How We Use Past Performance

Yesterday’s insight about the Morningstar Manager of the Year generated a lot of interest and feedback from readers.  Thank you!  I appreciate your replies and make it a point to respond to all of them.

While we’re on the subject, I also want to thank you for completing the satisfaction survey that we sent 10 days ago.

I am pleased to report that when you were asked whether you would recommend Daily Insights to a friend or colleague on a scale of 1-10, a full 66 percent of you answered with a nine or ten.

Only 12 percent effectively said you would not recommend Daily Insights with a ranking of 1-6, which is not too shabby and leaves me with some room for improvement.

Also, many of you left comments that were largely positive with some constructive criticism and suggestions for improvements.   For example, a few people asked for links to source material and I will make an effort on that front.

I digress.

Regarding yesterday’s article, I did watch a video today from Morningstar (click here, although you may have to subscribe), where they explicitly said that winning wasn’t a curse and that winning managers had outperformed on a risk-adjusted basis after the big win.

I don’t buy it, but thought I should mention it until I get a chance to run all of the numbers myself.

A few clients basically said, ‘well, if you don’t look at past performance, how do you do it?’  Or similarly, ‘you guys look at past performance all of the time.’

These are good questions, and I didn’t mean to bury the lead, but here are a few thoughts to consider.

Let me start with the second question.  There is no doubt that we look at past performance.  We look at it, inspect it, scrutinize it, probe it and generally run it through the ringer.

Here’s the difference: we aren’t looking at managers: we’re looking at assets.  We have decades of past performance on the return of ‘the market’ whether you define that is a commercial index or an academic index like CRSP 1-10.

We’re not trying to understand whether Bill Miller, Bruce Berkowitz or Bill Gross has skill because it’s just too unpredictable and even if they’ve been managing assets for a few decades, there often just isn’t enough data, statistically speaking.

In the final analysis, I don’t think we can really see whether any of them will outperform in the future like they did in the past.

So, I don’t believe in manager skill – well, I believe it exists, I just don’t think you can find it by looking at past performance.

What I do believe in, and what I do think you can find answers in past performance is how different assets behave.

The simplest example is that stocks have outperformed bonds over the long run.  That’s an analysis based on past performance – and we and academics have run those numbers through the ringer since the 1960s.

In addition to the data, there’s a good theory for why stocks should outperform bonds: an ownership interest in a business has potentially unlimited gains and risk of total loss.  A bond, depending on who issues it, has limited downside and no upside beyond the yield at the time of purchase.

Within stocks and bonds, there are a variety of different types of assets.  I alluded to one in the last paragraph – corporate bonds are different than government bonds because government is much more likely to make the interest and principal payments, making them less risky.

Stocks can be broken into different groups as well: cheap versus expensive, large versus small, etc.  We look at past performance to give us some idea of how these might perform in the future in various environments, knowing that we can’t predict the timing or magnitude of how the assets might react to different market environments.

So, yes, we look at past performance to try and understand asset prices, but not to pick the next Peter Lynch (something he failed to do when he retired).