Overview

                Let R be the interest rate paid on excess reserves. Let G be the real economic growth rate. Let V be inflation (or money velocity). For the purposes of this thought experiment, I am going to assume that the Fed can accumulate “profit” without remitting it back to the treasury and its bond portfolio generates a return of at least G. Assume banks are flush with reserves.

                The Fed announces that it is going to set R=G and will be following a policy of price level targeting. What do you think happens? I think there would be minimal effect on real economic growth, and some deflation/disinflation, particularly if the Fed is credible. I think real growth would slow only if there was a fiscal response. By slowly decreasing R and purchasing treasuries with the cash we save, we will eventually stimulate the economy and put inflationary pressure.

                The reason I like this policy is twofold. First, it keeps short-term rates above zero, by quite a margin during booms. I believe this will stop speculative bubbles from emerging in the first place. Second, by forcing the government to borrow more, we not only “keep them honest", but actually force them to create more safe assets, giving us more capacity to ease when we need to. This is sort of counter-intuitive, but I believe this works almost as a better form of NGDP targeting. After all, if the price level target is believable, the opportunity cost of letting go of reserves is still a function of NGDP. It’s just that we don’t sacrifice long-run control of the price level, or have to make forecasts of potential GDP growth.

                To be clear, the interest paid on reserves would be equal to NGDP minus inflation (that’s what real GDP is). If you think that inflation and NGDP are both procyclical variables, this policy is strictly more countercyclical than NGDP targeting.