MIXED-FREQUENCY AND MARKOV-SWITCHING MODELS FOR GDP GROWTH: A COMPARATIVE ANALYSIS OF DEVELOPED AND EMERGING ECONOMIES

This thesis evaluates the in-sample performance of mixed-frequency and Markov-switching models for GDP growth, using industrial production as the high-frequency indicator. Unlike most existing studies, which focus on developed economies, this research applies these models to various countries, of which five are developed and five are emerging economies. Results show that Markov-switching models generally achieve superior in-sample fit, though improvements are not always statistically significant. Developed economies exhibit stronger and more consistent performance than emerging economies, partly due to differences in sample size and volatility. Nonetheless, regime-identification scores indicate that these models capture different economic regimes across both groups of countries.