Are the Minutes of the Bank of Korea Informative on the Policy Rate Decisions?

Abstract

We estimate the policy stance on the monetary policy rate decision, which stands for the intensity of the committee’s consensus to adjust the policy rate, with the minutes of the Bank of Korea from 1999 to May 2018. We try to estimate the policy stance more directly without relying on external dictionary information by using a model that combines two observations—the text of minutes and the actual policy rate decisions, both of which are closely related to the policy stance. We also propose an algorithm generating additional N-grams to improve the quality of Korean text data. We observe that the estimated policy stance improves the identification of the policy rate shock in a vector-autoregression model, without losing many degrees of freedom by adding only one variable. The result could be one example showing that the text data work as alternatives to the traditional economic data, possibly even more parsimoniously and efficiently.