Open Access Open Access  Restricted Access Subscription Access
Cover Image

Dual Long Memory Properties and Stock Market Efficiency in External Shocks: Evidence from Leading Oil-Dependent Markets

Salma Jaghoubi, Bassem Salhi, Theodore Syriopoulos

Abstract


Long memory dynamics, persistence of volatility and market efficiency remain important information to investors and have had a great influence in theoretical and empirical finance. In this paper, we revisit and test the Efficient Market Hypothesis (EMH) to contribute challenging updated empirical evidence from leading oil exporting and importing markets, namely Saudi Arabia and the US respectively, taking into account potential implications from the recent 2014 petroleum crisis. To this aim, ARFIMA and FIGARCH model specifications are incorporated to test for long memory properties in stock market return and volatility persistence, in the context of the weak EMH form. It was seen that these oil-importing and oil-exporting markets exhibit different behaviour in returns and volatilities. To test simultaneously for stock market return and volatility behavior, a robust quantity technique, namely Autoregressive Fractionally Integrated Moving Average Fractionally Integrated Generalized Autoregressive Conditional Heteroscedastic or ARFIMA - FIGARCH model is estimated before and after the petroleum shock, under the Student’s t-distribution. The findings indicate that dual long memory returns are detected in the Saudi Arabian stock market during the pre-oil crisis period but this is not supported in the post-oil crisis period. In contrast, dual long memory is not detected in the US stock market during the sample period, implying an unpredictable path in return and volatility behavior, in convergence to the EMH framework. The recent petroleum shock is seen to reduce the estimation accuracy of stock market return dynamics and to weaken return persistence. This evidence  provides  useful  insights  to  participants  in  oil  and  stock  markets  and  to policymakers. 


Full Text:

PDF

Refbacks

  • There are currently no refbacks.