Monitoring supply and demand in Asia: an industry-level approach
Electronic Access:
Free download. Use the free Adobe Acrobat reader to view this PDF file
Disclaimer: IMF working papers describe the research being conducted by the authors and are published to provoke comment and encourage debate. The views expressed in IMF working papers are those of the authors and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.
Summary:
This paper provides a decomposition of GDP and its deflator into demand- and supply-driven components for 12 Asian countries, the United States, and Europe, following the forecast error-based methodology of Shapiro (2022). We extend that methodology by (1) considering a wide range of statistical forecasting models, using the optimal model for each country, and (2) providing a measure of idiosyncratic movements in supply and demand. The latter draws, for example, a distinction between inflation driven by aggregate demand and inflation driven by large shocks in only a small number of sectors. We find that lockdowns in 2020 are explained by a combination of demand and supply shocks in Asia, but that idiosyncratic demand shocks played an important role in some countries. Supply factors played an important role in the post-COVID recovery, mainly in 2021, and demand factors became more important in 2022. The mix of shocks during the sharp rise in inflation in 2021-22 differs depending on the country, with large and advanced economies generally experiencing more supply shocks (China, Australia, Korea), while emerging markets experienced significant demand pressures that drove up prices (Indonesia, Malaysia, Philippines, Vietnam, Thailand). We illustrate the usefulness of industry-level shocks in two applications. First, we consider whether industry supply shocks have created demand-like movements in aggregate prices and quantities, so-called Keynesian supply shocks. We find evidence of this mechanism in a minority of countries in our sample in Asia, as well as in Europe and the US, but these results are driven by the COVID-19 event. Second, we use the granularity of industry shocks to construct country-level GDP shocks, driven by idiosyncratic industry-level movements, to study cross-country growth spillovers for the three large economic units. of our sample: China, Europe and the United States.
