AFRREV STECH: An International Journal of Science and Technology

Log in or Register to get access to full text downloads.

Remember me or Register

Macorvian characteristics of the Nigerian stock market

Hamadu Dallah, Ismail Adeleke


Stock Market prediction has been one of the active research areas that have enjoyed attention in the fields of actuarial science and quantitative finance. This article investigates the Markovian characteristics of the Nigeria Stock Market using weekly data on All Share Index (ASI) market, 30- Index and five sub-sectors of Nigerian stock exchange from October 4, 2013 to September 30, 2016. The Chapman-Kolmogorov’s principles of handling transition probabilities and limiting distributions methods were employed for predicting future market behaviour. The findings suggest that compounded returns of the indices for the sectors and the market are highly volatile. The long-run distribution forecasts established that the market converged to stationarity after six weeks, while industrial sector has the shortest stationarity step period of five weeks. It is also observed that it will take about 31 weeks for the market and the 30-Index to reach the best return state, while about 78 weeks period is required to revert to the worst performing state. Further analysis findings of the mean return time suggest that it will take only about two weeks period for the indices returns of the market and the sectors under study to transit to the average state irrespective of the current state. Generally, the findings established the volatile nature of the market and its rapid tendency for deterioration. Finally, it is important to note that the 30-Index and ASI exhibit similar Markovian characteristics. It is pertinent to ensure strict compliance of the 30-Index stocks to the regulatory risk management frameworks for the robustness and sustainability of the market.

Keywords: Markov Process, Nigeria Stock Market, All Share Index, limiting distribution, Mean return and First Passage time, Prediction

AJOL African Journals Online