Why Expectations Based Econ Models Don’t Work

Jason Smith wrote a tremendously good post on the pitfalls of relying on expectations in economic models. We see this all the time in mainstream economics where the economist relies a good deal on market expectations to infer certain conclusions. For instance, if the Fed implements a new round of QE we often see economists cite changes in “inflation expectations” as represented by the change in break-even inflation rates or something like that. They will often say that a certain policy “works” or doesn’t because of the expectations channel.  If the Fed can boost expectations then they can change market behavior and future output. But this is often bunk because it doesn’t reflect how expectations actually work in reality.

First, as Jason notes, the markets have to believe that a policy action will result in fundamental economic change in order for it to be effective. Market Monetarists famously say that their policies can work through the expectations channel because of the “Chuck Norris effect”. That is, Chuck Norris doesn’t punch your beard off, he just threatens to punch your beard off and you tuck tail and run before the beard gets punched off. In other words, the policy works because of the threat, not the action. I’ve called this theory nonsense because it’s based on the assumption that the Fed is actually Chuck Norris. But as any good consumer of Way of the Dragon knows, Bruce Lee said:

“willing is not enough, you must do.”

I’ve called this the Bruce Lee effect. And we all know that Bruce Lee kicked Chuck’s ass right after he ripped out his chest hair. The point is, expectations only work if the markets really think the policy could have the effect that it threatens. You have to be Chuck Norris to threaten the markets with something (or better yet, Bruce Lee considering he’s the far superior martial artist).  But, for instance, as we all know now, QE doesn’t really create inflation so the threat of more QE just doesn’t seem to have the impact that many might expect it to have. At an operational level, I’ve described QE as lacking the Bruce Lee effect. That is, QE doesn’t kick your ass and rip your chest hair out.  It just results in asset swaps that have weak side effects on some markets (like some asset prices).  So there’s no reason for the threat of QE to result in higher inflation because the markets now know that the Fed isn’t Chuck Norris when it comes to creating inflation.

But there’s a more important understanding about expectations here. Expectations in markets often change because market participants expect to be wrong. That is, asset managers are generally risk managers. They don’t always position themselves to be right. They often position themselves to be wrong. For instance, anyone who studies Fed Funds Futures will know that the futures curve just about always slopes up. This has been consistently wrong for 5 years running. That’s not because fixed income traders are stupid. It’s because fixed income traders are hedging against the risk of being on the wrong side of the trade. In other words, the futures curve slopes up to hedge against being wrong. And anyone who was devising policy based on this sort of an expectations based market model has basically been wrong as well because they don’t understand how traders and markets actually work. Scott Sumner loves to say that we should keep finance out of macroeconomics, but his lack of understanding of finance is actually the cause of so many of his misunderstandings about macroeconomics.

All of this can be filed under the folder of “Economics is often a Political Theory Masquerading as a Reflection of Reality”.  And it’s a big problem because an economist who can’t accurately understand how markets operate in reality will totally misunderstand the application of many of their ideas.  Theory is nice and all, but only when it meshes well with reality.