I've been reading Lars Christensen's excellent blog posts on Friedman's exchange rate advice. Friedman claimed that fixed exchange rates are bad, partly since the market can never be sure that the country would not later devalue their currency. This makes the currency a target for speculative attacks that might cause the exchange rate to be abandoned. A monetary union, on the other hand, doesn't have this problem.
While reading those blog posts I started to wonder if NGDP level targeting could be vulnerable to a similar problem. Let's say the Fed announces that it will do whatever it takes to keep NGDP on the historical growth path. The market believes the Fed is credible and base velocity increases sharply over the coming months so that NGDP returns to that path.The Fed keeps on fiddling with their funds rate, but no matter what they do, NGDP keeps following the path. The Fed, not wanting to admit that they aren't important anymore, keeps on adjusting the funds rate and keeps on patting their backs for being so amazingly good. What really happened, though, is that the market expected NGDP to stay on the path, so the economy kept on growing at the expected rate.
However, after many years without any slowdowns, the economy took a nose dive (initiated by a war in some oil producing country). NGDP fell by one percent. Well, this shouldn't be a problem. The Fed will get NGDP back on the path. No need to worry. But there was something the market worried about. What if this recession got big enough that the Fed decided that NGDP level targeting didn't work? And since the market fears this, NGDP might fall even more, making the shift away from NGDP targeting even more likely. So NGDP kept on falling and eventually the Fed announced that NGDP level targeting didn't work and switched back to inflation targeting.
I don't think NGDP futures convertibility suffers from this problem. That's a bit like a monetary union. But do we need convertibility to make NGDP level targeting work? The problem, as I see it, is that the central bank might let the monetary base grow too slowly since the market expectations make NGDP stay on the path (the circularity problem). Eventually, the rubber band is stretched too much and snaps back.
So what's the flaw? I hope some competent market monetarist can tear this down.
Also, is this a problem even with an inflation target? Can market expectations keep inflation close to the target even if central bank keeps the monetary base growing too slowly? Did the rubber band cause the current recession?
Update: One big difference between the exchange rate defense and the defense of the NGDP level target is that the exchange rate case leads to tighter policy (Sweden raised its rates to 500% in 1992), while the NGDP case leads to easier policy. This should make the NGDP level target a lot easier to defend. Just increase the monetary base until the market is satisfied. Still, the circularity problem makes this situation so weird.