(a) EGX 30
(b) MOD1 ()
(c) MOD2 ()
Figure 7: Estimation of self-similarity parameter . for raw and absolute returns, respectively, for (a) the EGX 30, (b) the MOD1, and (c) the MOD2, respectively. To estimate the self-similarity parameter we follow Peng et al. [57] and perform a detrended fluctuation analysis (DFA). A linear relationship on a log-log scale plot indicates the presence of power-law scaling. The slope of the line relating to is the estimated scaling exponent, . The scaling exponent yields a value of for the EGX 30, for the MOD1, and for the MOD2, which are close to the theoretically expected value of the white-noise process. The scaling exponent reveals a value of for the EGX 30, for the MOD1, and for the MOD2, which indicate persistent long-range (power-law) autocorrelations in absolute returns for the three time series under investigation.