Abstract:

The purpose of this paper is to analyze the returns of the BEL-20 index in the search for evidence of long-term memory. An econophysics approach was used to test and measure the degree of dependence of the Belgian market. The Hurst exponent, estimated through Rescaled-Range Analysis (classical and modified) and Detrended Fluctuation Analysis, evidences slight long memory in the form of persistence. Moreover, Rescaled-Range Tests (classical and modified) and Fractional Differencing Test suggest a stochastic process of short-term memory that degenerates to insignificant levels over long periods. The shocks have persistent impact on returns and, therefore, constitute signals against the weak form of efficiency and imply a potentially predictable component of the index. The results of this paper should be useful to regulators and risk managers.