Alphacast Highlight – High-Frequency Global Growth
Much to analysts' and policy-makers chagrin, activity data often comes with substantial lags. Quarterly and/or monthly GDP reports reflect what happened in the recent past and they might not offer timely information about the present and future. Machine learning techniques allow researchers to fill that gap, by producing almost real-time estimates. One of such efforts is conducted by the OECD, which produces a weekly GDP growth tracker by applying a machine learning model to a panel of Google Trends data for 46 countries (see further information here). Check the OECD’s weekly growth tracker dataset here Global growth cooled down after the post-Covid 19 rebound. Central banks have been reacting to soaring inflation, with policy rates increasing worldwide. Not to mention that they have also started normalizing their balance sheets (the so-called "tapering"). It is not surprising that economic activity takes some hit, with spillovers everywhere. The first chart shows the simple average of OECD’s growth estimates for each country. Of course, any average hides heterogeneity. As expected, most developing economies are growing faster than developed ones (with the notable exception of Ukraine). However, dynamics seem pretty similar: a strong rebound, followed by some slowdown. Argentina, in particular, has suffered a...
