Let’s start with a question: Are you an above-average driver?
Chances are, about 85% of you answered yes. This means that 35% percent of the population answered this question incorrectly. Yet, even knowing that information, you are likely thinking to yourself “who are these people that are overconfident of their driving ability, because it’s surely not me.”
I mention this because, in past newsletters, we have discussed how the typical investor underperforms the market (link). I’m guessing you read that information and thought… other people underperform the market, not me (just like other people are below average drivers, not me). I have a challenge to see if you’re right. I’ve compiled historical returns from various mutual funds. I’ll provide the past performance figures and you pick the one that you expect to perform best (for now, just assume that I have done the research on these funds and they can be considered a wise investment).
Which fund did you pick? My guess is 90% of you picked B, E or I (highlighted below). In the next table, I show the performance of each of these funds for the ten years following the date that you made your selection.
At this point, many of you probably have buyer’s remorse. Funds B, E and I didn’t perform as well as you might have expected them to. If you could select again, which funds would you select? Funds A, D and H seem ideal given perfect hindsight. Yet personally, I never would have picked those funds, especially not A or D because they performed so poorly in the past.
This exercise is designed to explain a concept common in investing (and life) called mean reversion. Investopedia defines mean reversion as “the assumption that a stock's price will tend to move to the average price over time” and you can see this happening in the above example. Periods of underperformance are often followed by outperformance and vis-versa. I’ve include a brief three-minute video with some additional background on mean reversion, below.
Now to illustrate this point, I took a few liberties with the facts in my example above. The returns I used were actually provided were decade long returns for large stocks (via James O'Shaughnessy). Here is what the full chart looks like…
A graph helps to illustrate the up and down (mean reversion like) nature of stocks since 1910.
In order to clarify this concept, let’s discuss an example of mean reversion that doesn’t involve investing. Assume I play basketball and I’ve made 30% of the first one thousand three point shots I’ve attempted (shots 1 – 1000). Over the next one hundred three point shots I take (shots 1000 to 1100) I only make 17% of them. In the subsequent one hundred three point shots I attempt (shots 1100 to 1200), would you expect me to:
- Make 30% of the shots I attempt
- Make >30% of the shots I attempt
- Make <30% of the shots I attempt
We know if I take a large number of shots I should shoot 30%. In our most recent smaller sample (of 100 shots) I only shot 17%. Therefore, I’d suggest I’m most likely to shoot >30% over the next one hundred shots (which would push my average shooting percentage closer to 30%, thus reverting to the mean). Of course, there are no guarentees what will happen in the next hundred shots, any range of outcomes is possible.
So now that we have all had a quick introduction to mean reversion – let’s try this again. The following graph is from a value focused fund I manage. Between August 2015 and August 2016 it outperformed the S&P 500 (blue line). Can you guess what happened next?
From Aug 2016 and Aug 2017 it underperformed the S&P 500 (reverted to the mean).
Finally, let’s look at the performance of the fund since inception. Periods of underperformance (highlighted by triangles) made outperformance (from that point in time) more likely. Similarly, periods of outperformance (highlighted by circles) were often followed by periods of underperformance.
The takeaway from this post shouldn’t be a hard and fast rule – not everything reverts to the mean. Great companies, great funds and great housing markets can outperform their benchmarks for significant periods… and some companies (or even stock markets) tumble down all the way to zero. Yet, I’ve found using the concept of mean reversion to frame future expectations as a helpful tool. Hopefully you will too.
We are looking for an editor to edit posts like this... Interested? Email mccauleycapital@outlook.com