engleski

How accurately stock market expectations can be predicted?

This research is about the use of high frequency data for determining the prognostic power of the option pricing models. This research aims to forecast the future trends in the expectation, variance and other moments of the financial time series. The financial time series that are the scope of this research are the put and call options of European market indices. The research involves two steps. The first step is the estimation phase of forecasting the probability density function of the observed financial time series. The second step is a comparison of the estimated probability density functions with a benchmark density function based on high frequency data. The subject of research are the option pricing models used to predict the future risk neutral probability density function. The aim is to estimate and compare option pricing models. The purpose is to evaluate their prognostic power and to asses which of them has the best fit. This study gives estimation of probability density function of DAX index on specified expiration date and comparison with the benchmark density function. There are limitations in the use of high frequency data on illiquid financial markets and therefore the lack of data to analyse. The results provide the contribution to the existing literature as the benchmark density function is the one on the basis of high frequency data. Methods of comparing the benchmark density function with the estimated risk neutral probability function give applicative results for market participants and public authorities. Provided results give better insights in high frequency data issues and therefore motivation for a further research regarding volatility estimation.

High frequency data ; option pricing models ; probability density function ; benchmark

nije evidentirano