Summary: | 碩士 === 國立中山大學 === 應用數學系研究所 === 93 === Under path dependent consideration, we discuss vulnerable option pricing problem. Two pricing
models are proposed: Model(1) use stepwise regression and Monte Carlo simulation, and Model(2) is based on
multi-level regression method. Since the option price was approximated by quadratic surface at each time point
in Model(1), large mean square errors are induced. Therefore, we further propose a stepwise subset regression
method to improve Model(1) approach. At present, this proposed method can compute the option price accurately
for no credit risk options. For Model(2), we utilize a multi-level regression method to price vulnerable
options, and simulation results show that the method can also obtain accurate option prices.
|