A LOOK AT SHORT- AND LONG-TERM NONLINEAR DYNAMICS IN FAMILY BUSINESS STOCK RETURNS LISTED ON CASABLANCA STOCK EXCHANGE

Abstract

Moroccan family business companies are becoming an attractive investment in Casablanca stock exchange. In this study, presence of fractal and chaos in Moroccan family business stock returns is examined. In particular, return series are analyzed by stationary wavelet transform to decompose original series into long- and short-term variations. Then, detrended fluctuation analysis and artificial neural networks are used to estimate, respectively Hurst exponent and largest Lyapunov exponent. Results from fractal analysis show that short variations in returns are anti-persistent, whilst long variations are persistent. Besides, results from chaos tests reveal that both short and long variations in returns are chaotic. Such conclusions should be taken into account in active portfolio optimization and management.