The Investor’s Fallacy – Why Markets are Never “Due” for Anything
“In God we trust, others must bring data” is a quote attributed to Deming. But Nick Maggiulli is perhaps one of the most ardent followers of that. In this piece, Nick uses long term data to show why the fact that the US stock markets have had a great run in the past decade doesn’t necessarily mean the next decade will see a bust or a mean reversion of sorts. Nick puts up a regression of preceding and ensuing decadal returns to show that there is no evidence of an inverse correlation whatsoever. However, the 20yr return does show an inverse correlation with the ensuing decadal return, albeit with a smaller sample set. Even then, based on the past 20yr returns, the US stock market should give a 4x return for the 20’s if past trends were to hold.
“HHHHH…What is the probability that my sixth flip is also a heads (H)? Assuming the coin is fair (equal likelihood of heads and tails), you already know that the answer is 50%. Because coin flips are an independent process, prior flips have no bearing on future flips.
But it doesn’t feel that way does it? Even if you understand the basics of probability, after seeing five heads in a row, it can feel like a tails is “due” even though you know better. This feeling is known as the gambler’s fallacy and explains why it is hard for humans to understand random processes.
Investors have a similar problem when it comes to thinking about future market returns. I call it the investor’s fallacy. But what makes the investor’s fallacy more difficult than the gambler’s fallacy is that markets are not an independent process. What happened yesterday can affect what happens today. This explains why the very best days in the market and the very worst days tend to occur near each other. The same logic goes for what happened last week, last month, or last decade.
Because of this interdependence of market returns, it is easy for investors to convince themselves that markets can be due for good or bad years. And with the stellar performance in U.S. markets over the previous decade, it feels like a correction is warranted. However, if you examine the data you will realize that this thinking is just as flawed as the person expecting a tails after seeing five heads in a row. There is little to no relationship between prior 10-year returns and growth over the next 10 years.”