Research project: FOMC uncertainty and monetary shock identification

This paper introduces a novel set of indices using Natural Language Processing (NLP) techniques applied to Transcripts and Staff Reports of the Federal Open Market Committee (FOMC) meetings. These indices are designed to quantify uncertainty in monetary policy decisions and are applied in two key ways: first, to augment the standard Structural Vector Autoregression (SVAR) framework, and second, within an Instrumental Variables SVAR (IV-SVAR) framework to identify interest rate shocks. Evidence from the SVAR analysis indicates that these policy uncertainty measures are significant drivers of economic activity while exerting minimal impact on financial markets or prices. However, IV-SVAR results reveal that most of these indices are weak instruments, failing to fully resolve the price puzzle.