뉴케인지안 모형에서의 통화의 역할과 예측력 검정

This paper investigates the role of money in the business cycle in Korea. In 1998 the Bank of Korea switched the operating target for monetary policy from monetary aggregates to a short-term interest rate. Although monetary aggregates cease to be considered important in the course of official monetary policy decision making process, it is widely speculated that monetary aggregates may still play significant roles in economic fluctuations. That is, there may remain a causal link through which money affects other economic activities. Or, if not a causation, monetary aggregates may at least help forecast inflation and growth, the most important target variables of monetary policy. These are the issues we take up in this paper. In order to explore the causal link between money and economic activities, we set up a dynamic stochastic general equilibrium (DSGE) model with money demand. Following the literature on the topic, we assume a non-separable utility function for consumption and money holding. This is to allow money to play an explicit role in determining other variables in the model. Then we estimate the model for the Korean economy with both the maximum likelihood estimation method and the Bayesian estimation method. The estimation results indicate that money does not affect the marginal utility of consumption, which can be interpreted as the real balance effect is not present. Next, to investigate whether money has additional forecasting power for inflation and growth, we nest money in otherwise standard out-of-sample forecast models and obtain statistics for the encompassing test. We find that money did have considerable forecasting power for inflation until 1997, the period when the Bank of Korea targeted monetary aggregates. However, monetary aggregates have lost the additional forecasting power for inflation and growth since 1998.