Forecasting Nigerian Stock Exchange Returns: Evidence from Autoregressive Integrated Moving Average (ARIMA) Model

19 Pages Posted: 22 Jul 2010  

Emenike Kalu O.

University of Nigeria - Department of Banking and Finance

Date Written: June 1, 2010

Abstract

It is no news that the global economic crisis has led to shortage of financial resources and to a general downturn in stock prices across the globe. Forecasting stock prices will provide a way to anticipate and perhaps avoid the risk of a large adverse change in prices. This paper, therefore, models and forecasts stock prices of the Nigerian Stock Exchange using the Autoregressive Integrated Moving Average (p, d, q) model. Monthly All-Share Indices of the NSE from January 1985 to December 2008 provide the fit sample, whereas January 2009 to December 2009 provide out-of-sample forecast period. Several diagnostic tests were performed to select the p, d, q parameter that best fit the index. The selected ARIMA (1, 1, 1) model predicted index points and growth rates deviated from the actual indices and growth rates. The predictions failed to match market performance during the forecast period. As a result, the adequacy of the model was questioned by generating one-period forecasts for the subsequent 12 periods and theil U statistics shows that ARIMA (1,1,1) model forecast outperformed the naïve Model. Hence, the deviations indicate that the global economic meltdown destroyed the correlation relationship existing between the NSE All-Share Index and its past.

Keywords: Forecasting, Stock Price, ARIMA Model, Nigerian Stock Exchange

JEL Classification: C22, C52, G17

Suggested Citation

O., Emenike Kalu, Forecasting Nigerian Stock Exchange Returns: Evidence from Autoregressive Integrated Moving Average (ARIMA) Model (June 1, 2010). Available at SSRN: https://ssrn.com/abstract=1633006 or http://dx.doi.org/10.2139/ssrn.1633006

Emenike Kalu Onwukwe (Contact Author)

University of Nigeria - Department of Banking and Finance ( email )

Enugu
Nigeria
+2348035526012 (Phone)

Paper statistics

Downloads
309
Rank
78,658
Abstract Views
1,180