In this paper we use a battery of various mixed-frequency data models to forecast Czech GDP growth. The models employed are mixed-frequency vector autoregressions, mixed-data sampling models, and the dynamic factor model. Using a dataset of historical vintages of unrevised macroeconomic and financial data, we evaluate the performance of these models over the 2005–2014 period and compare them with the Czech National Bank’s macroeconomic orecasts. The results suggest that for shorter forecasting horizons the CNB forecasts outperform forecasts based on the mixed-frequency data models. At longer horizons, mixed-frequency vector autoregressions and the dynamic factor model are able to perform similarly or slightly better than the CNB forecasts. Furthermore, moving away from point forecasts, we also explore the potential of density forecasts from Bayesian mixed-frequency vector autoregressions.
Journal of Business Cycle Research, vol. 12(2), pages 165-185
David was a research fellow who contributed a better understanding of the current and future economic environment of a changing…David was a research fellow who contributed a better understanding of the current and future economic environment of a changing region. He was mainly interested in the development of models for policy analysis and forecasting. At the same time, he was involved in projects related to the Vision 2030 program, focusing on the economic transformation and diversification of the Saudi economy.
Prior to joining KAPSARC, David worked at the European Commission, European Central Bank, Moody's Analytics and the Czech National Bank. In these institutions, he participated in economic policy analysis, forecasting and research. He also served as a consultant to central banks in the CEE region, managing a variety of economic modeling projects. David led courses in econometrics and operations research during his Ph.D. studies.