Bivariate American Option Pricing Using GARCH Copula LSM Method
碩士 === 國立臺灣大學 === 財務金融學研究所 === 95 === A bivariate American option is an option with two underlying assets. In this thesis, a new pricing technique for bivariate American option pricing is presented. The pricing technique is implemented using GARCH, copula and the least squares Monte-Carlo (LSM) algo...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2007
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/36343417670590869326 |
| Summary: | 碩士 === 國立臺灣大學 === 財務金融學研究所 === 95 === A bivariate American option is an option with two underlying assets. In this thesis, a new pricing technique for bivariate American option pricing is presented. The pricing technique is implemented using GARCH, copula and the least squares Monte-Carlo (LSM) algorithm. The individual asset return process is described using the GARCH model. The correlation between the underlying assets’ returns is represented by modeling the joint distribution of the innovations of their respective GARCH processes. The joint distribution of the innovations is modeled by different copula models and traditional bivariate normal setting. Finally, the bivariate American option price is calculated by the LSM method. The numerical results of bivariate basket American call option are presented. Different copula models and bivariate normal model of innovations are provided. The results show that American option prices implied by GARCH-copula-LSM models can differ substantially from the prices implied by traditional bivariate normal model. The traditional bivariate normal innovations setting will overestimate the value of the bivariate American call options.
|
|---|