Kent Osband writes:
About 15 years ago you kindly linked an article I wrote on “rational turbulence.” I’d like to let you know that I have recently summarized much more research along these lines in a short book Rationally Turbulent Expectations.
I have published it as cheaply as color printing allows and also posted all chapters for free on ssrn, starting with this overview.
The main finding, summarized in the first few pages of Chapter 4, is that–once we allow for even tiny doubts about the the stability of an iid process–Bayesian learning has calm and turbulent phases, with fast learning more turbulent. It explains why differences in opinion between two reasonable people often widen before they narrow. I think this deserves broader attention in that people can learn to disagree more respectfully.
Hardly anyone will listen to me but many listen to you, so I am hopeful you will persuade yourself of this quickly and help persuade others.
I clicked through and took a look. Lots of things there resonate with various ideas I’ve discussed various times without ever fully thinking them through. For example:
“Words like ‘ahead’ and ‘forward’ that point to space in front of us also point to future time. However, that isn’t the only way to align directions”: This reminds me of the idea discussed here that there’s a logic of causality going forward in time and a logic of inference going backward in time. Generative models in statistics are a way of going from one of these to the other.
“Most surprises are outliers from a still intact trend. We tweak the next round of forecasts and move on. Occasionally the surprises get under our skin. They shock us less by their size than their persistence. They make us suspect that what we thought of as a rare outlier is now the new norm”: This reminds me of the ideas of Shewhart, Deming, etc., on quality control, an approach to statistics which I think is important and underrated.
“Contemporary finance theory rests on an unstable truce between two opposing schools. . . . The Rational Expectations school treats the market as a knowledge machine that assesses risks correctly and prices them appropriately. The Behavioral Finance school treats the market as a ship of fools prone to long stretches of complacency and short bouts of panic”: This reminds me of what I’ve called the two modes of reasoning in microeconomics: people are sometimes considered to be rational, so that the role of economists is to observe and analyze behavior and, from that, deduce values and motivations; and sometimes people are considered to be irrational, and the role of economists is to set them straight. Either way, the economist (or “freakonomist”) is portrayed as a culture hero, either in protecting us from pinheaded academics who don’t trust the ordinary Joe to make his own damn decisions, or in helping people avoid deadweight losses all around them.
Osband frames the economy in terms of turbulence, which fits in well with the idea that economics occurs on the phase transition of equilibrium. “Turbulence” seems like an appropriate term.
I don’t have the energy to read the whole book–I guess there’s an economic message there too!–so I can’t say more, but I’m happy to spread the word, and if you’re interested you can take a look into it yourself.