An Empirical Comparison of Benchmarking Methods for Economic Stock Time Series*Irene Brown, U.S. Census Bureau
Keywords: benchmarking, Time Series, Stock, Interpolation, Splines
Economic Programs at the U.S. Census Bureau use the Causey-Trager method for benchmarking monthly and quarterly series to annual series and to the Economic Census every five years. This procedure uses an iterative, nonlinear constrained optimization technique to obtain the benchmarked series. Although constrained optimization works well for flow variables, Statistics Canada suggests using a different method for stock variables based on inconsistencies they found in their final benchmarked series. They use a simple method of interpolating the ratios of benchmark value to the original series value with a natural cubic spline and then apply the interpolated ratios to the original series to obtain the benchmarked series. Using historical economic time series from the Census Bureau, this paper investigates the effects of the two different methods in terms of level estimates of the benchmarked series, ratios of benchmarked series to original series, and preservation of the original series’ period-to-period changes. Lastly, we examine stock to flow ratios (e.g. inventory to sales).