Evaluating the Effectiveness of the BOK Communication using Text Mining

Overview

How well does a central bank communicate its intentions to the public? This project tackled that question by applying text mining and NLP techniques to over a decade of Bank of Korea (BOK) monetary policy communications. By quantifying the “tone” of policy statements, we measured how effectively the BOK’s words aligned with its actions and how markets responded.

Research Approach

We analyzed 152 Monetary Policy Board meeting minutes spanning from 2005 to 2017, building a comprehensive dataset of central bank language patterns. Using the eKoNLPy library (developed as a companion tool for this research), we classified policy statements along a hawkish-dovish spectrum and tracked how communication tone evolved over time.

The analysis examined whether shifts in language preceded actual policy changes, and how financial markets reacted to different communication styles.

Key Findings

The research demonstrated that text mining can effectively capture the nuances of central bank communication in Korean, and that measurable tone shifts in BOK statements provided meaningful signals about future policy directions.

Publications

This research was published in the Korean Economic Review (Vol. 35, No. 2, pp. 471–511, 2019) and as BOK Working Paper No. 2019-1 by the Bank of Korea.

Collaborators

• Bank of Korea, Economic Research Institute: Lead research partner
• Research Team: Park Ki Young, Lee Young Joon, Kim Soohyon