At the Philadelphia Fed's conference on Real-Time Data Analysis, Methods, and Applications, I just discussed a very neat paper by Jim Nason and Gregor Smith on "Reverse Kalman Filtering U.S. Inflation with Sticky Professional Forecasts". Basically, Jim and Gregor take observed forecasts from the SPF and try to back-out the implied inflation trends under different assumptions of information processing. These estimates are less parametric than a formal Kalman filter. Strikingly, the results correspond quite well to what my model would yield. Here are my slides: pdf |

News >