Improving Futures Hedging Performance Using Option Information

The drastic fluctuation of financial market often makes investors suffer great systemic risk.Then,more and more financial derivatives were designed to avoid the special risk.In particular,the appearance of futures contracts has become a tool of avoiding stocks' price volatility.This paper will focus on how to estimate the optimal futures hedge ratios.Many researches describe different theoretical models to the optimal futures hedge ratios,with most using historical volatility information from stock and futures data.The relevant unconditional hedging models don't consider the time-varying distribution of the spot and futures price,such as ordinary-least-square model and vector error correction model.In addition,the conditional hedging approaches including the random coefficient autoregressive model and the generalized autoregressive conditional heteroscedasticity model,but these time-varying hedge ratios' performances are not always better than unconditional hedging performances.Based on above method can't get consistent results.This paper proposes a optionimplied approach to estimate the optimal dynamic hedge ratio.Under this method,we calculate the correlation between the spot and futures returns using spot modelfree implied volatility and futures volatility.Then we compute the estimated hedge ratio.This article consists of the following parts.The first chapter is an introduction of this article.Summing the advantages and disadvantages of different existing models.On the same time,taking a simple introduction about option implied hedging strategies and pointing theoretical and practical significance.The second part is an expatiation of all mentioned models,including the least squares regression model,naive hedging model,error correction model,random coefficient autoregressive model,correlation coefficient autoregressive conditional heteroscedasticity model,dynamic correlation coefficient autoregressive model,asymmetric dynamic correlation coefficient autoregressive model and option implied correlation coefficient autoregressive conditional heteroscedasticity model,option implied dynamic correlation coefficient autoregressive model and option implied asymmetric dynamic correlation coefficient autoregressive model.On the besides,choosing evaluation indexes of the performance of the in-sample and outof-sample.These are variance reduction,value at risk 1%,value at risk 5%,expected loss 1%,expected loss 5%,sharp ratio reduction and the mean-variance utility reduction.Next part is also the most important part,we evaluate the in-sample performance of various hedging models using S&P 500 indices.The empirical results show option-implied hedging processes generate better in-sample performance than the relevant conditional hedging processes based on most evaluation indicators.And we choose from 1st January,2007 to 5th November,2013 as window period,to forecast the optimal hedging ratios from 6th November,2013 to 31 th October,2016.The out-of-sample performance of the error correction model and the option implied asymmetric dynamic correlation coefficient autoregressive model is relatively better,and they have the optimal performance under the different dimensions.After considering transaction cost,option-implied hedging strategies also perform well.It is noteworthy that transaction cost is also considered as an important factor to evaluate the performance of different models.Finally,putting forward contribution suggestions on the hedging strategy selection in different situations.And pointing out the deficiencies of this paper and future research points.