MCMC Analysis of Stochastic Volatility Models Joining Evidence from Spot and Option Prices
2004/03/09
研討會日期 : 2004-03-09
時間 : 15:00
主講人 : 鄭藹如女士
地點 : B棟110室
演講者簡介 : 鄭藹如女士為Ph.D. Candidate in Economics， University of North Carolina at Chapel Hill (Ph.D. expected May 2004)。
其主要研究領域為Finance及Econometrics。
演講摘要 : This paper proposes a Markov Chain Monte Carlo (MCMC) method to estimate a class of stochastic volatility (SV) models, including one- and two- factor log volatility models, using information from both the underlying asset and option markets. Parameters for both objective and risk-neutral measures are estimated simultaneously. Since option prices ought to be computed at each iteration of MCMC sampler, the lack of closed-form pricing formulas for the log volatility models causes a serious problem. This paper uses a new method to numerically approximate these option prices. The computational burden is considerably reduced, making the overall estimation problem feasible. Using the information embedded in option prices enables us to obtain more precise estimates than can be obtained from returns alone.
