Emenogu, N.G.Adenomon, Monday OsagieNweze, N.O.2023-12-142023-12-142019-06-05Adenomon, M.S. et al. (2019) MODELING AND FORECASTING DAILY STOCK RETURNS OF GUARANTY TRUST BANK NIGERIA PLC USING ARMA-GARCH MODELS, PERSISTENCE, HALF-LIFE VOLATILITY AND BACKTESTINGhttps://keffi.nsuk.edu.ng/handle/20.500.14448/6237This study investigated the forecasting ability of GARCH family models, and to achieve superior and more reliable models for volatility persistence, half-life volatility and backtesting, the study combined the ARMA and GARCH models. The study modeled and forecasted the Guaranty Trust Bank (GTB) daily stock returns using data from January 2, 2001 to May 8, 2017 obtained from a secondary source. The ARMA-GARCH models, persistence, half- life and backtesting were used to analyse the data using student t and skewed student t distributions, and the analyses were carried out in R environment using rugarch and performanceAnaytics Packages. The study revealed that using the lowest information criteria values alone could be misleading so backtesing was also carried out. The ARMA(1,1)-GARCH(1,1) models fitted exhibited high persistency in the daily stock returns while it took about 6 days for mean-reverting of the models, but failed backtesting. However, backtesting showed that ARMA(1,1)-eGARCH(2,2) model with student t distribution passed the test and was suitable for evaluating the GTB stock returns, and required about 16 days for the persistence volatility to return to its average value of the stock returns. The study recommended addition of backtesting approach in evaluating the performance of GARCH model in order to avoid misleading results. Also, the GTB stocks can be predicted since most of the estimated models were stable.enStock returns, Guaranty Trust (GT) Bank, Generalized Autoregressive Conditional Heteroskedasticity (GARCH), Persistence, Volatility, BacktestingArticle