason is that many "investors" are really speculators--buying and selling stocks at the drop of a hat instead of purchasing partial ownership of an underlying company. Those traders create noise because they aren't trading on fundamentals in most cases, they are trading based on chart patterns and intuition.

Another reason for noise is that a lot of estimates go into financial reporting. Companies must estimate how much of the credit they extend won't be paid, how much their inventory will be worth, how long their factory and equipment will last, how much they will have to pay into pension plans, etc. It can take years to see what these numbers really turn out to be. That's why so many companies must restate their financials over time.

A third reason is simply human psychology. People tend to over-emphasize recent data and under-emphasize old data. They tend to anchor themselves to a price they paid or a price they want to buy or sell at. They tend to follow the herd instead of gathering and analyzing relevant data. These psychological tendencies lead prices to trend too far in one direction and then the other.

That's only a few of the reasons, but you get the idea. Stock prices, in the short run, reflect a whole bunch of noise and very little signal. You have to look at the data over much longer periods of time to start to see signal. It's like trying to pick a good place to farm based on a weeks worth of rain and temperature data. A week just isn't a long enough period of time to look at--you need to look at YEARS worth of data.

How many years of investment performance do you need to distinguish signal from noise? It depends (I must sound like an economist). But, the absolute minimum you should use is 3 years, it's better to use 5 years, and even better still to use 10 years.

This may seem like a stiff test, especially when a manager has a limited record to observe. But, think of it like you would about picking a place to farm, too little data is just plain gambling, so wait until you have enough information before you act.

Nothing in this blog should be considered investment, financial, tax, or legal advice. The opinions, estimates and projections contained herein are subject to change without notice. Information throughout this blog has been obtained from sources believed to be accurate and reliable, but such accuracy cannot be guaranteed.