Simon Beyeler and Sylvia Kaufmann, "Reduced-form Factor Augmented VAR
-- Exploiting Sparsity to Include Meaningful Factors," Journal of
Applied Econometrics, Vol. 36, No. 7, 2021, pp. 989-1012.

The data used in this paper are from the FRED-QD database available
for download from the website of the Federal Reserve Bank of St. Louis
(https://research.stlouisfed.org/econ/mccracken/fred-databases/). The 
data set is a quarterly companion to the Monthly Database for
Macroeconomic Research (FRED-MD) assembled by McCracken and NG. It
consists of 248 macroeconomic time series for the U.S. economy which
are regularly updated and reported at a quarterly frequency starting
in 1959Q1. For our analysis, we use the vintage 2020-06. In addition,
we include utilization adjusted total factor productivity (TFP) from
Fernald (2012) available online

(https://www.frbsf.org/economic-research/indicators-data/total-factor-productivity-tfp/)In

In our analysis, we focus on the period 1965Q1 - 2019Q4 and drop the
series with missing observations. Where appropriate, we transform
series by applying first differences to either logs or levels. The
complete data set Is stored in bk_data.csv and a complete list of
included series is available in Web Appendix W6.

The second row in bk_data.csv contains the applied transformation
(t_code), where lv=level, fl = first difference in logs and fd = first
difference.

The MATLAB script bk_read_data.m can be used to read and transform the
series from bk_data.csv.

The files bk_data.csv and bk_read_data.m are zipped in the file
bk-files.zip. They are ASCII files in DOS format.

Please address any questions to:

Simon Beyeler
simon.beyeler [AT] snb.ch

Or

Sylvia Kaufmann
silvia.kaufmann [AT] szgerzensee.ch