My favorite economist is Tyler Cowen. But this isn’t because he has a great track record at predicting things—none of the greats do. Unfortunately, our society places way too much weight on prediction. Even worse, we don’t know how to evaluate whether people have done a good job at predicting. Bubbles are a great example of this phenomenon.
Over at TheMoneyIllusion, I have a new post claiming that bubbles do not exist. They are mythical entities, like ghosts, alien spacecraft, or fiscal multipliers. But I’m confident that I will lose this debate, and in this post I’ll try to explain why.
Consider the following thought experiment. Once during each of the next 99 months, Tyler predicts an asset price bubble will burst. Each time, I write a snarky post explaining that the EMH “proves” that bubbles don’t exist. And let’s also assume that each of Tyler’s 99 predictions turn out to be spectacularly correct. Each time, the asset he singles out quickly goes to zero.
Who wins that argument? Who should win that argument? (Answers: Tyler, me.)
Tyler’s track record would need to incorporate all of his bubble predictions, including this one from 10 years ago when Bitcoin was at $30:
With apologies to Scott Sumner, I say Bitcoin is a bubble. Outside of war and rebellion, do “normal” new currencies behave this way?
I know what you are thinking: “Oh, come on, you can’t seriously suggest that getting 1 out of 100 predictions correct makes you a better forecaster than someone who is completely correct 99 times out of 100?”
Yes, I’m dead serious. For a bubble claims to be meaningful, they must be in some sense useful. A bubble claim is an implied prediction that an asset is overpriced, and is not a good long-term investment. Maybe it will go somewhat higher, but eventually it will crash.
I know what you are thinking: “Even so, in your thought experiment Tyler was correct 99 times out of 100. So he’d be better even using your pragmatic definition of truth.”
Not quite. Investors don’t care about the number of wins and losses; they care about the total return of a portfolio. An investor that invested $30 in Bitcoin, and also invested $30 in each of the 99 other (lousy) assets, would have ended up 10 years later with nearly $30,000, a 10-fold return. Those that took Tyler’s advice and refrained from all 100 of those highly speculative investments would have missed out on those outsized gains.
Thus despite being “wrong” about 99 out of the 100 cases, I would have been right about the broader point—bubble claims are meaningless nonsense. In an efficient market, the vast majority of assets that have the potential to go up 1000-fold will in fact go down sharply.
Now I think you can see why it’s inevitable that I’ll lose this debate. Most people just don’t have brains that are wired to think this way. Put me in a room with a really bright blogger (say Scott Alexander or Matt Yglesias) and I might eventually be able to convince them with my arguments. But there’s no way I’ll ever move the needle for society as a whole. The argument that the guy wrong in 99/100 cases was more correct than the guy who made accurate predictions in 99/100 cases is just too counterintuitive, too much against common sense. It would be like convincing people that the 2008 financial crisis didn’t cause the Great Recession, or that rising interest rates don’t mean that money is getting tighter, or that price gouging is good. There are some things that are just too counterintuitive to be accepted as conventional wisdom. And yet it moves . . .
Bubbles don’t exist, but “bubbles” are never going away.
And I’ve accepted that. But I’ll keep beating my head against the wall, in the slim hope that I’m wrong to be so pessimistic.
PS. This post also has implications for reputations based on predictions. (BTW, I’d argue that Tyler’s reputation is based on his talent as a critic and analyst, not prediction.) Do we overrate pundits that seem to make correct predictions? Do we sufficiently downgrade people after failed forecasts, such as the people who made lots of false financial crisis predictions before getting it right in 2008? Or who got it right in 2008, and then did poorly afterwards? Do we have such a deep innate hunger for soothsayers that we seek Delphic oracles in areas where prediction is essentially impossible, such as the timing of business cycles?
PPS. Ten months ago, Bloomberg told us that there was a 100% chance of recession within the next 12 months. Rest assured, I keep track of these things. Like a dog holding onto a bone, I’ll never let them forget.
PPPS. The title of this post reminds me of a classic pop song from 40 years ago (“99 Red Balloons” in English.). Surprisingly dark lyrics for such a poppy earworm.
