Latest update:

The nowcast for the U.S. output gap is based on the method described in Berger, Morley, and Wong (forthcoming, Journal of Econometrics). Please cite the paper when referring to the nowcast estimate.

The nowcast is reported in the figure, using an orange line extending from the realized estimated output gap for previous quarters. Conditional forecasts can be toggled on or off by clicking on the label in the legend and are displayed as a green line. Exact values can be seen by hovering over each point in the figure. For more information on the monthly indicators used in the nowcasts please click the 'Monthly Indicators' dropdown. The realized estimated output gap is based on parameter estimates for the mixed-frequency Bayesian VAR specified in Berger, Morley, and Wong and data up to the latest quarter for which real GDP has been released.


The table contains the latest values of monthly indicators that feed into the nowcast estimates.




Below are the three data files that can be used to run the MATLAB code linked to below. These files are updated daily. These data are mostly gathered from FRED, except for some indicators reported on FRED with a lag, which are scraped from elsewhere.

Contents by column (based on FRED codes - starting in 1967):
Monthly: FEDFUNDS, UMCSENT, UNRATE, CPIAUCSL, INDPRO, HOUST, SP500
Monthly spread: BAA, AAA, DGS10, DGS1
Quarterly: GDPC1


Reference:

Tino Berger, James Morley and Benjamin Wong, forthcoming, "Nowcasting the Output Gap", Journal of Econometrics.

Link to journal version

Link to working paper

Link to code