The seminar will be held at 10:00 a.m. in the Visitor Centre at the MNB.
Abstract
I perform a non-linear discrete regression analysis of the manner in which the FOMC sets the federal funds rate target. Using daily data for broad spectrum of economic indicators, I estimate ordered probit models for discrete changes in the target. They can be applied as a useful benchmark for explaining past and predicting future Fed policy decisions. Along with using more appropriate analysis techniques, my focus is on getting more and better data, and looking more carefully on the timing and institutional details of how the target is determined. Instead of employing quarterly or monthly data averages as is common practice in the literature, the “FOMC meetings” data frequency has been used. I demonstrate the superior performance of my models compared with the performance of simple policy rules (such as Taylor’s ones) as well as existing discrete models from the literature. I show how the time aggregation affects the identification of monetary policy rules.
The paper will be published on 8 th. February.