Study On Option Pricing Models

Although option pricing models have been studied for long,in China we have little empirical research on our own vanilla plain stock options since we didn't have any until February last year in 2015.Now with data covering over one year,we can do research to check whether the classical or the newest model is suitable for our option market.According to Merton(1974),there are four criteria that are used to assess option pricing models,this paper focuses on one of them which is whether a model can capture the reality.Base on this,firstly,I introduced there models:the Black-Scholes option pricing model?the double fractional option pricing model?the temperd stable process and Monte Carlo simulation option pricing model.The first model here is the classical one,but like many researches in the past,it does not fit the 50ETF option price well due to the 50ETF return has power-like tails just as many other stock markets do.The second model is based on the log-levy process which captures the characteristic of many more extremely plunged price than people thought.So this model gives the put option price compensation,especially the one that is deeply out of the money.This paper did not do empirical research on the last model considering the complex computation,instead,the Monte Carlo simulation to price model is performed based on the geometry Brownian process since it is much more easier while have the similar researching process.The empirical result shows that this model causes many errors which is the result of some combined reasons and surely the TS model can solve these problems.This paper finally looks forward to volatility models and reforms in the newly-born stock option market.