Source item

Is Kevin Warsh argument that AI justifies a lower equilibrium Fed Funds rate correct? Not necessarily. Here is why… The equilibrium/neutral nominal Fed Funds rate - or r-star - is the sum of the equilibrium/neutral real r-star and the equilibrium/target expected inflation rate. A positive aggregate supply shock - like the AI/tech revolution - has two opposite effects on the nominal r-star : it increases it as real potential GDP growth is higher and there is a positive correlation between real potential growth and real equilibrium rates; and it decreases it as inflation and its expectations are sharply reduced. Thus, the net impact on the nominal neutral r-star is ambiguous. But most likely it is about a wash as real r-star is much higher while expected inflation is lower. Real potential GDP is positively correlated with the real r-star, regardless of the monetary regime as the monetary regime has noting to do with the real r-star that depends on potential GDP growth. Before the GFC, real r-star was positive and even when inflation was at target the neutral nominal r-star was close to 4% - ie 2% inflation and 2% real r-star. After the GFC we had 15 years of Secular Stagnation when the nominal r-star was close to zero or even negative as depressed growth and capex implied a negative real r-star and inflation was near deflation mode. But since Covid the nominal and real r-star have surged. If inflation will be close to 2% target that puts the floor to the nominal r-star. And now the real r-star is higher as secular stagnation is over and potential growth is back closer to 2%. Thus, estimates of nominal r-star are around 3.5% , 2% for inflation and 1.5% for real r-star. If AI/tech secular boom leads to higher potential closer to 3-4% , then the real r- star could surge closer to 3% . But the positive supply shock could push inflation close to 0% over time and inflation expectations could drop to say 0.5%. Then the nominal r-star would still be round 3.5% but its composition would radically change: near 3% for the real r-star and 0.5% for its inflation component. As opposed to today where the same 3.5% nominal r-star is 1.5% real and 2% inflation. So AI doesn’t necessarily imply a lower nominal r-star and it certainly implies a higher real r-star. This is the point that Steve Miran and Kevin Warsh don’t seem to get in arguing that the equilibrium Fed Funds rate is now lower.