一种寻找Heston期权定价模型参数的新方法

Heston随机波动率模型的期权定价比Black-Sholes模型更符合市场情况,是金融衍生品定价研究的热点。但应用时需要确定五个待估参数，参数的确定属于组合优化问题，此问题的求解通常比较困难。本文利用遗传算法解决该优化问题,从而得到Heston模型的待估参数。该算法避免丢失最优解，具有群体搜索的特点，有着很好的概率跳出局部极小值,从而以概率1收敛到全局极小值。在实证研究中,利用香港恒生股票指数期权在2014年6月10日和2014年6月25日交易的数据为样本,得到待估参数,并用该参数对2014年6月12日的买入期权和2014年6月27日的卖出期权进行了模拟定价。数值结果与进化过程表明本文方法的有效性和可行性。

A New Method of Finding the Parameters of Heston's Option Pricing Model

In this paper,we use the Genetic Algorithm together with the minimum residual sum of squares to estimate the 5 parameters in Heston model. This method allows us jump out of the local minimal values at certain probability,then converges with a probability 1 to the global minimal value. It also has the characteristic of avoiding missing the optimal solution and the ability of group search. In the empirical study,we have determined the suitable parameters for Heston model by using Hong Kong Hang Sheng equity index call option data on June 10,2014 and put option data on June 25,2014. Afterwards,we have used these parameters to price the index call options for June 12,2014 and put options for June 27,2014. The numerical results and the evolutionary process show that the method is effective and feasible for application.