Jump Risk Implicit in Options Market
提出从期权数据中恢复半矩和累积矩的方法，推导基于双指数跳跃模型的跳跃风险度量，发现其优于现有度量，并能预测标普500指数的已实现方差和超额收益。
Abstract We propose a simple procedure to recover (semi-)moments and cumulants from option data. We further derive jump risk measures based on a general asset return model with double-exponential jumps. Numerical and empirical results show that our jump variation measures outperform existing measures under specific conditions. Using return and option data on the S&P 500 index, we examine the information content of our measures, with a focus on large jumps (LJ). Our measures contribute to market realized variance and excess return prediction suggested by in- and out-of-sample tests. Accounting for LJ identified by jump variation improves market return forecast, implying a distinct impact of large and non-LJ.