Tony Chernis, Patrick Coe, and Shaun Vahey, "Reassessing the Dependence 
Between Economic Growth and Financial Conditions since 1973", Journal of 
Applied Econometrics, Vol. 38, No. 2, 2023, pp. 260-267.

All files may be found in the file cvv-files.zip. Text files are in
DOS format.

***** Data *****

Data were downloaded from FRED on Jan 14, 2022 and are included in the
folder Data_Files in 3 formats.

.xls are those downloaded from FRED and include headers etc.
.xlsx are the same files with headers and dates removed -- these are
      used by ABGRep_Data.m to create ABGRep.mat
.csv are comma-separated versions of the .xlsx files.
 
The three data files are:

A191RL1Q225SBEA.xlsx -- contains ABG's growth measure from
1973Q1--2021Q3 for a total of 195 observations.

GDPC1.xlsx -- contains the real GDP series from 1972Q1--2021Q3
(199 observations).  This series is used to calculate our 1-quarter
and 4-quarter growth measures from 1973Q1--2021Q3.

NFCI.xlsx -- contains the financial conditions index from
1973Q1--2021Q3 (195 observations).

See section A.1 of the Appendix for further details.


***** Matlab Files *****

These are included in the folder Matlab_Files.

*** ABGRep_Data.m

Loads .xlsx data files and writes to a single data file ABGRep.mat.  

ABGRep.mat is used by ABG_Figures to produce Figure 1 of the paper and
Figure A1 of the Appendix.

*** ABGRep_Stats.m

Produces output for Table A1 of the Appendix.

It calculates first 4 moments for 1-quarter growth, 4-quarter growth and
financial conditions. It also calculates the Shapiro-Wilks statistic.
Requires swtest.m (included).


*** ABGRep_Estimate.m

Produces output used by ABGRep_Figures.m to produce Figures 2-7 and
Figures A2, A13 and A18-A22.

It estimates ABG's QR on observed data and produces QR lines.  It also
estimates copula QR using pseudo data and produces QR lines.  It
produces non-linear copula QR lines using ECDF fitted marginal.

Before running, choose the growth measure by setting g_flag = 1 for
ABG's series (1dp), g_flag = 2 for the 1-qtr growth from GDPC1 (1qtr)
or g_flag = 3 for the 4-qtr growth from GDPC1 (4qtr). Also choose the
sample by setting s_flag = 0 for ABG's sample ending 2015Q4 (abg) or
s_flag = 1 for the extended sample ending 2021Q3 (ext). Output is
written to files EST_xxx_yyy.mat, where xxx indicates growth measure
used, 1dp, 1qtr or 4qtr and yyy indicates sample used, abg or ext.
Requires ssvkernel.m (included) and rq.m (from ABG's code at
https://www.aeaweb.org/articles?id=10.1257/aer.20161923).

*** ABGRep_HetTests.m

Produces output used by ABGRep_Tables to produce Tables A2 and A3 and
calculates Machado-Santos Silva and White statistics.

Before running, choose g_flag and s_flag. Output is written to files
HET_xxx_yyy.mat where xxx denotes growth measure and yyy denotes
sample.


*** ABGRep_BB.m

Produces output used by ABGRep_Tables to produce Tables A4 and A5.

It performs a block bootstrap to obtain confidence intervals for QR
coefficients. Before running choose g_flag and s_flag.  

Output is written to files BB_xxx_yyy.mat where xxx denotes growth
measure and yyy denotes sample. Requires rq.m.


*** ABGRep_Dummy.m and ABGRep_Dummy_BB

Produces output used by ABGRep_Tables to produce Tables A6 and A7.

The first file modifies ABGRep_Estimate.m to allow for a dummy for the
Covid pandemic observations. The second file modifies ABGRep_BB.m to
incorporate the dummy variable. Output is written to files
EST_Dummy.mat and BB_Dummy.mat. Requires the file rq.m


*** ABGRep_MCSim.m

Produces output used by ABGRep_Figures to produce Figures A3 and A4.

It performs the Monte Carlo experiment described in section A.2.4 of
the appendix. Output is written to file MCSim.mat. Requires the file
rq.m


*** ABGRep_Bands.m

This is a modified version of ABG's code VulnerabilityBands.m to allow
for pseudo data. 

It produces output used by ABGRep_Figures to produce Figures A5, A6,
A14 and A15. Generates confidence bands from VAR as in ABG.

Before running, choose s_flag and d_flag, where d_flag = 0 for
observed (standardized) data and 1 for pseudo data.  Output is written
to files ABGBands_yyy_zzz.mat where yyy refers to the sample and zzz
to either observed data (obs) or pseudo data (ps). Requires the file
QRBoot.m (from ABG's code) and uses the data file
DataVulnerability.mat which includes dates in the format expected by
ABG's code.


*** ABGRep_CopSim.m

Produces output used by ABGRep_Figures to produce Figures A16 and A17.

It simulates from the copula density to produce a swarm for growth and
financial conditions.  Output is written to files CopSim.mat. Requires
the file scatter_kde.m (included).


*** ABG's Code

Finally, we use ABG's code and our extended sample to produce Figures
A7 to A12 inclusive. 

ABG's code can be downloaded from
https://www.aeaweb.org/articles?id=10.1257/aer.20161923.

Output is written to files ResMatch_H1.mat and ResOOS_H1.mat -- these
files are included in this archive.