This paper examines the effects of macroeconomic variables on the returns of a broad cross-section of emerging stock markets (ESMs) for a relatively recent time period. Specifically, the paper examines the quarterly data of select local and global macroeconomic variables for 9 ESMs over the period 2002-09 using the same methodology that was applied in Fifield et al. (2002) on similar sets of data. Applying the methodology used in Fifield et al. (2002) we find that the local economic variables included in the study can be summarized by net exports, interest rates, and currency, while global variables can be summarized by world-market returns and US interest rates. The paper uses principal component analyses (PCA) to reduce the number of the variables. The principal components (PCs) are then selected by way of ad hoc rules-of-thumbs. A scree test is then applied in conjunction with an analysis of the acceleration factors of each scree plot to provide robustness. Essentially, a minimum of 0.5173 to a maximum of 0.7775 of the variation can be explained by the first PC, while approximately 0.76 to 0.95 of the cumulative variance can be explained by both the first and second PC. We retain the first and second PCs; thus, we can reduce the dimensionality of the variables from six to two variables. The retained PCs are used as inputs into two regression analyses in order to explain the variation of index returns within each of the 9 ESMs over the period 2002-09. The first regression analysis only includes PCs retained that contain global macroeconomic variables, while the second includes both the PCs that contain global macroeconomic variables as well as PCs that contain information at the local level or local macroeconomic information. The R2 and adj. R2 of each regression analysis was compared for robustness. The regression analysis indicates that while global factors are consistently significant with a high degree across the cross-section of ESMs when both the first and second recession analysis is investigated, local factors, do not show consistent significance across the cross-section of ESMs when the second regression analysis is investigated. Additionally, we use the retained global and local PCs as inputs for a third regression analysis in which the residuals of the first model are used as an input for the dependent variable in order to make sure the improvement in the R2 and adj. R2 between the first and second regression analysis are attributed to a robustness versus general improvements of R2 and adj. R2 due to adding additional variables. After examining the R2 and adj. R2 we find that although the first regression analysis has a relatively higher R2 and adj. R2 compared to the second linear mode the first linear model does not provide a high enough R2 or adj. R2. Essentially, both linear models lack predictive prowess because Additionally, the second linear model does not show much improvement to the first when we add additional explanatory variables. This was validated when we examined the R2 and adj. R2 of the third linear model as both variables were significantly lower than the R2 and adj. R2 of the first model. Furthermore, for certain ESM the variance among local variable show a degree of significance, but they do not show the same high degree of significance as compared to the level of significance indicated by the global macroeconomic variables. Finally, cross-validation (CV) was applied to both models. We find that for the ESM that had significant local variables for some α the second model had a lower mean squared error (MSE) compared to the MSE of the first model.