Husted-Rogers-Sun MPU Index for the United States

Download Data

Husted, Rogers, and Sun construct a monthly index of Monetary Policy Uncertainty (MPU) by searching for keywords related to monetary policy uncertainty in the New York Times, Wall Street Journal and Washington Post. In particular, they count articles containing the triple of (i) "uncertainty" or "uncertain," (ii) "monetary policy(ies)" or "interest rate(s)" or "Federal fund(s) rate" or "Fed fund(s) rate," and (iii) "Federal Reserve" or "the Fed" or "Federal Open Market Committee" or "FOMC". They divide the daily count of such articles by the total number of news articles mentioning the Federal Reserve (or more precisely, any of the words in (iii) above) for each newspaper in a given period. Next, they standardize the resulting scaled counts for each newspaper to have a unit standard deviation from 1985 to 2015. To aggregate over newspapers, they sum the resulting series across newspapers by month. Finally, they normalize the series to have a mean of 100 from 1985 to 2015.

Husted, Rogers and Sun perform an extensive audit of their search results, constructing a "human index" based on human readings of 8,000 randomly selected newspaper articles. The human index is highly correlated with the baseline, computer-automated index. They also examine the sensitivity of their baseline, computer-based index by considering several adjustments to its construction. In the most noteworthy refinement, they narrow their search to articles in which the word uncertainty/uncertainties is in close proximity to Federal Reserve or monetary policy. Specifically, they restrict "uncertainty" or "uncertainties" to be within either 5, 10, or 20 words of the phrase "Federal Reserve" or "The Fed" or "monetary policy."

The HRS MPU index lines up very closely with a measure of uncertainty about monetary policy actions based on the Primary dealers' survey conducted in 2005 through 2013 by the New York Fed. Unlike market-based measures, such as swaptions, their MPU index has sizable movements during the zero-lower-bound period, capturing episodes such as the Taper Tantrum and liftoff uncertainty.

The HRS MPU index differs from the Baker, Bloom, and Davis (BBD) MPU indices along three dimensions. First, one version of the BBD MPU index uses the Access World News database of over 2,000 U.S. newspapers, and a second version uses a balanced panel of 10 major national and regional U.S. papers. In contrast, HRS focus on three leading newspapers that cover national economic and financial news. Second, the HRS keyword search features a focus on monetary policy in the U.S., while BBD include a broader set of terms designed to capture domestic and foreign sources of monetary policy uncertainty. Third, BBD scale by the total number of articles, while HRS scale by the number mentioning "Federal Reserve" (broadly). HRS conclude that the factors explaining the imperfect correlation between the HRS MPU Index and the BBD MPU Index based on Access World News can be ranked: (1) newspapers, (2) keywords, and (3) scaling.

See Husted, Lucas, John H. Rogers, and Bo Sun, "Monetary Policy Uncertainty," working paper, Board of Governors of the Federal Reserve Board, 2017 for additional information. This paper should be cited by persons who use the Husted-Roger-Sun MPU Index.