Instructions for reproducing figures and tables for ``Reputation and Liquidity Traps.'' Taisuke Nakata (taisuke.nakata@frb.gov or taisuke.nakata@gmail.com) September 5th, 2017 =========================== =========================== =========================== Overview =========================== =========================== In this file, I describe how to generate figures and tables in the paper and the online appendix. In general, to replicate each figure in the main text, find its folder with its respective figure number or table number. Also, one should have a license to use IMSL Fortran library to run Fortran files properly. Fortran codes were complied using Intel Fortran Complier in Linux. Matlab codes can be run in both Windows and Linux. Fortran codes were complied using Intel Fortran Complier in Linux. Matlab codes were also run on Linux. ***** ***** ***** Basic Structure of the "Codes" folder: ***** ***** Each folder is named after its respective figure(s). In folders for figures, Fortran files yield .dat files which are used as inputs for MATLAB files to create figures. In folders for tables, Fortran files yield numerical results that are reported in tables. Detailed step-by-step instructions for creating each figure and table are found below. ***** ***** ***** Expected computation time. ***** ***** All Fortran files should not take any longer than 5 minutes, and most of them will take less than a minute to complete their runs. All MATLAB files will create figures in less than 30 seconds. =========================== =========================== =========================== =========================== Main body of the paper =========================== =========================== =========================== =========================== Figure 1: The Discretionary Outcome and Value Sequence Figure 2: The Ramsey Outcome and Value Sequence =========================== 1. Run Fig_Intro.f90. 2. Run Fig_Intro.m file, which will create Fig1.eps (figure 1) and Fig2.eps (figure 2) by reading in the following .dat files, created from running the Fortran code: IRF_Intro_hypo_ipH0.dat, IRF_Intro_hypo_ipH0.dat, IRF_Intro_hypo_ipH0.dat, IRF_Intro_mp_ipH0.dat, IRF_Intro_mp_ipH1.dat, IRF_Intro_mp_ipH2.dat, IRF_Intro_ramsey_ipH0.dat, IRF_Intro_ramsey_ipH1.dat, IRF_Intro_ramsey_ipH2.dat. =========================== Figure 3: Credibility of the Revert-to-Discretion Plan =========================== 1. Run Credible_Region.f90. 2. Run Fig_CredibleRegion.m file, which will create Fig3.eps by reading in the following .dat file, created from running the Fortran code: iCredible.dat. =========================== Figure 4: The Discretionary and Ramsey Outcomes/Value sequences: Frequent vs. Infrequent Shocks =========================== 1. Run Fig_IRF_pH.f90. 2. Run Fig_IRF_pH.m file, which will create Fig4.eps by reading in the following .dat files, created from running the Fortran code: IRF_mp_ipH1.dat, IRF_mp_ipH2.dat, IRF_ramsey_ipH1.dat, IRF_ramsey_ipH1.dat. =========================== Figure 5: The Discretionary and Ramsey Outcomes/Value sequences: Transient vs. Persistent Shocks =========================== 1. Run Fig_IRF_pL.f90. 2 Run Fig_IRF_pL.m file, which will create Fig5.eps by reading in the following .dat files, created from running the Fortran code: IRF_mp_ipL1.dat, IRF_mp_ipL2.dat, IRF_ramsey_ipL1.dat, IRF_ramsey_ipL1.dat. =========================== Figure 6: Credibility of the Revert-to-Discretion(N) Plans (i.e., Plans with Finite-periods Punishment) =========================== 1. Run Credible_Region_FinitePunishment_max_pH_loop_N.f90. 2. Run Credible_FinitePunishment_max_pH_loop_N.m file, which will create Fig6.eps by reading in the following .dat file, created from running the Fortran code: iCredible_FinitePunishment_max_pH_loop_N.dat. =========================== Figure 7: Sensitivity Analysis =========================== 1. Run the following Fortran codes: Credible_Region_FinitePunishment_max_pH_loop_N.f90, Credible_Region_FinitePunishment_max_pH_loop_N_chiC.f90, Credible_Region_FinitePunishment_max_pH_loop_N_CostPush.f90, Credible_Region_FinitePunishment_max_pH_loop_N_kappa.f90, Credible_Region_FinitePunishment_max_pH_loop_N_lambda.f90. 2. Run SA.m file, which will create Fig7.eps by reading in the following .dat files, created from running the Fortran code: iCredible_FinitePunishment_max_pH_loop_N.dat, iCredible_FinitePunishment_max_pH_loop_N_chiC.dat, iCredible_FinitePunishment_max_pH_loop_N_CostPush.dat, iCredible_FinitePunishment_max_pH_loop_N_kappa.dat, iCredible_FinitePunishment_max_pH_loop_N_lambda.dat. =========================== Figure 8: GR and GD Calibrations of Denes, Eggertsson, and Gilbukh (2013) =========================== 1. Run the following Fortran codes: Credible_Region_FinitePunishment_max_pH_loop_N_GD.f90, Credible_Region_FinitePunishment_max_pH_loop_N_GR.f90. 2. Run SA_GR_GD.m file, which will create Fig8.eps by reading in the following .dat files, created from running the Fortran code: iCredible_FinitePunishment_max_pH_loop_N_GD.dat, iCredible_FinitePunishment_max_pH_loop_N_GR.dat. =========================== Figure 9: Nonlinear Models =========================== 1. Run the following Fortran codes: Credible_Region_FinitePunishment_max_pH_loop_N.f90, Credible_Region_NonLinear_FinitePunishment_max_pH_loop_N.f90, Credible_Region_NonLinear_FinitePunishment_max_pH_loop_N_BKW2013.f90, Credible_Region_NonLinear_FinitePunishment_max_pH_loop_N_CE2012.f90. 2. Run SA_NonLinear.m file, which will create Fig9.eps by reading in the following .dat files, created from running the Fortran code: iCredible_FinitePunishment_max_pH_loop_N.dat, iCredible_NonLinear_FinitePunishment_max_pH_loop_N.dat, iCredible_NonLinear_FinitePunishment_max_pH_loop_N_BKW2013.dat, iCredible_NonLinear_FinitePunishment_max_pH_loop_N_CE2012.dat. =========================== =========================== =========================== =========================== Online Appendix =========================== =========================== =========================== =========================== Figure D.1: Existence of Four Markov-Perfect Equilibria =========================== 1. Run Existence_Region.f90. 2. Run existence_plots.m file, which will create FigD1.eps by reading in the following .dat files, created from running the Fortran code: iExistence_case1_npH051_npL051.dat, iExistence_case2_npH051_npL051.dat, iExistence_case3_npH051_npL051.dat, iExistence_case4_npH051_npL051.dat. =========================== Figure D.2: Allocations in Type-I Markov-Perfect Equilibrium, Figure D.3: Allocations in Type-II Markov-Perfect Equilibrium =========================== 1. Run analytic_analysis.m file, which will create FigD2.eps and FigD3.eps. =========================== Figure E.1: Repeated Crises: Scenario 1, Figure E.2: Repeated Crises: Scenario 2, Figure E.3: Repeated Crises: Scenario 3 =========================== 1. Run RecurringCrises.f90. 2. Run RecurringCrises.m file, which will create FigE1.eps, FigE2.eps, and FigE3.eps by reading in the following .dat files, created from running the Fortran code: IRF_Ramsey_A1_v1.dat, IRF_Ramsey_A1_v2.dat, IRF_Ramsey_A2_v1.dat, IRF_Ramsey_A2_v2.dat, IRF_Ramsey_A3_v1.dat. =========================== Figure F.1: Credibility of the Revert-to-Deflation Plan =========================== 1. Run Credible_Region_DeflationPunishment_pcy.f90. 2. Run deflation_punishment_credibility_dotplot_pcy.m file, which will create FigF1.eps by reading in the following .dat file, created from running the Fortran code: iCredible_DeflationPunishment_pcy.dat. =========================== Figure I.1: The Short-run Gain and the Long-run Loss of Deviating from the Ramsey Policy (with Alternative Shock Frequencies) =========================== 1. Run Fig_Mechanism_pH.f90. 2. Run Fig_Mechanism_pH.m file, which will create FigI1.eps by reading in the following .dat file, created from running the Fortran code: Mechanism_pH.dat. =========================== Figure I.2: The Short-run Gain and the Long-run Loss of Deviating from the Ramsey Policy (with Alternative Shock Frequencies) =========================== 1. Run Fig_Mechanism_pL.f90. 2. Run Fig_Mechanism_pL.m file, which will create FigI1.eps by reading in the following .dat file, created from running the Fortran code: Mechanism_pL.dat. =========================== Figure J.1: Credibility of the Revert-to-Discretion Plan: Sensitivity Analysis =========================== 1. Run the following Fortran codes: Credible_Region_bet_maxpH_i.f90, Credible_Region_chiC_maxpH_i.f90, Credible_Region_kapp_maxpH_i.f90, Credible_Region_L_maxpH_i.f90, Credible_Region_lambd_maxpH_i.f90. (1 <= i <= 4) 2. Run FigJ1.m file, which will create FigJ1.eps by reading in the following .dat file, created from running the Fortran code: iCredible_maxpH_ibet010_i.dat, iCredible_maxpH_ibet020_i.dat, iCredible_maxpH_ichiC010_i.dat, iCredible_maxpH_ichiC020_i.dat, iCredible_maxpH_ikapp010_i.dat, iCredible_maxpH_ikapp020_i.dat, iCredible_maxpH_iL010_i.dat, iCredible_maxpH_iL020_i.dat, iCredible_maxpH_ilambd010_i.dat, iCredible_maxpH_ilambd020_i.dat. (1 <= i <= 4) =========================== Figure J.2: The Discretionary/Ramsey Outcomes and Values with Alternative Sizes of the Shock =========================== 1. Run Fig_IRF_L.f90. 2. Run Fig_IRF_L.m file, which will create FigJ2.eps by reading in the following .dat file, created from running the Fortran code: IRF_mp_iL1.dat, IRF_mp_iL2.dat, IRF_ramsey_iL1.dat, IRF_ramsey_iL2.dat. =========================== Figure J.3: The Discretionary/Ramsey Outcomes and Values with Alternative Discount Rates =========================== 1. Run Fig_IRF_beta.f90. 2. Run Fig_IRF_beta.m file, which will create FigJ3.eps by reading in the following .dat file, created from running the Fortran code: IRF_mp_ibet1.dat, IRF_mp_ibet2.dat, IRF_ramsey_ibet1.dat, IRF_ramsey_ibet2.dat. =========================== Figure J.4: The Discretionary/Ramsey Outcomes and Values with Alternative Slopes of the Phillips Curve =========================== 1. Run Fig_IRF_kappa.f90. 2. Run Fig_IRF_kappa.m file, which will create FigJ4.eps by reading in the following .dat file, created from running the Fortran code: IRF_mp_ikapp1.dat, IRF_mp_ikapp2.dat, IRF_ramsey_ikapp1.dat, IRF_ramsey_ikapp2.dat. =========================== Figure J.5: The Discretionary/Ramsey Outcomes and Values with Alternative Risk Aversion =========================== 1. Run Fig_IRF_chiC.f90. 2. Run Fig_IRF_chiC.m file, which will create FigJ5.eps by reading in the following .dat file, created from running the Fortran code: IRF_mp_ichiC1.dat, IRF_mp_ichiC2.dat, IRF_ramsey_ichiC1.dat, IRF_ramsey_ichiC2.dat. =========================== Figure J.6: The Discretionary/Ramsey Outcomes and Values with Alternative Weights on Consumption Stabilization =========================== 1. Run Fig_IRF_lambda.f90. 2. Run Fig_IRF_lambda.m file, which will create FigJ6.eps by reading in the following .dat file, created from running the Fortran code: IRF_mp_ilambd1.dat, IRF_mp_ilambd2.dat, IRF_ramsey_ilambd1.dat, IRF_ramsey_ilambd2.dat. =========================== Figure K.1: Credibility of the Revert-to-Discretion Plan: Sensitivity Analysis in the Structural-Parameter Space =========================== 1. Run the following Fortran codes: Credible_Region_alpha_maxpH_i.f90, Credible_Region_chiC_maxpH_i.f90, Credible_Region_chiN_maxpH_i.f90, Credible_Region_theta_maxpH_i.f90. (1 <= i <= 4) 2. Run FigK1.m file, which will create FigK1.eps by reading in the following .dat file, created from running the Fortran code: iCredible_maxpH_ialph010_i.dat, iCredible_maxpH_ialph020_i.dat, iCredible_maxpH_ichiC010_i.dat, iCredible_maxpH_ichiC020_i.dat, iCredible_maxpH_ichin010_i.dat, iCredible_maxpH_ichin020_i.dat, iCredible_maxpH_itheta010_i.dat, iCredible_maxpH_itheta020_i.dat. (1 <= i <= 4) =========================== Figure K.2: Sensitivity Analysis for $\kappa$ in the Reduced-Form Parameter Space (left) versus Structural Parameter Space (right) =========================== 1. Run the following Fortran codes: Credible_Region_FinitePunishment_maxpH_loop_N_kappa.f90, Credible_Region_FinitePunishment_maxpH_loop_N_kappa_alt.f90. 2. Run SA_kappa_altkappa.m file, which will create FigK2.eps by reading in the following .dat file, created from running the Fortran code: iCredible_FinitePunishment_maxpH_loop_N_kappa.dat, iCredible_FinitePunishment_maxpH_loop_N_kappa_alt.dat. =========================== Figure K.3: Sensitivity Analysis for $\chi_{c}$ in the Reduced-Form Parameter Space (left) versus Structural Parameter Space =========================== 1. Run the following Fortran codes: Credible_Region_FinitePunishment_maxpH_loop_N_chiC.f90, Credible_Region_FinitePunishment_maxpH_loop_N_chiC_alt.f90. 2. Run SA_kappa_altchic.m file, which will create FigK2.eps by reading in the following .dat file, created from running the Fortran code: iCredible_FinitePunishment_maxpH_loop_N_chiC.dat, iCredible_FinitePunishment_maxpH_loop_N_chiC_alt.dat. =========================== Figure L.1: Threshold Crisis Frequencies with Various ``Modified'' GR Calibrations =========================== 1. Run the following Fortran codes: Credible_Region_FinitePunishment_maxpH_loop_N_GR.f90, Credible_Region_FinitePunishment_maxpH_loop_N_GR_chiC.f90, Credible_Region_FinitePunishment_maxpH_loop_N_GR_kappa.f90, Credible_Region_FinitePunishment_maxpH_loop_N_GR_L.f90, Credible_Region_FinitePunishment_maxpH_loop_N_GR_lambda.f90, Credible_Region_FinitePunishment_maxpH_loop_N_GR_pL.f90. 2. Run SA_GR_GD.m file, which will create FigL.eps by reading in the following .dat file, created from running the Fortran code: iCredible_FinitePunishment_maxpH_loop_N_GR.dat, iCredible_FinitePunishment_maxpH_loop_N_GR_chiC.dat, iCredible_FinitePunishment_maxpH_loop_N_GR_kappa.dat, iCredible_FinitePunishment_maxpH_loop_N_GR_L.dat, iCredible_FinitePunishment_maxpH_loop_N_GR_lambda.dat, iCredible_FinitePunishment_maxpH_loop_N_GR_pL.dat. =========================== Figure M.1: IRFs under different shock realizations =========================== 1. Run Fig_Mechanism_pH_loop.f90. 2. Run Fig_Mechanism_pH_loop.m file, which will create Fig_M1.eps by reading in the following .dat file, created from running the Fortran code: Mechanism_pH_pcy_high_pH.dat, Mechanism_pH_pcy_cons_inf_high_pH.dat. =========================== Figure M.2: The incentive to deviate is the largest in the aftermath of a crisis lasting for a few years =========================== 1. Run the following Fortran codes: Fig_Mechanism_pH_loop.f90, Fig_Mechanism_pH_loop_lowpH.f90, Fig_Mechanism_pH_loop_midpH.f90. 2. Run Fig_low_pH_high_pH.m file, which will create Fig_M2.eps by reading in the following .dat file, created from running the Fortran code: Mechanism_pH_pcy_high_pH.dat, Mechanism_pH_pcy_cons_inf_high_pH.dat, Mechanism_pH_pcy_mid_pH.dat, Mechanism_pH_pcy_cons_inf_mid_pH.dat, Mechanism_pH_pcy_low_pH.dat, Mechanism_pH_pcy_cons_inf_low_pH.dat.