Forecasting the Jumps of Stock Index Using Volatility Skew

Kim, Sol; Park, Hye Hyun

Korean Journal of Financial Studies 2012;41(2):189-231.
Published online April 30, 2012.

Sol Kim, Hye Hyun Park

변동성 스큐를 통한 주가지수 점프예측력 검증

김솔, 박혜현

Abstract

This paper investigates whether the volatility skew has predictive information about the jumps of stock index. When estimating volatility skew, we use three different methodologies, Doran, Peterson and Tarrant(2007), Corrado and Su(1996), Bakshi, Kapadia and Madan(2003) models to give more robustness to this paper. We categorized jumps on the base of the percentage change in stock index over a previous day. Additionally, We employ Lee and Mykland(2006) methodology which isolate the volatility effects on jumps. Followings are the major findings and implications drawn from the empirical analysis of the Korean options market. First of all, Bakshi, Kapadia and Madan(2003) skew and Corrado and Su (1996) skew has power in predicting the market crash while does not have for the market spikes. As for Doran, Peterson and Tarrant(2007) skew, put volatility skew has some information about downward market crash while call volatility skew has for upward market spikes. However, the predictive power of put volatility skew is stronger than that of call volatility skew. Second, the predictive power of volatility skew for forecasting the jumps of underlying assets weakens as time to maturity of options increases while implied volatility has still predictive power in the longer maturity options. Third, as a result of conducting probit model using LM jump, we find that the volatility skew still has predictive information for forecasting the movements of underlying assets under controlling volatility impact on jumps. In contrast with Doran, Peterson and Tarrant(2007), we observe that there is some information about future market movements in the volume variable and the predictive power of the volume variable is getting stronger as it is expected to be rather larger market movements. Also, put volume has more information contents in forecasting market spikes or crash than call volume does.

Key Words: 내재변동성,변동성 스큐,점프,정보효과,프로빗 모형,Implied Volatility,Information Contents,Jumps,Probit Model,Volatility Skew