دنباله قدرت توزیع شاخص در بازار بورس چینی

The power αα of the Lévy tails of stock market fluctuations discovered in recent years are generally believed to be universal. We show that for the Chinese stock market this is not true, the powers depending strongly on anomalous daily index changes short before market closure, and weakly on the opening data.

In summary, we have analyzed the asymptotic behavior of the distribution of normalized 1-min returns for the SSE index. We find that the tail properties of the distribution are caused by the near-closure returns, the relevant fall-off powers being α≃3α≃3. Moreover, for time scales ΔtΔt from 1 min up to 64 min, the distribution shows a slow convergence to Gaussian behavior as required by the central limiting theorem. It is argued that to remove opening and closure returns in the study of high-frequency changes of the emerging stock markets qualifies as a standard empirical methodology and should be taken into account as a standard caveat in empirical research.