“…First, there are numerous methods suitable for investigating possible linear dependence of return time series, 25,34,54 such as the portmanteau Q statistic, 40 the AR model, 27 the runs test, 12 and the variance ratio test. 43 Second, there are also methods for testing nonlinear dependence in return time series, 25,34,38,39,54 such as the test for ARCH e®ect, 47 Tsay's test, 53 the ARCH Lagrange multiplier test, 16 the bicorrelation test, 24 Broock et al test, 8 and the generalized spectral test. 17 Third, there are econophysical methods for the test of long-range dependence by estimating the Hurst index, 26 such as the rescaled range analysis [44][45][46] the structure function approach, 14,15 the detrending moving average analysis, 1,9,22 the detrended°uctuation analysis, 29,31 and so forth.…”