A new methodology able to handle unbalanced panel with mixed frequencies data in the context of a multiple dynamic factor model with a block structure (e.g., Kose, Otrok and Whiteman 2003) is proposed. Therefore, monthly and quarterly data at country levels can be included in order to decompose national business cycles into a common and a country-specific component. By introducing high frequency data to the model, a wide range of economic indicators for each of the 20 countries in the sample are included, resulting in a proper characterization of business cycle fluctuations both, across countries and at a world level. The existence of a world factor even at monthly frequencies is confirmed, although its relative importance differs among advanced and emerging economies. Major world and country-specific economic events over the last 25 years are captured well by the model. A Monte Carlo experiment is also conducted in order to verify the proposed methodology.
Resum
Leave Comment