회사채 시장에서의 모멘텀 현상

Abstract

본 연구는 국내 회사채 시장에서의 모멘텀 현상에 대해 분석하였다. 첫 번째, 회사채시장에서 모멘텀 현상이 발견되었으며, 이는 기존의 채권과 주식에 대한 체계적인위험 요인들(Fama and French, 1993, Carhart 1997)로 설명되지 않았다. 두 번째, 모멘텀 전략의 수익성은 주로 평가기간과 보유기간이 6개월 이내로 짧은 경우에강하게 나타났다. 세 번째, 회사채 모멘텀 전략의 수익성은 금융위기를 제외한 기간과경기 확장기에서 강하게 나타났다. 네 번째, 회사채 모멘텀 현상은 주로 신용등급이낮은 그룹에서 강하게 나타났다. 다섯 번째, 모멘텀 현상이 크게 나타나는 신용등급이낮은 채권 그룹들은 주로 규모가 작고, 유동성이 낮았다. 이러한 결과로 볼 때, 모멘텀효과가 신용등급이 낮은 쪽에서 발생하는 것은 정보의 지연 반응 효과(Hong and Stein, 1999)에 의한 가능성이 있는 것으로 추측할 수 있다. 여섯 번째, 과거 연구(Gerbhart, Hvidkjaer and Swaminathan, 2005)에서 발견되었던 주식과 채권 모멘텀 간의전이효과(Spillover)는 나타나지 않았다. 마지막으로 채권의 신용등급이 상향 조정되어발생하는 수익률의 증가는 모멘텀에 큰 영향을 끼치지 못했다. The momentum effect is the phenomenon whereby the higher the past return on an asset, the more persistent this return will be in the future. The momentum effect can be found not only in equity, but also in several other asset classes. For example, recent studies show that momentum occurs in the corporate bond market (Pospisil and Zhang, 2010; Jostova et al., 2013; Israel et al. 2017; Houweling and Zundert, 2017, Ho and Wang, 2018). This phenomenon has been documented in the U.S. and various other regions, where various strategies for promoting momentum are appearing. However, empirical studies on the momentum effect in Korea are mostly limited to stocks. AAlthough bonds represent a large proportion of Korea’s financial market, studies of the existence of momentum in this market are still rare. Therefore, we attempt to determine whether the momentum effect exists in the Korean corporate bond market, and if so, what drives this effect. We focus on the corporate bond market, not the government (treasury) bond market, for the following reasons. Depending on the characteristics of the firm that issues a corporate bond, the speed at which investors respond to this information may vary. Given the characteristics of corporate bond issuers, the momentum effect is more likely to occur if there are differences in the speed at which information is reflected. In contrast, in the government bond market, issuers are not diverse. Therefore, the information asymmetry among market participants is smaller for government than corporate bonds, so investors have less influence in the government bond market. Participants in the corporate bond market are more likely to interpret private or public information differently from each other than their counterparts in the government bond market are. Our main results are as follows. First, the higher the past return of a bond, the higher its future return. In other words, we verify the momentum effect. A momentum strategy comprising a six month formation period and a six month holding period shows an average return of 0.17% per month (2.02% per year). The momentum in the corporate bond market is not explained by previously observed systematic risk factors for bonds and stocks (Fama and French, 1993, Carhart 1997). Thus, we cannot conclude that the observed momentum of corporate bonds is associated with compensation for systematic risk. The profitability of the abovementioned corporate bond momentum strategy remains robust even when we control for various characteristics, such as the duration and age of the bond. Second, the profitability of the bond momentum strategy is strong when the formation period and holding period are the short-term periods of three to six months each. The corporate bond momentum is mostly sustained in the short term. Third, the bond momentum strategy is profitable during the period excluding financial crisis and economic expansion, but not during the period including financial crisis and contraction. Fourth, the corporate bond momentum is strong in the low credit rating group. In other words, the higher the past return, the higher the future return in the group of firms with low credit ratings. Fifth, we find that most of the firms in the low credit rating group, which shows a significant momentum effect, are small and have low liquidity. Accordingly, we suggest that the momentum effect occurs at the lower credit level due to the gradual information diffusion phenomenon reported by Hong and Stein (1999). For example, momentum is high under small market capitalization, when private information is difficult to spread. In addition, it is difficult to interpret information from firms with low credit ratings. Sixth, we do not find evidence of the spillover between stock momentum and bond momentum reported in previous studies (Gerbhart, Hvidkjaer, and Swaminathan, 2005). For example, we do not observe a significant relationship between high stock returns in the past and high future bond returns. Finally, after controlling for the effect of bond rating changes on bond returns, the performance of the momentum strategy remains statistically and economically significant. Our study has the following academic and practical implications. First, it reveals a momentum phenomenon in non-stock assets, namely corporate bonds, in the Korean market. Many studies focus on the momentum phenomenon in stock markets. Our results suggest that we also need to research the momentum phenomenon that can occur in various asset classes in Korea. Second, portfolio managers can use our findings to develop an effective bond investment strategy. In the overall asset management industry in Korea, investments in corporate bonds are increasing, especially among institutional investors such as pension funds and insurance companies. Therefore, based on the results of this study, the abovementioned momentum strategy can be used to obtain excess returns in the Korean bond market.