Az előadás a Széchenyi teremben 10:30 órakor kezdődik ,angol nyelven.
Victor Bystrov (EUI)
Abstract
The adoption of inflation targeting in emerging market economies makes accurate forecasting of inflation and output growth in these economies of primary importance. Since only short spans of data are available for such markets, autoregressive and small-scale vector autoregressive models can be suggested as forecasting tools. However, these models include only a few economic time series from the whole variety of data available to forecasters. Therefore dynamic factor models, extracting information from a large number of time series, can be suggested as a reasonable alternative. In this paper two approaches are evaluated on the basis of data available for Brazil and Russia. The results allow us to suggest that the forecasting performance of the models considered depends on the statistical properties of the series to be forecast, which are affected by structural changes and changes in operating regime. This interaction between the statistical properties of the series and the forecasting performance of models requires more detailed investigation