Yahoo! Finance ran an article on Saturday showing how stock analysts struggle to understand Warren Buffett’s company, Berkshire Hathaway. Here’s the quote that hit me like a freight train:
Mr. Buffett is very good at investing but you never know what he’s going to buy next. I never would have thought he’d buy Precision Castparts – even though after the fact it makes sense. Anticipating what he’s gonna buy next is tough.
I can’t say I’ve ever heard a purer manifestation of the narrative fallacy. Nassim Taleb describes the narrative fallacy in his book The Black Swan this way:
The narrative fallacy addresses our limited ability to look at sequences of facts without weaving an explanation into them…Where this propensity can go wrong is when it increases our impression of understanding.
The long and short of it is, no one can predict what investment Warren Buffett will make next. It’s impossible. There are way, way too many variables.
Yes, stock analysts try to predict these very events. Then, when their predictions fail (if they even make predictions at all), they act surprised. And of course, it makes sense in hindsight. That part of the above quote still kills me: “even though after the fact it makes sense.” Incredible.
That’s what The Black Swan is about: the human compulsion to make predictions, even though our predictions are often laughably wrong. It’s not that we’re necessarily bad predictors. It’s that some things we try to predict are fundamentally unpredictable. We’re not even playing the right game, which means we lose. And when we lose, we often lose big.
I’m engrossed by the whole topic of prediction, mostly because of Taleb’s book. Two parts of the prediction puzzle intrigue me most: (1) the compulsion to make a prediction in the first place, and (2) the action or decision that is made because of the prediction.
We predict things that can’t be predicted (like what interest rates will do next year). And then, we can’t even identify a meaningful action or decision that our prediction would affect. In a ton of cases, we’d do exactly the same thing, no matter what our prediction says.
In this sense, it doesn’t matter whether we predicted increasing interest rates or decreasing interest rates. Our actions, as a bank, will unfold depending on whether interest rates actually rise or actually fall. The prediction offers us false confidence. We pay an opportunity cost. We spend resources fruitlessly predicting outcomes, when we could have spent those resources preparing for contingencies.
I’m coming down a little hard on predictions here. They are sometimes useful. Are you trying to predict whether a given mechanical assembly will survive real world wear and tear? Great! You have a lot of models and experimental tools available to help you make a useful prediction.
If you’re making predictions in well-ordered, deterministic environments, then you’re in good shape. If you’re trying to predict what investment decisions another human being will make…you’re not in good shape.
What kinds of things do you predict, as part of your job? How often are you surprised by reality, compared with what you predicted? Do you tend to make sense of things after they occur, but struggle to make a confident prediction in advance? Beware…you might be succumbing to the narrative fallacy.