Abstract:
This study examines how effectively common factors, extracted using both the partial least squares method and principal component analysis from the business tendency survey and the banking loan tendency survey, can predict Turkiye's economic growth. The findings indicate that integrating this survey data with macroeconomic variables has the potential to improve the accuracy of Turkiye's real GDP growth predictions. When examined at the sector level, models employing factors from the Durable Consumer Goods sector exhibited the strongest predictive capabilities. Regarding firm size, models based on factors from large companies yielded superior out-of-sample prediction performance. Moreover, refining the prediction models by strategically reducing the number of factors using variable selection algorithms and choosing the most significant ones further enhanced their forecast accuracy. In conclusion, this study offers invaluable insights for policymakers, investors, and households in Turkiye by introducing a new approach to improving the accuracy of economic growth forecasts.
