Hongwei Zhang, Qiang He, Ben Jacobsen, Fuwei Jiang, "Forecasting Stock
Returns with Model Uncertainty and Parameter Instability", Journal of
Applied Econometrics, Vol. 35, No. 5, 2020, pp. 629-644.

All the data used in this paper are publicly available at the FRED
website, Amit Goyal's website, and Michael W. McCracken's website. 

All the data files are ASCII files in CSV (DOSD) format. They are
zipped in the file zhjj-files.zip. Unix/Linux users should use "unzip -a".

The "MacroData" folder contains seven CSV files that represent the
monthly U.S. macroeconomic conditions that are used in Table 6 of our
paper. The seven macroeconomic condition variables are the Chicago Fed
National Activity Index (CFNAI), Smoothed U.S. Recession Probabilities
(SRP), Industrial Production Growth (IPG), Jurado et al. (2015)'s
Macroeconomic Uncertainty Index (MU), Output Gap (Gap), Lettau and
Ludvigson (2001)'s Cay (Cay), and Civilian Unemployment Rate (UNRATE).
Detailed variable definitions and data source descriptions are provied
in Section 5 (Forecasting Macroeconomic Conditions) of our paper.

"GoyalData_NBER.csv" contains the monthly Welch-Goyal dataset to
construct 12 economic return predictors used in our paper. The sample
period is 1926:12 to 2016:12. This dataset is publicly available from
Amit Goyal's website, http://www.hec.unil.ch/agoyal/. Detailed
variable definitions and data source descriptions are given in Welch
and Goyal (2008). 

"FRED_MD_Factors.csv" contains the first seven monthly PCA macro
factors that are used in Table 7 of our paper. The PCA factors are
extracted from FRED-MD dataset. The sample period is 1959:03 to
2016:12. FRED-MD dataset is publicly available at Michael W.
McCracken's website,
https://research.stlouisfed.org/econ/mccracken/sel/. Detailed variable
definitions and data sources descriptions are given in Mccracken and
Ng (2015). The MATLAB code for processing the FRED-MD dataset and for
estimating the PCA factors is also publicly available at Michael W.
McCracken's website.