In this study, the information content of diverse range of economic variables for Tehran house price forecast has been examined and then the performance of some forecast combination methods for this target variable has been evaluated. The results show that using the information of various variables through forecast combination techniques can improve the forecast accuracy. Among these techniques, simple combination methods are more accurate than the optimal weight methods although latter has theoretical background. Also, in general, putting more emphasis on the recent forecasts (in the discounted squared error methods) and less information aggregation (in clustering methods) can improve forecast accuracy. On the other hand, shrinking weights toward the equal weights (in shrinkage methods) can improve forecast performance through reducing estimation error.
atrianfar, H., barakchian, S., & fatemi, S. (2013). Evaluation of Forecast Combination Methods. Journal of Applied Economics Studies in Iran, 2(6), 123-138.
MLA
hamed atrianfar; seyedmehdi barakchian; seyedfarshad fatemi. "Evaluation of Forecast Combination Methods". Journal of Applied Economics Studies in Iran, 2, 6, 2013, 123-138.
HARVARD
atrianfar, H., barakchian, S., fatemi, S. (2013). 'Evaluation of Forecast Combination Methods', Journal of Applied Economics Studies in Iran, 2(6), pp. 123-138.
VANCOUVER
atrianfar, H., barakchian, S., fatemi, S. Evaluation of Forecast Combination Methods. Journal of Applied Economics Studies in Iran, 2013; 2(6): 123-138.
)