The fact that most assets are not traded around the clock raises a natural decomposition of the daily return.The intraday return covers the price movement between open and close,while the overnight return cov-ers the price movement between the previous close and current open.Previous literature documented that overnight information has a significant impact on financial activities.It can help explain the market anomalies and improve the volatility forecasting accuracy.However,there is little research investigating the effects of option pricing.In this paper,the daily log returns are decomposed into intraday and overnight components and a new model that extends the Heston-Nandi GARCH framework to a bivariate structure is proposed to describe the two return processes simultaneously.Such decomposition is different from those with high-frequency data(such as semivariance-based good-bad volatility framework)as we only require daily frequency data.Using the variance-dependent pricing kernel,a closed-form option pricing formula is derived and the pricing performance of SSE 50 ETF options is assessed.The empirical results using SSE 50 ETF options from 2015 to 2019 show that distinguishing the overnight component from daily returns can potentially reduce the pricing errors,both in-sample and out-of-sample.The results enrich the current literature on return decomposition by adding a piece of option pricing evidence and call for more research on option pricing in this new direction.
NETL
NSTL
万方数据
