- The FOMC has an asymmetrical loss function: avoiding a recession is more important than avoiding the risk of overheating
- With this comes the necessity of a symmetrical inflation objective: a temporary overshooting is acceptable
- Given the unclear relationship between unemployment and inflation, the Fed’s tone remains cautious despite upbeat growth projections
The Federal Reserve faces two asymmetries. The first is the asymmetric loss function. This is largely by choice because it wants to avoid triggering a recession (fighting recessions has become increasingly difficult). Chairman Powell insisted in his press conference on Wednesday on the need not to tighten too quickly. Otherwise inflation would rise insufficiently which would make it more difficult to tackle the next recession when it comes. The second asymmetry is imposed on the central bank. At the same time, we note that since 2000, 75% of observations of core inflation were below 2% (see Chart). Engineering an uplift in inflation towards the 2% target looks like a huge task.
Faced with these two asymmetries, insisting that the inflation objective is symmetrical, like Jerome Powell re-iterated on Wednesday, becomes an obvious, not to say inevitable, conclusion: allowing a temporary inflation overshoot is the price to pay for the wish to avoid the most uncomfortable part of the loss function (a recession) that could ensue from what would turn out to be a prematurely restrictive policy stance. This is all the more rational given the difficulty of assessing the level of the natural rate of unemployment and the uncertainty about the relationship between the unemployment rate and inflation. Taking the risk that the Phillips curve re-emerges then makes sense. What also makes sense is avoiding creating the impression that one firmly believes that it will re-emerge. That explains the, on the surface, puzzling changes to the FOMC members’ projections: the median unemployment rate in 2020 is now put at 3,6%, far below the longer run projection of 4,5% whereas PCE inflation in 2020 is projected at 2,1%, a tad above the longer run projection of 2,0%. Nobody seems to believe in the resurrection of the Phillips curve. All this helps explaining the relaxed reaction of bond markets: a temporary inflation overshoot won’t stop the Fed from sticking to a gradualist approach and if the FOMC has doubts about the Phillips curve, why speculate against this? At the end of the day, financial market positioning has become as data-dependent as central bank policy