Analysis of Nigeria Stock Market Using Bayesian Approach in Stochastic Volatility Model (2012 – 2016)

摘要

This research was conducted to analyzed volatility in Nigeria stock market using Bayesian approach in stochastic volatility model. The sample data used for this study were daily and weekly closing prices of All Share Index over the period of January 30~(th), 2012 to December 8~(th), 2016. The analysis was carried out based on posterior credible interval estimates produced by running MCMC algorithm for 10,000 iterations, with a burn-in of 1000000 and the first 1000 iterations discarded. The result of the study revealed evidence of significant variation in the price movement, as shown by the large differences between minimum (-4.35316) and maximum (7.984833) with mean estimate (0.018). The result for unit root test based on Posterior means and 95% posterior intervals for and shows that there is a significant evidence for unit-root in log- volatility model for All share index (the corresponding 95% posterior intervals include the point 1) i.e [0.9995,1.0]. The inter-correlation among the Posterior parameters revealed that there is a less or no collinearity between the parameters of estimate. Mean estimate for were found to be 0.7504 and 2.286 respectively for daily and weekly share price, indicating volatility clustering in the series. The study that revealed volatility persistence were significantly high with mean estimated close to unity for daily and weekly series respectively. The estimate for leverage parameter showed -0.566 and -0.561 for daily and weekly respectively. Arising from this study, stochastic volatility model could be used to test price movement in Nigerian stock market to make it better utilized by financial experts, econometrician and researcher.