Improving inflation forecasts using robust measures

Theory and extant empirical evidence suggest that the cross-sectional asymmetry across disaggregated price indexes might be useful in forecasting aggregate inflation. Trimmed-mean inflation estimators have been shown to devices for headline PCE But is this because they signal underlying trend or implicitly distribution? We address question by augmenting a "hard" beat benchmark model with robust trimmed-mean measures of skewness, both computed using 180+ components index. Our results indicate significant gains point density accuracy forecasts over medium- longer-term horizons, up through including COVID-19 pandemic. Improvements stem mainly from information implicit estimators, but skewness also useful. An examination goods services provides similar inference.