Decoding the directives: a textual analysis of monetary policy

Abstract

Purpose This study aims to explore the linguistic and syntactic textual features of Central Bank of Nepal (Nepal Rastra Bank [NRB])’s monetary policy. Considering the recent literature and methodological advancement in computational linguistic analysis, this study intends to explore the features of published monetary policy reports. Design/methodology/approach Text mining technique has been used on the monetary policy published by Central Bank of Nepal for the period between 2002/03 and 2021/22 to describe the textual features such as length, tone, degree of forward looking, use of numerical contents and readability. The raw text was tokenized using Python’s Natural Language Toolkit. Considering the LM dictionary, the frequency of tokens matching with dictionary is computed and divided by total number of words to normalize the obtained value. Findings This study found that NRB publishes lengthy monetary policies during economic contractions and vice versa. The tone of the policies are pessimistic most of the time. NRB’s policies are not sufficiently forward looking and complex to be comprehended by layman. Ergo, NRB shall form a team of communication experts to ensure publication of optimistic policies with appropriate linguistic features. Furthermore, publishing the minutes of monetary policy meetings will help enhance effective communication through transparency and proper functioning of expectations channel. Originality/value To the best of the authors’ knowledge, no similar study has been conducted to assess the textual features of monetary policy in Nepal. And this study will be helpful to gauge the status of central bank communication in the context of emerging and least developed countries.