A Lesson in Randomness Via the Super Bowl Indicator

The observed trend is likely a coincidence

Senior Quantitative Analyst
Feb 9, 2022 at 7:00 AM
facebook twitter linkedin

The Super Bowl Indicator is a theory that suggests the winner of the Super Bowl can tell us what to expect from stocks over the next year. It’s an outlandish theory that doesn’t work anymore (more on that soon), but I’m going to go on about it for a couple reasons. First, I believe it provides a valuable lesson in randomness. And second, we’re based in Cincinnati, whose Bengals are playing the Los Angeles Rams this weekend, so there has been heightened discussion on the topic, locally.

Super Bowl Indicator

First observed in the late 1970’s, stock markets tended to be bullish after a team from the National Football Conference (NFC) won the Super Bowl, rather than a team from the American Football Conference (AFC). The table below summarizes the return of the Dow Jones Industrial Average (DJI) for the rest of the year after the Super Bowl, depending on which conference the winner happened to come from. When the winner is from the NFC, which would be the Los Angeles Rams this year, the Dow averaged a return of 10.5% for the rest of the year, with 86% of the returns positive. When the winner came from the AFC, or this year the Cincinnati Bengals, the index averaged a gain of about 5% with 70% of the returns positive.

IotW Feb8

I doubt such a dubious indicator would have gained notoriety given the numbers I just presented. The two tables below explain why this indicator gained such prominence and why I said earlier the indicator doesn’t work anymore. This indicator was uncanny in its predictive power for the first 27 Super Bowls. The first table below shows from 1967 through 1994, when an NFC won the Super Bowl, the DJI gained an average of 12.4% over the rest of the year, with 93% of the returns positive. When an AFC won the big game, however, the index averaged a loss of 2% for the rest of the year with just half of the returns positive. That is a wide gap. The second table shows the data over the past 28 Super Bowls and it’s what you would expect given a random indicator. In fact, in the most recent Super Bowls, stocks have done slightly better after an AFC team has won the game.

IotW Feb8 Chart 2

It’s obvious that the winner of the Super Bowl has no effect on the stock market. We know the difference in returns is strictly due to randomness. Even when the difference is huge, like the first 27 Super Bowls. It’s an observed trend that we would not base a trade off of because it’s clearly dumb luck that it turned out the way it did.

A Takeaway on Random Indicators

In terms of random indicators, what about when the study is more plausible? Below is a table from my article last week about the January Barometer. It shows that when January has been positive, stocks tended to do much better for the rest of the year, as compared to when January was negative. It sounds plausible that the first month of the year creates a psychological momentum that carries forward. Perhaps that’s true, but we also need to realize that some of that difference is due to randomness.

IotW Feb8 Chart 3

I’m not saying these types of studies are worthless. Otherwise, I’d be telling you to ignore a good amount of work that I do. I’m telling you to never discount the effects of randomness. If a fluke can lead to the Super Bowl Indicator, then it can also lead to a January Barometer. That’s why it’s important to have a reasonable theory behind any indicator that influences your outlook on the market.

But here’s another thing about randomness. Not only can it create bogus indicators, it can also work in the other direction. Perhaps the January Barometer is even more pronounced than the actual numbers observed, and randomness is making the January return seem less influential. If the theory seems sound, but the numbers aren’t backing it up, keep observing it.


Special Offers from Schaeffer's Trading Partners