우리나라 통화정책에서 통화·신용량의 역할 (The Role of Money and Credit Aggregates in Conduct of Monetary Policy in Korea)

Korean Abstract: 글로벌 금융위기를 계기로 그동안 경시되어 왔던 통화·신용량의 중요성이 학계 및 중앙은행을 중심으로 재인식되고 있다. 그러나 양적지표를 실제 금리중심의 인플레이션 타게팅 운용체계에 어떻게 접목시킬 것인가에 관한 연구는 이제 시작단계이다. 이와 관련하여 본고는 한국은행 금융통화위원회의 월별 의사록 내용을 토대로 금통위원들이 통화·신용량을 언급하며 표명한 금리결정 의사 및 강도를 지수화(indexation)하고 동 지수를 이용하여 우리나라가 인플레이션 타게팅 제도를 도입한 이후 통화/신용량을 금리결정에 어떻게 반영해 왔는지를 분석하였다. 분석결과 금융통화위원회는 통화·신용량의 움직임에도 유의하면서 기준금리를 결정한 것으로 나타났는데, 이는 한국은행이 자산가격과 신용경색 등을 고려하여 인플레이션 타게팅 제도를 신축적으로 운용해 왔음을 시사한다. English Abstract: After having been de-emphasized from around the early 1990s, the importance of financial factors such as the money and credit aggregates in the implementation of monetary policy under inflation targeting regimes has gained renewed attention since the recent global financial crisis that erupted in August 2007. Further discussions and rigorous analysis are however needed, to reach a broadly shared consensus among academia and policymakers on the issue of when and how the central bank should incorporate these quantity variables into its actual conduct of interest rate-focused inflation targeting.Our aim in this paper is to address the issue of how the Monetary Policy Committee (MPC) of the Bank of Korea took into consideration the actual money and credit aggregates in its policy decisions under inflation targeting over the period 2003-2011. A novelty of our empirical analysis is its construction of an index as a measure for capturing, using the monthly Committee meeting minutes for the period concerned, how many MPC members mentioned their intentions with regard to the policy stance by taking the current movements of money and credit aggregates into consideration just prior to their policy decisions. For the analysis we set up a discrete choice framework for the index. We assign integer values of from 4 through -4, depending upon how many MPC members show their intentions that the policy stance be “tight,” “neutral,” or “easy,” in consideration of the movements of money and credit aggregates for a specific month. The integer value 4 denotes that more than five MPC members mentioned favoring a policy rate hike while the value 2 indicates that it was less than two MPC members. The same classification applies to the negative integer values. A zero value indicates that no MPC member has shown any favoring of a “tight” or “easy” policy stance at all, implying an intended “neutral” stance of the MPC members.We conduct OLS estimation of the effects of the index on the actual policy rate decisions by the Committee members, using an equation controlling for two key macroeconomic indicators such as the inflation gap and the output gap that the Committee considers in setting policy. Interestingly, the estimated index coefficient turns out to be positive and statistically significant. The regression suggests that when the MPC members frequently and strongly mentioned the need to take notice of the movements of money and credit aggregates around the time of the Committee meeting, monetary policy was dominated by concerns over the stabilization of asset prices or massive credit squeezes in the financial markets. We, in addition, estimated two different equations, as a way of checking the robustness of the OLS estimates. The first specification is an ordered probit model in which we estimated the effects of the index on the policy changes in discrete increments. The integer values assigned for the policy changes range from -2 through 2. The values of 2(-2) and 1(-1) correspond to 50 bp and 25 bp hikes (cuts) in the policy rate, respectively, while a value of 0 indicates no policy change. It seems more accurate to use this five-stage classification of policy rate changes as the dependent variable, because the evolution of changes in the policy rate does not follow a smooth process but a step function. In a second experiment, we first estimated the standard Taylor rule over the period concerned, and then the effects of the index on its residual estimates. In both cases, the estimated index coefficients are positive and significant, implying that the results support the OLS estimates mentioned above.To summarize, the evidence suggests that the Bank of Korea has been implementing flexible inflation targeting.