Testing for Sibex Market’s Long-Term Memory

Long-term memory or long-term dependencies of financial time series’ returns allows testing the efficiency of capital markets. This paper investigates the long-term memory in Sibex market by both static and dynamic approaches. Identifying the long-term memory in futures market can be useful for detecting arbitrage opportunities in this market and thus for managing portfolios of financial assets. For measuring the long-term dependencies we used the Generalized Hurst Exponent because it has the smallest estimation error compared to other methods. The implementation of GHE test on rolling windows allows for robust results concerning the degree of markets’ informational efficiency in time.