By Brad Thomason, CPA
When we go looking for an investment advisor or new fund, a common step is to ask about the track record, the past performance of the thing we’re thinking about getting involved with.
Now, we’ve all heard the endless warnings about past performance not being a guarantee of future results. And we get it. Sort of. Yet it still seems like we ought to find out something about the past before we get involved. That’s part of “due diligence,” right?
Let me point out a few things you might want to keep in mind.
First, the context of the track record matters. You often hear people say things like, “My guy’s done a real good job for me the last few years.” But what does that really mean? If the market was up 20%, and your account went up 20%, how much of that is really attributable to “your guy.” To simply know the degree of movement doesn’t tell you much. A far more important measure is the degree of movement after taking out the part that the market naturally gave to everyone who just showed up.
Or is it? The second thing to remember is that market returns don’t repeat. Like, ever. If you look at twenty years of annual returns, from any period in US history, you are unlikely to see the same rate of return for any two years in the data set. Probably you are going to have twenty different numbers. If the historical record all but proves that the results are going to be different year to year, and the market return has a big impact on the return you experience, how much insight about what you will earn in the future can you really discern from looking at the past? Moreover, to get back to the point made above, even if your return in one year really was meaningfully helped by the actions of the advisor or fund manager, how does that imply that a repeat is coming?
Which brings me to my final point. Understand that track record comparisons in the investing world are fundamentally different than what we do when we look at stats and win/loss records for athletes. Why? Because there is a sameness to sporting events which doesn’t translate into the world of the market.
In a baseball game, for instance, while it may be the case that you have different people playing the positions, different order of pitches, different weather conditions, and other unique features from one game to the next, at the structural level, at the rules level, one game of baseball is exactly like every other one.
The variances are minor, and exist within a rigid framework; meaning that the possible permutations and the range of variance are essentially fixed. This stable environment, in turn, makes it possible to draw some meaningful comparisons. When we look at the batting averages of two players, for instance, we can get some insight as to how each player may perform when presented with essentially identical circumstances in the future. As a measure of relative performance, it really can tell us which guy is the better hitter.
But if we get into a land where the events are not repeating, this mode of analysis loses its ability to tell us much of anything about what to expect.
Now, to be clear, markets are not brand new inventions each new day. There is some sameness from one day to the next. But unfortunately the small amount of sameness has a tendency to draw our attention away from the far more expansive degrees of difference.
The paradox is that markets are always just the composite of people buying or selling. So at the broadest level they seem like an unchanging, rather simple, thing. The problem though is the vast range of possibilities for why people transact, when they transact, and in what quantities. Those ranges are so much broader than steal second/don’t steal second that the variances become the drivers, and the sameness loses most of its meaning. They keep the market from attaining the kind of stable environment status that we wish it had. Instead, it’s just an accumulation of historical events.
Historical events simply don’t repeat themselves in a way that makes standard measures of probability suitable for gaining insight. There was only one Waterloo, and there will only ever be one Waterloo, because Napoleon and Wellington are never going to relive that day again. The ever-changing aspect of global political and economic events, combined with the ever-changing cast of market participants who happen to show up on a particular day, make a day in the life of the market much more similar to Waterloo than just another baseball game.
So before you spend too much time comparing and considering track records, make sure you understand what you are really looking at and what value (if any) it can really offer.
I’ll leave you with this thought, which you may find instructive. Professional traders often look at past studies of various strategies in order to try to gain a sense of how they will perform in the future. But the best traders understand the inherent limitations of this approach, and proceed accordingly.
Many ask the rather simple and general question, “Does this event happen more often than not?” They want to know, is there something to indicate that a particular set up or indicator identifies a greater than or less than 50% proposition? If it does, they often stop the analysis right there and turn their attention to how they will manage the position: when to get in, when to take profit, when to stop out.
In other words, they use the track record to gain a small bit of insight. But they don’t try to take it much further than that. Because they realize there isn’t much more it can tell them. Just because the particular thing happened 63.7% of the time during the test period, doesn’t imply that the same results will be repeated. It just confirms that the thing did in fact happen more than half the time. With that out of the way, they turn their attention to the practical matters of what they will do if the investment works out the way they hope, and what they will do if it doesn’t.
If the pros allocate their time and attention in that manner, well that probably says something worth hearing about the way other investors should approach the matter, too.
Older blogs (2015-2017)