마르코프 국면전환모형을 이용한 국제곡물 가격 예측

Abstract

The purpose of this study is to analyze the volatility structure and fluctuation factors of international grain prices and forecast future price. The price dynamics are divided into normal and crisis periods according to the price fluctuations of the grain market and analyzed by applying the Markov Regime Switch Model, which can be estimated and predicted. The analyzed data comprises international grain prices and the corresponding determinants that explains them. The analysis period spans from January 2006 to June 2021. The empirical findings are summarized as follows: First of all, as a result of estimating the phase shift model considering the Markov process, it is found that it is more appropriate to estimate price volatility in the grain market by dividing the model into two phases - normal and crisis - than OLS which assumes a single phase. This implies that the presence of various phases holds significance in accordance with the price fluctuations observed in the grain market. Furthermore, it is cautiously expected that the phase-change model may prove to be an effective tool for predicting international grain prices in the future, showing superior predictability compared to other models in the evaluation of predictive performance using RMSE.