Parameter Estimation Of Stock Price Jump-diffusion Process

This paper is based on the fluctuation law of underlying asset price, which is oneof the important problems of financial derivatives pricing.Stochastic volatility is the most fundamental characteristics of the stock market.Usually Brown motion is introduced to model the stock price given that the stockprice volatility is a constant. In fact，in the real financial market, the occurrence ofmajor events on the market may cause the fluctuation of stock price which appearsdiscontinuous jump. So in the experimental research papers, it was found that thestock prices were discontinuous with jumps.This paper focuses on the parameter estimation of jump-diffusion model, andthe main methods used are the optimal clustering and moment estimation. Afterobtaining the parameter estimators, Monte Carlo simulation is used to test theaccuracy of estimators in the sample data and out the sample data respectivelyThrough the simulation results we found that there is small error in the short term,and long term prediction error was higher. That is to say. the model can be used topredict in the short term with higher accuracy.