Yeni bir hisse senedi fiyatlama modelinin parametre kestirimi

In this study, a stock price process is considered as an integral with respect toa Poisson random measure which governs several parameters of the trading agents. This model is powerful since it reflects two important properties of high frequencyfinancial data, long-range dependence and self-similarity. We estimate parameters of this model using real data. The estimation procedure is demonstrated on log-returnsof a particular stock in banking industry from Istanbul Stock Exchange between February 2007 and December 2009. We estimate the Hurst parameter describing long range dependence. The numerical values found here verify the long-range dependenceassumption. We also estimate the order duration parameter which follows a Pareto distribution. Interarrival rate of orders are calculated under the assumption that they arrive according to a Poisson Process. Effect rate and effect function are numericallyfitted to the price data.