Here’s Scott Aaronson:

I love when important decisions fall into the hands of people who constantly second-guess themselves and worry that their own ‘tribe’ might be mistaken, who are curious about science and have a sense of the ironic and absurd.

This is clearly good advice, and got me thinking about what would push me away from market monetarism. For instance, I recently did a post suggesting that negative interest on reserves is expansionary (ceteris paribus). On the other hand, this hypothesis is not easy to test, at least using conventional economic tools, as negative interest on reserves will generally be associated with bad outcomes over any extended period of time, for the same reason that brain surgery is often associated with bad outcomes.

One nice thing about market monetarism is that it comes attached to a clear procedure for refuting the theory. There is a simple answer to the question, “What would make me abandon market monetarism?”


If that seems too cute, let me explain. Suppose that over the next few months I noticed that exchange rates were appreciating on news of larger than expected reductions in interest on reserves. My market monetarist theory predicts that larger than expected cuts in IOR should lead to currency depreciation. If I was seeing one example after another of currency appreciation, all in response to negative IOR surprises, I’d have to reject that particular tenet of market monetarism, at least below the zero bound. Much as I’d like to believe that increasing the cost of holding the medium of account would reduce its value, and hence increase the price level, if the markets suggested otherwise then I’d have to abandon my hypothesis.

If, in addition, markets suggested that inflation expectations were not affected by the announcement by a central bank that it would buy up whatever assets it took to hit an inflation target, then market monetarism would be further weakened. Add to that a couple cases where the announcement of higher government spending did boost inflation expectations, and I’d be inclined to convert from market monetarism to market Keynesianism. Or more accurately, I’d be inclined to invent market Keynesianism, as no one has done so thus far.

[Just think about that for a moment. No one has invented market Keynesianism. What does that tell us? I’m not sure, but it’s a question worth pondering.]

I hope it’s apparent where I’m going with all this. The set of all “market isms”, including market monetarism, market Keynesianism, market RBCism, market Austrianism, market MMTism, and, well why not, market Marxism, are embedded in a sort of meta-theory called “market macroeconomics.”

Market macroeconomics suggests that the markets provide the best way of evaluating the impact of policy shocks on the key macro variables. If the yen plunges on news of negative IOR, then we can infer that negative IOR is expansionary. But that raises an even deeper question–what makes us think this market macro meta-theory is valid? Doesn’t the meta-theory also need to be constantly re-evaluated?

Yes it does, but at the moment it would take for more evidence to push me away from market macro than to push me away from market monetarism. I believe it’s quite possible that at least some kinds of fiscal stimulus surprises would boost NGDP futures prices, if we had such a market, and indeed would boost equity prices even without such a market. In other words, I think it’s quite possible that I’m wrong about monetary offset.

In contrast, I think it’s exceeding unlikely that if a surprise cut in IOR causes the yen to depreciate by 2% in 5 minutes, that the optimal forecast of the one month forward yen doesn’t also fall by something on the order of 2%. If, for instance, the optimal forecast of the one month forward yen did not change at all, then there’d be lots of 10,000 yen notes on the sidewalk, just waiting for traders to pick them up.

What I’m trying to say here is that for the meta-theory of market macro to be wrong, it’s not enough that the EMH doesn’t hold “true”. Heck, it’s obviously not literally true. Rather, the EMH would have to be spectacularly wrong, making it easy to earn large excess returns. That’s possible, but seems exceedingly unlikely to me.

So if someone far smarter than me, such as Scott Aaronson, asks me whether I should constantly re-evaluate the correctness of market monetarism, I’d say yes. If he then asked me about confirmation bias, I’d have to agree that’s a problem. We can’t be trusted to evaluate the accuracy of the theories that we hold dear, because we’ve fought for them for so long. So who can we rely on to evaluate market monetarism? Not people with other ideologies like Keynesianism, because they also have an axe to grind. Not people with no alternative theory, because they are not well enough informed. There’s only one group I trust—market participants who have money on the line. To paraphrase Richard Rorty might have said truth is what the market let’s you get away with.

Can the market also answer “scientific” questions? Sure it can. Aaron Jackson and I once wrote a paper advocating futures markets for both global average temperatures and CO2 levels. (Robin Hanson has pushed this line of thinking in all sorts of directions.)

Of course you need to eventually have some clear observational data to answer the question, or a market won’t work. AFAIK, there are some deep scientific questions (string theory, “many worlds”, etc.) where we currently lack the ability to do an empirical test. But for global warming we can. Time will tell. Even better, markets can provide an optimal forecast of what exactly “time will tell”.

Many readers of this blog favor market solutions to many of society’s problems. So do I. But I’m even more confident about the utility of markets in science, in deciding what’s true and what’s false, than I am that markets are a good way to organize health care or education. Indeed, as a follower of Rorty, I’d be inclined to define “truth” as “what the market believes”. And yes, that obviously means that truth is always provisional. How provisional? It depends on the hypothesis, and to find out you need options markets.

To a computer scientist like Scott Aaronson, the universe looks sort of like a giant computer (or at least I got that impression from the interview linked to above). To an economist like me, science looks kind of like a giant set of markets.

HT: David Levey