No. 1075 - Let's talk about it: an NLP approach to evaluate monetary policy communication

Questioni di Economia e Finanza (Occasional papers)
by Giulia Iadisernia and Giulio Carlo Venturi
November 2026
Vai alla versione italiana Site Search

This paper develops an automated natural language processing (NLP) methodology to produce a quantitative measure of the tone of monetary policy communication (textual tone score index), i.e. of the degree to which it is hawkish or dovish. The analysis focuses on transcripts of monetary policy statements and press conferences following ECB monetary policy decisions from 2000 to the present.

The tone of the ECB's communication was generally hawkish prior to the global financial crisis and during the high-inflation period of 2021-23, whereas it was dovish during the global financial crisis, the euro-area sovereign debt crisis, and the Covid-19 pandemic. The analysis confirms that the tone index evolves consistently with the ECB's monetary policy stance and that the ECB's aversion to positive or negative deviations from the inflation target, measured through the analysis of sentences concerning inflation, is symmetric.

Full text

  1. 12 November 2026
    No. 1075 - Let's talk about it: an NLP approach to evaluate monetary policy communication
    PDF 3 MB

Sezione di approfondimento