README FILE FOR THE REPLICATION CODES 

JANUARY 2021

TAMON TAKAMURA

1) MATLAB codes for replicating the figures in the main text and the online appendix. 

a. plot_fig1.m: This file loads the data file (VAR_data.mat) and runs VAR to create the empirical impulse responses in Figure 1. A subroutine requires MATLAB's statistics toolbox. This file saves results in "IRFV_empirical.mat", which is used for estimation. 

b. plot_fig2_3_4.m: This file creates Figures 2 to 4 in the main text. One needs to specify a figure number (2, 3 or 4) as well as a panel number (1 for panel a and 2 for panel b in Figures 3 and 4) at the beginning of the file. The file calls LINLAGEX subroutine developed by Meyer-Gohde (2010) to solve the dual stickiness model in this paper. The enclosed "private" folder contains LINLAGEX. 

c. plot_fig5.m: This file creates Figure 5 in the main text. The file loads pre-computed estimated parameters and the empirical impulse responses from "ResultKnitro_mg.mat". Then, it computes the model-implied impulse responses by calling LINLAGEX to plot the figure.

d. plot_figA1.m and plot_figA2_A3.m: These files replicate the figures in the online appendix. 


2) Minimum-distance estimation codes

All files are contained in the folder named "Estimation".

a. mde_knitro_mg.m: This file implements the minimum-distance estimation to estimate the information stickiness parameters. To run this program, one needs to have Artelys KNITRO (https://www.artelys.com/solvers/knitro/). We use its optimization subroutine with a multi-start algorithm to search for an optimum. This file saves results in "ResultKnitro_mg.mat", which is used by "plot_fig5.m".  

b. mde_knitro_ir.m and mde_knitro_ir_noSI.m: These files estimate parameters in the model with the interest-rate rule discussed in the online appendix. These files save results in "ResultsKnitro_ir.mat" and "ResultsKnitro_ir_noSI.mat", which are used by "plot_figA2_A3.m".


3) Data

Data files are available in "VAR_data.csv" (ASCII format) and "VAR_data.mat" (MATLAB format).
