Hurricanes bring out the worst in people. No, I don’t mean looting or storming Costco like it’s 12:01 on Black Friday. I mean it brings out the worst analytical side of people. Hurricanes take away our ability to think rationally and return us to earlier times where we attributed things we didn’t understand to the Gods being angry with us or ancient curses.
For those unaware, confirmation bias is “the tendency to interpret new evidence as confirmation of one’s existing beliefs or theories”. In other words, we know what we want to believe so we seek out facts that seem to support our predetermined conclusion. Often, these “facts” are subject to interpretation and someone with an opposite confirmation bias will come to an exactly opposite conclusion.
This behavior has been programmed in humans likely from the beginning. At some point in our evolutionary history, it served a purpose and it didn’t do enough harm to lead us to extinction. Some argue that the purpose of this behavior is about leadership. One gains status by being persuasive and an effective orator. Confidence in one’s argument tends to boost persuasiveness.
Thus, the social gain from being a strong leader overcomes the loss from being wrong. Our politicians clearly understand this lesson! Spin matters more than being right and fact checkers don’t change many minds.
Studies have even shown (though I can’t find them at the moment) that the worst way to change someone’s mind is to show them conclusive evidence they are wrong. This just makes people more obstinate. It is better to show them why they not be entirely right and gradually get them to move their position.
The “Hurricane Trade”
There are many examples of confirmation bias in investing. Today, though the topic is hurricanes, so I will just focus on one. Every June, a bunch of sell side analysts publish research to buy reinsurers for the “hurricane trade”.
The idea is that almost every year reinsurers outperform for the summer because, if there is no hurricane, they go up since they will beat earnings. However, they still go up when there is a hurricane because people will want to buy the stocks for better pricing. At least that is the theory.
The problem is there are very few data points to support this. When the idea first started circulating, there were just a handful of years of data. Even now, 15-20 summers isn’t all that significant, especially given we had a long run of years without any significant landfalling hurricanes.
The analysis also was very liberal in counting “wins”. For example, if the market was down as long as the reinsurers were down less than the market, than that was a “win”. Never mind that several of those summers had significant effects from the financial crisis and the Euro scare. Reinsurers would have done better in those environments regardless of hurricanes.
The idea of shaping the data in a way to make the evidence look more compelling is classic confirmation bias. The analysts seek data points to support a conclusion they already believe rather than fully analyze the data set to see how much of the performance is really due to the occurrence or lack of a hurricane each year.
Next to politics, climate change may be the greatest area of confirmation bias in our society. Believers see every adverse weather event as proof of the impact of man-made warming while deniers discredit every study that shows climate change is growing more severe. There are very few people in the middle who are dispassionate about the debate and are motivated to get the facts right.
Naturally, every time there is a major hurricane, there will be a contingent blaming it on climate change. “Hurricane Twitter” is a community of meteorologists who tweet often during storms that can be very informative. One of them went into great detail today refuting the confirmation bias that tends to show up around these events.
This is actually part of a very long string of tweets. If you click on the above and scroll up, you can go to the beginning of the argument. Scroll down and you come to what’s below.
There is even more after this, but hopefully you see the point. The Times writer has seen an increase in severe hurricanes and jumped to the conclusion it is cause by climate change without any evidence.
He is clearly coming to this conclusion because of confirmation bias. But does it matter? Recall the earlier point that the theory for why confirmation bias has persisted in our species is because it is effective…at winning arguments.
Will most people who already believe in climate change question the author’s “facts” or cheer him on for agreeing with them? While the deniers will still disagree with the article (remember, even absolute facts don’t convince people), I think it’s safe to say the Times has more readers who believe in climate change than don’t.
Thus, he will get lots of clicks and his editor will think he is doing a great job. There is no penalty for being wrong, as long as you don’t mind that an obscure corner of Twitter calls you out.
The Sad Irony
The recognition of our cognitive flaws has become better understood over the last 20+ years. Many of these terms (confirmation bias, anchoring, etc.) have started to become commonplace. Yet, at the same time we have become more self-aware, our country as a whole has become more polarized and thus lapsed into deeper use of these flawed practices. Maybe we were better off not knowing?