<![CDATA[Although volatility modeling in emerging stock markets has received increasing attention, limited research has jointly compared GARCH-type model structures under alternative symmetric and skewed innovation densities in the Nigerian capital market. This study aims to evaluate the forecasting performance of selected GARCH-type models under alternative innovation densities using daily returns of the Nigerian Stock Exchange All Share Index (NSE-ASI) from February 2012 to July 2023. A quantitative econometric time-series design was employed, involving 2,820 daily observations selected through purposive sampling based on data availability. Data were obtained from the official market database and analyzed using Maximum Likelihood Estimation, model selection criteria comprising Log-Likelihood (LL), Akaike Information Criterion (AIC), and Bayesian Information Criterion (BIC), and forecast accuracy measures including Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE). The findings indicate that the APARCH(1,1)-GED model provides the best in-sample fit, whereas the APARCH(1,1)-SGED specification produces the most accurate out-of-sample forecasts. These results demonstrate the importance of innovation density selection in capturing asymmetry and fat-tailed behavior in stock return volatility. The study concludes that incorporating skewed heavy-tailed distributions enhances volatility forecasting accuracy in the Nigerian capital market. The findings contribute to the theoretical development of conditional heteroskedasticity modeling and offer practical implications for risk management, portfolio analysis, and regulatory forecasting in emerging markets. Future research may extend this work by examining advanced nonlinear and regime-switching volatility models across broader emerging market contexts.]]>
