Description

This study introduces a novel information-adaptive multinomial inverse regression (IA-MNIR) model to assess the clarity of central bank communications, specifically FOMC statements. By quantifying the alignment between text-implied policy stances and actual interest rate targets, the paper develops a clarity measure that captures signal-to-noise ratio in policy communications. Empirical findings suggest that higher clarity leads to a significant decrease in market uncertainty (as measured by VIX), with a 22% reduction per standard deviation increase in clarity. The model isolates influential phrases, reveals temporal clarity patterns, and validates robustness across multiple confounding factors (similarity, statement length, MPU, MPS).