Forecast Pooling or Information Pooling During Crises? MIDAS Forecasting of GDP in a Small Open Economy

This study compares two distinct approaches, pooling forecasts from single indicator MIDAS models versus pooling information from indicators into factor MIDAS models, for short-term Singapore GDP growth forecasting with a large ragged-edge mixed frequency dataset. We investigate their relative predictive performance in a pseudo-out-of-sample forecasting exercise from 2007Q4 to 2020Q3. In the stable growth non-crisis period, no substantial difference in predictive performance is found across forecast models. We find factor MIDAS models dominate both the quarterly benchmark model and the forecast pooling strategy by wide margins in the Global Financial Crisis and the Covid-19 crisis. Reflecting the small open nature of the economy, pooling single indicator forecasts from a small subgroup of foreign-related indicators beats the benchmark, offering a quick method to incorporate timely information for practitioners who have difficulty updating a large dataset. Nonetheless, the information pooling approach retains its superior ability at tracking rapid output changes during crises.