LOG IN창 닫기

  • 회원님의 아이디와 패스워드를 입력해 주세요.
  • 회원이 아니시면 아래 [회원가입]을 눌러 회원가입을 해주시기 바랍니다.

아이디 저장

   

아이디 중복검사창 닫기

HONGGIDONG
사용 가능한 회원 아이디 입니다.

E-mail 중복확인창 닫기

honggildong@naver.com
사용 가능한 E-mail 주소 입니다.

우편번호 검색창 닫기

검색

SEARCH창 닫기

비밀번호 찾기

아이디

성명

E-mail

학술자료 검색

가격제한폭이 꼬리위험에 미치는 영향

  • 오세경 건국대학교 경영대학 교수
  • 기혁도 건국대학교 경영학과 박사과정, 마일스톤자산운용
본 연구는 위험지표로서의 꼬리위험과 가격통제장치로서의 가격제한폭의 관계를 조명하고 기업규모와 거래소 유형이 꼬리위험의 미래 주식 수익률 예측력에 어떤 영향을 미치는지 우리나라 주식시장을 대상으로 실증 분석하였다. 주요 분석 결과는 다음과 같다. 첫째, 꼬리위험은 가격제한폭이 12% 이상으로 확대된 이후부터 통계적으로 유의한 주식 수익률 예측력을 가지는 것으로 나타났다. 둘째, 규모가 작은 기업의 경우 꼬리위험의 민감도가 장기간에 걸쳐 지속되고 시간이 흐를수록 영향력이 커지는 반면, 기업 규모가 큰 경우에는 꼬리위험이 단기간에 유의한 설명변수가 되지만 장기적으로는 영향력이 약해짐을 발견하였다. 셋째, 규모가 큰 기업의 경우 규모가 작은 기업에 비해 가격제한폭의 영향을 덜 받는 것으로 나타났다. 넷째, 유가증권시장에서는 꼬리위험의 주식 수익률 예측력이 단기의 예측기간에서 통계적 유의성이 있는 것으로 나타난 반면, 코스닥시장에서는 주가 수익률 예측력이 장기에 걸쳐 나타나고 꼬리위험의 주가 민감도도 크다는 것을 발견하였다. 다섯째, 꼬리위험 기대값에 대한 민감도 분석 결과가 상하한가에 의한 가격 왜곡현상을 최소화하면서 주가의 급변동을 억제하고자 하는 정책 목적을 달성하고자 하는데 보조적인 활용이 가능할 것으로 기대된다.
꼬리위험; 가격제한폭; 수익률 예측력; 왜도(skewness); 규모효과; Tail Risk; Price Limit; Predictive Power; Skewness; Size Effect

Impact of Price Limits on Tail Risk

  • Sekyung Oh
  • Hyukdo Kee
This study analyzes the relationship between tail risk as a risk indicator and price limits as a price control system. We investigate the impact of price limits on tail risk and examine the predictive power of tail risk on future stock returns. We use the measure of tail risk proposed by Kelly and Jiang (2014), which is based on the commonality inherent in the tail risk of individual stocks and is useful for capturing the common factors associated with tail risk at each time point. We analyze the sensitivity of Oh, Park, and Kee’s (2017) expected value model of tail risk under price limits and simulate the random returns under various price limit conditions. The sensitivity analysis for the expected value model of tail risk shows that the tail risk is underestimated when the price limits are tight. The sensitivity results can also be supplemented to find the appropriate price limits needed to achieve the policy objectives of mitigating the extreme fluctuation of stock prices and minimizing the price distortions caused by the upper and lower limits. We then examine these findings using a simulation of 50,000 random returns with leptokurtic distribution. The tail risk measured suggests that the tail risk is underestimated as the price limit becomes increasingly tight. However, with a well-expanded price limit, the tail risk based on leptokurtic distribution is estimated to be about 20% to 30% higher than that based on the standard normal distribution and the negative impact of the price limit on tail risk is generally attenuated after a price limit greater than 12% is applied. The price limit system, which limits the range of prices within which stocks are allowed to be traded, restricts the intrinsic fluctuation demand of the stock market and is believed to have a significant impact on tail risk. Price limit systems have been implemented in many Asian countries such as Japan and China, which are the third and fourth in the world in terms of stock market capitalization, and Korea, Thailand, Malaysia, and Taiwan. Our empirical analysis focuses on the Korean stock market because the price limit has been gradually eased from an average of 4.6% to the current 30%. Our sample covers trading on the Korean stock market from January 1990 to October 2015. We find that the tail risk during the period with the price limit of 12% or more increases more than twice compared to that of previous period. We interpret this as reflecting the inherent nature of the tail risk associated with stock movements without being subject to the artificial control of the price limit. We separately analyze the predictive power of tail risk in the KOSPI and KOSDAQ markets and examine the effect of firm size on the predictive power of tail risk for future Korean stock returns. Our main empirical results are as follows. First, among groups of the same firm size, the coefficient of tail risk increases as the forecasting period becomes longer, which means that the sensitivity to tail risk persists over a long period. Second, when examining the size effect in relation to the forecasting period, we find that in the short term, the sensitivity of tail risk increases as the firm size increases whereas for longer forecasting periods, the sensitivity diminishes as the firm size increases. Third, in the largest firm size group, we find statistically significant predictive power in the short-term forecasting period for the whole sample period. We find that the stock returns of large firms are less affected by price limits than those of small firms. Fourth, in all of the forecasting periods for the KOSDAQ market and in the short-term period of the KOSPI market, the tail risk has statistically significant predictive power on future stock returns. In the KOSDAQ market, the longer the forecasting period, the more significant the predictive power and the more sensitive the tail risk. In contrast, in the KOSPI market, the tail risk is only found to have predictive power in a less than one year forecasting period and the sensitivity of the tail risk is lower than that of the KOSDAQ market. Fifth, by estimating the predictive power of tail risk on the future skewness, we find that the coefficients of tail risk have negative values for all forecasting periods and Newey West t-statistics are sufficiently high. This confirms that the two variables share the same dynamics and the tail risk is closely related to the left-skewed distribution.