Investigation of Multifractaly Models in Finance

Subject Areas : Financial Knowledge of Securities Analysis

Fraydoon  R. Roodposhti ¹ (Professor of Finance, Faculty member of Islamic Azad University, Science and Research)
Mahdeyeh  Klantari Dehaghi ² (PhD Student of Financial Management, Islamic Azad University, Science and Research)

Received: 2014-09-01 Accepted : 2014-10-17 Published : 2014-12-22

Keywords: multifractaly process, Stochastic criteria, Stochastic Volatility, Forecasting,

Abstract :

Specifying the governing process of stock market’s return with the goal of making optimal decisions and reducing the risk cost has a great importance for investors and policy makers. The importance of market analysis on one hand and the effort for comprehending the markets on the other hand resulted in that, after the assumptions of efficient market were challenged and universal financial facts such as “fat tails” and “volatility clustering” were discovered, analysts leaned toward multifractaly and Lévy models and moved away from models with random characteristics and normal distribution. This caused multifractal models to be used in different segments of the market. In this article the multifractal approach that in recent years has been used for forecasting and modeling volatility, will be examined. In the beginning the origin of this approach that stems from turbulent flows in statistic physic will be introduced and in the following sections the details about the specifications and features of multifractal time series models, available approaches for estimating them and empirical uses of this model will be mentioned. The results of this research show that the dynamic nature of capital market has caused the approaches, methods and models of market analysis to be permanently changing. In addition, for volatility clustering of time series, smaller scales are considered.