This paper considers the use of trimmed means as monthly indicators of Japanese core inflation. As in Bryan, Cecchetti, and Wiggins (1997) for the United States, and Roger (1997) for New Zealand, we find that trimming the tails of the price change distribution substantially improves high-frequency estimates of Japanese core inflation. These estimators yield efficiency gains of roughly two-thirds over the Japanese CPI. While we find that trimming approximately 35 percent from each tail of the price change distribution produces the most efficient monthly estimator over the full 27-year period, a range of trimmed-mean estimators (between 21 percent and the median price change) provide nearly the same signal. Moreover, we find that these estimators are superior to the standard monthly core inflation estimator in Japan, the CPI less fresh food. At lower frequencies (12-month percent changes and beyond), the differences between the candidate estimators were found to be small, and the trimmed estimators were nearly the same as the CPI less fresh food and energy along many dimensions.
Keywords: Inflation measurement; Core inflation; Trimmed means; Median
Views expressed in the paper are those of the authors and do not necessarily reflect those of the Bank of Japan or Institute for Monetary and Economic Studies.