自從Black-Scholes發表其選擇權訂價理論後,對於選擇權的發展貢獻極大,更加速各種新奇選擇權的發展,其中之一就是彩虹選擇權,也就是選擇權的標的資產不僅一個股票,它可能包括兩個股票(two color rainbow option)、三個股票 (three color rainbow option)或甚至n個股票皆有可能,本研究在介紹兩個標的資產的彩虹選擇權的定價模型,比較雙變數GARCH以及Copula這兩個定價模型的差別。本研究以蒙地卡羅模擬法對標的資產進行股價模擬,而發現到以假設兩標的資間的相關性是常態分配或T分配時,其績效都比假設Copula來的差,而尤其又以動態調整的Copula來的最為理想。
After Black-Scholes advert to the option pricing theory, it has a great contribution to the advance of option;speed up all kinds of Exotic Options, one of it is Rainbow Options. Rainbow Options are their underlying assets which are not only one stock, but it may include two stocks or three stocks even more. Our studies introduce the two color rainbow option of its pricing model, to compare the difference with Bivariate garch model and Copula model. Our study use Monte Carlo Simulation to simulate underlying assets. We discover that if we assume the relationship of two underlying to yield to normal distribution or T distribution it results are worse than Copula. Specifically dynamic Copula are the best.