GARCH模型下的美式期权模拟定价--来自中国权证市场的经验

American style option simulation pricing under GARCH models——Evidence from Chinese warrant Market

分别运用GBM模型和GARCH模型对我国市场上的股票价格进行参数估计,并使用最小二乘蒙特卡罗模拟方法对我国具有百慕大性质的权证进行蒙特卡罗模拟定价,发现GARCH模型的定价效果明显优于GBM假设下的定价,虚值程度越高的权证,定价误差越大。定价误差的对数与上证综合指数的对数之间存在明显的协整关系,权证没有卖空机制,使得套利无法实现,投机气氛较浓,是我国权证的市场价格明显高估的重要原因。实证对GARCH条件下的定价误差更加具有解释力。

We estimate the parameters of GBM model and GARCH model using stock prices on Chinese market. and then simulate the prices of warrants on the market by LSM,we found that GARCH models are better than GBM models in the pricing ,when it is out of moneyness, the Pricing error is large. Log of pricing error and that of Shanghai Composite Index have the significant co-integration relationship, the reason why the market prices are higher than the theory prices is that no mechanism of short of warrants make arbitrage can not be achieved. GARCH models'pricing error can be explained more properly by empire study.