You can get the paper here.

Email ed.herbst@gmail.com with comments/questions.

Data used in the paper:

• data/CHdata.txt contains the macroeconomic series used to estimate the proxy SVARs and the local projections.

  • The federal funds rate is the average effective rate over the last week of the month
  • The unemployment rate, and the producer price index for final goods are taken from Coibion (2012) "Are the effects of monetary policy shocks big or small?"
  • We use manufacturing industrial production index (NAICS);
  • The BAA spread is Moody's Seasoned Baa Corporate Bond Yield Relative to Yield on 10-Year Treasury Constant Maturity, downloaded from the St. Louis Fed FRED database.
  • The Arouba, Diebold, Scotti business condition index is downloaded from the Federal Reserve Bank of Philadelphia website.
  • We use real personal consumption expenditures in nondurable goods, deflated using its own price deflator (downloaded from FRED);
  • We use the value-weighted total stock market index; Core loans are the sum of loans to households and businesses.
  • Business loans include commercial and industrial (C&I) loans and business loans secured by commercial real estate; household loans include residential mortgages, credit card loans, and other consumer loans. All series were obtained from the Federal Reserve's H.8 Statistical Release.
  • mhf is the series of monetary policy surprises constructed using high frequency data
  • mrr and mrrcs are the series of Romer and Romer shocks constructed following Equation (16);
  • mcgcs is the series of monetary policy shocks constructed following Equation (18)

• data/RR-CG-data.dta contains the data used to estimate the Romer and Romer (2004) and the Coibion Gorodnichenko regressions (equations (16) and (18) in the paper). The naming of the variables is consistent with these two papers.

For most models, you can either use Matlab (Gibbs Sampler) or Python/Fortran (Sequential Monte Carlo) to estimate the models. See the anaconda environment file (proxy-svar.yaml) for the python packages necessary.

1. Figure 1: Impulse Reponse to a Monteray Policy Shock

   cd python
   python compare_irfs.py --model 4eq 5eq 

2. Figure 2: Contribution of Monetary Policy Shocks to FEVD

   cd python
   python compare_fevds.py --model 4eq 5eq

3. Table 1: Coefficients in the Monetary Policy Equation

   cd python
   python compare_elasticities.py --model 4eq 5eq

4. Figure 3: Impulse Reponse to a Monetary Policy Shock

   Python code:

   cd python
   python compare_irfs.py --model 5eq_cholesky --overlay 5eq

5. Figure 4: Macroeconomic Implications of Financial Shocks

   Python code:

   cd python
   python compare_irfs.py --model 5eq_cholesky --overlay 5eq_fin

6. Table 2: Coefficients in the Monetary Policy Equation

   cd python
   python compare_elasticities.py --model 5eq 5eq_tight

7. Figure 5: Impulse Responses to a Monetary Policy Shock

   cd python
   python compare_irfs.py --model 5eq_tight

8. Table 3: Determinants of the Change in the Intended Fed Funds Rate

   cd stata
   stata -b do RR-CG-reg.do

9. Figure 6: Impulse Response to a Monetary Policy Shock

   python compare_irfs.py --model 4eq_cholesky_RRCS 5eq_cholesky_RRCS --overlay 4eq_cholesky_RR 5eq_cholesky_RR 

10. Table 4: Local Projections

    cd python
    python local_projections.py

11. Figure 7: Impulse Responses to a Monetary Policy Shock

    The figure is created automatically after the matlab estimation.

12. Table 5: Determinants of the Federal Funds Rate

    cd stata
    stata -b do RR-CG-reg.do