Volatility of High-Frequency Returns on Foreign Exchange and Stable Innovations
Ece Oral1, Evrim Oral2
1Research and Monetary Policy Department, Central Bank of the Republic of Turkey, Ankara, Turkey; 2School of Public Health, LSU Health Sciences Center, New Orleans, LA, United States

Many models in finance are often based on the assumption that the random variables follow a Gaussian distribution. It is now well known that empirical data have frequently occurring extreme values and cannot be modeled with the Gaussian distribution. The stable distributions, a class of probability distributions that allow skewness and heavy tails, have received great interest in the last decade because of their success in modeling financial data that depart from the Gaussian distribution. This study examines the statistical distributions of intra-daily TRY/USD foreign exchange changes. The volatility of the return series are calculated using the Stable GARCH models. It is found that the GARCH model with stable innovations fits returns better than the Normal distribution.

Keywords: Stable distribution; Parameter estimation; high-frequency data; Conditional heteroskedasticity

Biography: Ece Oral has a M. Sc. in statistics from Middle East Technical University and a Ph.D. in statistics from Hacettepe University of Turkey. She worked as a visiting assistant professor at the Department of Economics and Finance at the University of New Orleans, between January 2010 and July 2010. She is currently working in the Central Bank of the Republic of Turkey.