My readers will know that I think that the Federal Reserve has taken a step in the right direction with its latest policy action. I do think that the fed finally after four years of failure is moving towards a more rule based monetary policy. However, it is certainly far from perfect and there is still a lot of risks involved.
The Minutes from the latest FOMC meeting was published yesterday and it is clear that the FOMC is well-aware that it needs to address it’s communication problem. That’s positive. However, it is also clear that the fed still don’t have a proper communication policy in place and even though we are moving towards a more rule based monetary policy it still not completely clear what the rule is and it is not entire clear how it should be implemented. We are still far away from Milton Friedman’s ideal of having a computer control monetary policy. However, I think that the fed should move in the direction of that ideal and it could start the journey toward this goal by introducing what we could call “policy futures”.
It is obvious that the fed is aware that there is problems with the present forecasting set-up within the fed. The key problem is one that every central bank in the world is facing – should the central bank forecast that it will fail? That is effectively what the fed has been doing so far when it is saying that it expect a fragile and weak recovery.
Scott Sumner has suggested that monetary policy should be “pegged” to a NGDP future, which would mean that the money base is increased or decreased continuously as market expectations for future level NGDP changes. This is basically the Friedman ideal of a computer – or rather the market – controlling monetary policy. However, a less radical plan where futures are “just” used for policy guidance and forecasting is also possible and that is what I suggest that the fed should look at.
There are some very clear advantages of using the market to forecast. First of all the fed would not have to know the “real model” of the US economy. Second the forecasts would be unbiased. Third the fed would have real-time forecasts of its policy variables.
It is pretty clear that the fed is now moving towards some kind of Evans rule where changes in the money base is a function of unemployment and inflation. We don’t know fed’s reaction function, but a version of the Evans rules could take the following form:
(1) ∆b = α(U-U*)-β(π-πT)
Where ∆b is the change in the money base, α and β are coefficients, U is unemployment and U* is the fed’s unemployment target or the structural unemployment, π is inflation and πT is the inflation target.
There plenty of reasons to be skeptical about the fact that the fed is so clearly targeting real variables (employment/unemployment). However, by using policy futures it might be possible to greatly reduce these risks.
I imagine that the fed set up a futures or an options market on for example inflation, employment/unemployment and obviously NGDP on different time horizons.
Let’s say that the fed has the target of reducing unemployment to 6%, but also want to maintain long term price stability (keeping inflation around 2%). If structural unemployment is higher than 6% then that would obviously not be possible – and if the fed tried to push unemployment below 6% then inflation would explode. A policy future would greatly help assess this risk.
Hence, the fed could issue a put option that would be knocked in if unemployment dropped by 6% and inflation was below 2 or 3% at some future date – for example January 1 2013. Such an option would give an assessment about whether it is likely that the fed will hit it’s policy objectives. If the market assess that structural unemployment is above 6% then that would be reflected in the pricing of the put option.
If the fed issued a number of different policy futures and options on the key policy objectives it could get the markets’ assessment of whether it is on the right track in terms of fulfilling it’s monetary policy objectives or not by cross-checking the pricing of different policy futures.
Such policy futures could also greatly help the fed in it’s communication with the markets and it would probably also be much easier to get consensus on the FOMC about the possible risk to monetary policy.
The fed would very easily be able to set up such policy futures markets, but the informational gains would in my view be tremendous. The only “problem” would that the fed would need fewer economists to do forecasting…
Yet another argument for prediction markets: “Reputation and Forecast Revisions: Evidence from the FOMC”
Benn & Ben – would prediction markets be of interest to you?
Prediction markets and government budget forecasts
Central banks should set up prediction markets
Markets are telling us where NGDP growth is heading
Scott’s prediction market
Robin Hanson’s brilliant idea for central bank decision-making