Macroeconomic risk factors and REITs returns predictability in African markets: Evidence from a new approach

Abstract

This study examines the predictability of REITs returns considering the role of macroeconomic risk predictors for three emerging African REITs markets, namely Ghana, Nigeria and South Africa. To account for the inherent statistical properties of persistence, endogeneity, autocorrelation and conditional heteroscedasticity, we apply the Feasible Quasi Generalized Least Squares (FQGLS) estimator. In addition, the role of asymmetries is considered. Our findings show the heterogeneous performance of the risk factors. All the risk predictors matter in predicting REITs returns in a non-linear fashion in Nigeria, while asymmetries appear to matter only for both interest rate and exchange rate in Ghana, and only interest rate in South Africa. These results suggest that the predictors are important forecast factors for the REITs market. Taking a step further to examine and compare their forecast performance with the traditional model, we find that the macroeconomic risk factors-based predictive model outperforms the traditional ARFIMA model in several cases in the countries, especially Nigeria. Robustness is provided for these results through the use of different sample periods and forecast horizons. Notable policy implications for investors and portfolio managers can therefore be derived from these findings.