The Optimal Hedging Ratio

Hedging is that people buy or sell the goods amount is equal to but opposite to the direction of trading in commodity futures contracts, so that can offset the loss cause by spot markets price adverse change thought sell or buy future contracts. In other words, the hedge is to avoid the risk of spot prices for the purpose of the Futures Exchange, as investors avoided it to transfer or divert price risk provides a useful tool.Naive hedging strategy is hedging ratio of 1.0, but it did not use any information on the market, so it is not ideal. Traditional hedging strategy is constant hedging method, the cash assets return regress to futures assets return, by the least squares method we can achieve optimal hedge ratio, but inadequacy of this method is disturbance may exist hereto- elasticity, and do not reflect market information on the impact of portfolio assets. Recently the more popular method is time-varying hedge ratio, this approach reflect market information on the impact of portfolio assets, and in-sample has good nature.Hamilton (1989) offers a Markov state transition model and its application widespread in econometrics increasingly, and Hamilton and Susmel(1994) offers SWARCH model, think variance is change over time. However the paper construct a model is under Hamilton and Susmel(1994)model and Brooks, Henry and Persand(2002) model, offer model that hedging ratio is over state, that is hedging ratio that not only change over time, but also change with the state. Finally come to a new hedging ratio based on Bayesian theory. And as an example of WUGANG warrant, compared with the previous method of calculation, show that in sample good or bad of model is not very clear, Not-hedging model have small risk, but also have small return, and naive-hedging model have high return, but have high risk, but out of sample, the best model is the model of hedging ratio over time, its return is biggest and risk is lest. It said when you hedge WUGANG warrant, in sample, can not hedge, because its risk is very small, out of sample could use model of hedging ratio over time, whether in sample or out of sample, can not use model of hedging ratio over state. Although the empirical results of the article did not reflect the superiority of this model. This is not negating the use of this model, because, after all is very special to WUGANG warrant, but this model also reflects the sub-optimal nature, it changes with the state of the hedging model is a model worth considering.