추계학술연구발표회
An Empirical Study on the Statistics Properties of Time Dependency using High-Frequency Data of KOSPI and KOSPI200 Futures(정대성, 엄철준)
작성자 관리자
등록일2009.06.09
조회수6414
We investigate the scaling behavior of the Korean financial market, this study examines the statistical properties of KOSPI and KOSPI200 Futures using methods including probability density function, cumulative distribution function, autocorrelation function, and conditional probability across time scales. This study confirms a few time dependent features of financial market: (1) the center part of the return distribution aggregating to that of Gaussian distribution, (2) the tail parts deviating from the Gaussian and the Levy distribution, (3) a short range correlation for returns and a long range correlation for absolute returns, and (4) volatility clustering.
×