Reliable Output Gap Estimates for Emerging Asian Economies

Abstract

Estimating output gaps is challenging for emerging Asian economies due to limited data availability and the potential effects of outliers. I apply the Beveridge-Nelson (BN) filter, comparing it with commonly used filters—Hodrick-Prescott (HP), Christiano-Fitzgerald (CF), and Hamilton—across four emerging Asian economies and argue that it provides more reliable and informative estimates than alternative methods. When benchmarked against narrower indicators of slack, BN filter output gap estimates provide a more informative indicator than capacity utilization and unemployment, given longer data coverage and controlling for long-run structural changes. I also document two systematic results for these economies. First, cyclical consumption is more volatile than the output gap. Second, decomposing GDP growth volatility shows that less than one-third of growth fluctuations is accounted for by movements in trend growth, with most variation attributed to the cyclical component. Taken together, these findings contrast with the interpretation in Aguiar and Gopinath (2007) that shocks to the trend are the primary driver of fluctuations in emerging economies and departs from their view that the “cycle is the trend.” Crucially, the BN filter estimates are also subject to smaller and less frequent revisions when faced with large changes in economic conditions, which benefits real-time policy decision-making.