earticle

영어

Since the seminal work of Irving Fisher, the question of how stocks covary with inflation has been extensively examined because inflation risk erodes purchasing power and threatens invetor’s long-term objectives (Ang et al., 2012). Most studies found that nominal stock market returns and inflation are negatively correlated. However, there are some evidence that some non-cyclical industries tend to covary positively with inflation, even though the inflation beta is not significant. In addition, the previous literature has focused on the hedging ability of aggregate stock market indices. Recently, Ang et al. (2012) found that some stocks have the ability to be good inflation hedges, suggesting that an investor seeking to hedge inflation risk would optimally hold this firm-level constructed portfolio rather than a market-weighted index. Therefore, if equity funds construct their portfolios based on individual stocks whose returns covary strongly with inflation, those funds have the potential to provide a better inflation hedge for investors, especially who want to avoid inflation risk. Based on this conjecture, this paper examines the inflation hedging ability of Korean equity funds during the January 2001 to May 2013 period. To measure the inflation hedging ability of individual equity funds, we compute fund-level inflation betas following Bekaert and Wang (2010), by regressing individual fund returns on inflation. This allows us to conduct an ex-post analysis of which funds provided the strongest realized covariation between fund returns and inflation. As Jiang, Yao, and Yu (2007) noted, evidence of timing ability can result simply from “luck.” For instance, even if no funds have timing ability, when there are a large number of them, some will have significant timing measures based on t-statistics, due to random chance (Jiang et al., 2007)and the sampling variation. To consider this problem and identify whether the inflation hedging ability is genuine, we use the cross-sectional luck distribution, estimated using the bootstrap approach, to distinguish between skill and luck (due to sampling variation), as proposed by Kosowski et al. (2006). The advantage of the cross- sectional bootstrap approach is that it allows researchers to obtain a distribution of the inflation betas for all funds; specifically, it does not consider the luck distribution of a particular fund but, rather, considers that of all funds, which allows us to draw a statistical inference of funds in the extreme tails of the cross-sectional distribution (i.e., extreme positive inflation betas). When using the realized and expected inflation, we find that while some Korean equity funds have inflation hedging ability, it is due more to fund managers’ luck (or sampling variation) than skill. Some previous studies have shown that inflation betas are time-varying. To consider this finding, we divide our sample period into two sub-periods ranging from January 2001 to December 2005 and from January 2006 to May 2013, respectively. This sub-period analysis shows that slightly more funds have returns that covary significantly and positively, in the second sub-period than in the first period. However, the significantly positive inflation betas in both periods are due to luck, consistent with those of the whole- period analysis. The bootstrap procedure used by Kosowski et al. (2006) assumed that the residuals from the regression analysis were independently and identically distributed for funds. However, it is possible that the residuals have serial dependence over time or cross-sectional correlations across funds. To evaluate their effect on the bootstrap results for inflation hedging skill, we adopt the sieve and stationary bootstrap approaches. The empirical results for these two approaches do not differ from those of the previous results. Lee and Jeon (2012) found that most Korean equity funds tend to invest in large and growth stocks relative to the KOSPI 200 index, implying that their inflation hedging ability may influence our results. To examine this effect, we divide our sample funds into their investment styles: large/small and value/growth using the Fama-French three-factor model. The results for their investment styles are consistent with those of whole sample funds. Finally, examining the persistence of Korean equity funds’ inflation hedging ability reveals no persistence, indicating that it is difficult for fund investors to construct portfolios of equity funds that are good inflation hedges. Overall, our results indicate that the good inflation hedging abilities of some Korean equity funds are merely the result of luck (or due to sampling variation) rather than fund managers’ skills.

한국어

본 연구에서는 2001년 1월부터 2013년 5월까지 국내 주식형펀드의 인플레이션 헤지 능력에 대해 살펴본다. 선행연구에서 볼 수 있듯이, 인플레이션 헤지능력은 여러 가지 방법으로 측정될 수 있지만, 본 연구에서는 Bekaert and Wang(2010)과 Ang, Biere, and Signori(2012)에서와 같이, 주식형펀드 수익률이 인플레이션과 양(+)의 관계를 가지는 경우 그 주식형펀드는 인플레이션 헤지능력이 있다고 정의한다. 또한, 주식형펀드에서의 인플레이션 헤지능력이 펀드매니저의 능력(skill)에 기인한 것인지, 표본추출오차(sampling variation)에 의한 운(luck)에 기인한 것인지를 구분하기 위해 부트스트랩(bootstrap)을 이용한 횡단면 운분포(cross-sectional luck distribution)을 이용한다. 분석결과, 실제 인플레이션과 기대인플레이션을 이용한 분석에서는 소수의 주식형펀드가 인플레이션과 양(+)의 관계를 가지는 것으로 나타났으나, 횡단면 운분포를 이용하여 검토한 결과, 이러한 결과는 펀드매니저의 능력이라기보다는 운에 의한 것으로 조사 되었다. 또한, 인플레이션 베타가 시간가변적이기 때문에, 표본기간을 두 기간으로 구분하여 분석하였다. 그 결과 두 하위 표본기간 모두에서 운에 의해서 일부 주식형펀드가 인플레이션과 유의한 양(+)의 관계를 가지는 것으로 나타났다. 강건성을 검증하기 위해 시이브 부트스트랩(sieve bootstrap)과 안정적 부트스트랩(stationary bootstrap)을 이용하였으며, 이 분석에서도 운에 의해, 즉 표본 추출오차에 의해, 소수의 주식형펀드 에서 인플레이션과 유의한 양(+)의 관계가 나타나는 것을 나타났다. 또한, 펀드의 운용 스타일을 ‘대형/소형’ 그리고 ‘가치/성장’으로 구분하여 추정한 결과 역시 전체 표본을 이용한 분석결과와 유사하게 나타났다. 마지막으로 인플레이션 헤지능력의 지속성을 검토해 보았을 때, 국내 주식형펀드에서는 지속성이 관찰되지 않았다. 이는 투자자들이 인플레이션 헤지능력을 가진 펀드로 포트폴리오를 구성하기는 쉽지 않다는 것을 의미한다.