기간 및 신용 스프레드의 경기 변동 예측력에 관한 연구 (A Study on the Predictive Power of Term and Credit Spreads for Economic Cycles)

서지정보

· 발행기관 : 한국무역연구원
· 수록지 정보 : 무역연구 / 20권 / 5호 / 363 ~ 380페이지
· 저자명 : 전병조

초록

Purpose – This study empirically analyzes whether term spread and credit spread provide useful information to predict domestic business cycles in Korea, and whether there is a difference in predictability between the two.
Design/Methodology/Approach – To overcome the limitations of linear analysis, this study employs transfer entropy and the random forest algorithm, which are known to be effective for nonlinear causality analysis. Monthly data from March 2003 to May 2024 are used. The term spread is calculated as the difference between the 10-year treasury bond yield and the call rate. The credit spread is calculated as the difference between the AA-rated corporate bond yield and the 10-year treasury bond yield, and the difference between the BBB-rated corporate bond yield and the 10-year treasury bond yield. The cyclical component of the coincident composite index represents business cycle fluctuations.
Findings – Our findings reveal that both term spread and credit spread have significant predictive power for business cycles, with credit spread, particularly for BBB-rated corporate bonds, exhibiting stronger predictive capabilities. Moreover, the predictive power of spreads varies across different economic conditions, increasing during periods of heightened economic volatility. Our analysis also highlights the persistent influence of spreads on business cycles, extending over a considerable period before and after cyclical turning points.
Research Implications – These findings imply that financial market participants should monitor both term spread and credit spread to predict business cycle fluctuations. Furthermore, analyzing the predictive power of each spread can help predict changes in trade and the terms of trade.

영어초록

Purpose – This study empirically analyzes whether term spread and credit spread provide useful information to predict domestic business cycles in Korea, and whether there is a difference in predictability between the two.
Design/Methodology/Approach – To overcome the limitations of linear analysis, this study employs transfer entropy and the random forest algorithm, which are known to be effective for nonlinear causality analysis. Monthly data from March 2003 to May 2024 are used. The term spread is calculated as the difference between the 10-year treasury bond yield and the call rate. The credit spread is calculated as the difference between the AA-rated corporate bond yield and the 10-year treasury bond yield, and the difference between the BBB-rated corporate bond yield and the 10-year treasury bond yield. The cyclical component of the coincident composite index represents business cycle fluctuations.
Findings – Our findings reveal that both term spread and credit spread have significant predictive power for business cycles, with credit spread, particularly for BBB-rated corporate bonds, exhibiting stronger predictive capabilities. Moreover, the predictive power of spreads varies across different economic conditions, increasing during periods of heightened economic volatility. Our analysis also highlights the persistent influence of spreads on business cycles, extending over a considerable period before and after cyclical turning points.
Research Implications – These findings imply that financial market participants should monitor both term spread and credit spread to predict business cycle fluctuations. Furthermore, analyzing the predictive power of each spread can help predict changes in trade and the terms of trade.

참고자료

· 없음