期权具有独特的风险规避功能，随着我国金融市场的完善，期权产品的推出被提上日程。如何建立合理的期权定价模型，并对期权合理定价成为政府和金融部门关注的热点。已有的期权定价模型大都在标的资产服从几何布朗运动的基础上得到。然而大量的实证研究表明金融资产对数收益率的分布具有尖峰厚尾性，是有偏分布，波动率呈微笑型，并且存在聚类现象。.首先，本项目引入调和稳定分布为标的资产的价格过程建模，并结合GARCH模型建立能刻画标的资产的上述特性的调和稳定GARCH定价模型。其次，对模型下欧式、美式、奇异期权的定价方法进行研究，提出了一种更加高效的期权定价方法：Markov Chain数值定价法，并证明其收敛性。最后，采用国内外金融市场的实际数据，对调和稳定GARCH模型和Markov Chain方法进行实证研究。研究成果不仅完善了期权定价理论，而且能有效的指导投资者进行风险管理，为政府和金融部门提供理论依据。

The important function of options is hedging. With the development of the financial market in our country, introducing options in Chinese financial market is put on the agenda. The reasonable option pricing model and efficient option pricing method have become major focus of attention from government and regulatory sectors. Previous studies dealing with option pricing usually assumed that the underlying asset followed the geometric Brownian motion. However, numerical empirical analysis illustrate that the distribution of asset log-return are not symmetrical and have the properties of heavy tail and aiguilles, and the volatility has the shape of smile, also there is volatility clustering..We first introduce tempered stable distributions to model the underlying assets, combined with GARCH process, we present tempered stable GARCH model to capture the above properties of underlying assets. Then we introduce a more efficient numerical option pricing method: Markov chain approximation method, and prove the convergence of the method. Finally we apply historical data to do empirical analysis to the option pricing model and method. The results of our research can improve the option pricing theories, conduct investors to avert the risk, and provide the theoretic foundation for our country’s financial policy.