This article produces monthly GDP estimates in response to the need for growth indicators that are updated more frequently, providing timely information about the Turkish economy. In this context, a dynamic factor model and an expectation-maximisation algorithm were applied using high-frequency macroeconomic indicators for the period 1998?2024. A 29-month indicator was chosen to represent the general structure of the Turkish economy, which is characterised by the dominance of the industrial and services sectors. A high level of agreement was observed between the estimated common factor and the quarterly GDP data published by the Turkish Statistical Institute. It was noted that the model was particularly successful in detecting turning points. In addition, significant and strong correlations were found between the leading indicators of the Central Bank of the Republic of T?rkiye and the estimated monthly GDP series. This supports the usability of the indicator in monitoring current economic activity for policy-making purposes. The model is evaluated both insample and out-of-sample, achieving 90% directional accuracy in out-of-sample forecasts. However, the model?s performance deteriorates during structural shifts, such as major monetary policy transitions. Nevertheless, the study contributes to the literature by providing a monthly GDP estimate with a real-time and structural approach for the first time in T?rkiye.
Şahin et al. (Thu,) studied this question.