These codes reproduce the results in:
Cúrdia, V., and M. Woodford (2011)
The Central-Bank Balance Sheet as an Instrument of Monetary Policy
Journal of Monetary Economics, 58(1), pp. 54-79.
These replication codes are available online at:
https://github.com/vcurdia/CW2011-ReplicationCodes
The codes were tested using Matlab (R) R2016b with the following toolboxes
- Symbolic Toolbox
- Optimization Toolbox
LaTeX is used by some tools to compile certain documents.
epstopdf, included in most LaTeX releases, is used by some tools.
Codes by Vasco Cúrdia:
Codes by Chris Sims:
All auxiliary codes included in this repository in subfolders.
These codes produce the figures comparing the response of the Central Bank's incentive to engage in credit policy displayed in figures 4 and 5 of the paper.
Exercise2.m
Generates results under the assumption that interest rate is optimal.
Exercise3.m
Generates results under the assumption that interest rate follows a simple
Taylor rule.
Both Exercise2 and Exercise3 can be set to ignore or not the zero lower
bound (ZLB); consider different sizes of the shock to the spread; and to
consider different persistency of shocks.
IRFPlotCompareExercise2.m
Plots figure 4, based on results generated by Exercise2.
IRFPlotCompareExercise3.m
Plots figure 5, based on results generated by Exercise2 and Exercise3, with
and without ignoring the ZLB.
Exercise4.m
Exercise4SearchSequence.m
Generate solution and IRF with and without credit policy, under optimal
interest rate policy. First need to run Exercise4 to generate base results
under a given sequence of regimes (with and without credit policy and with and
without ZLB binding). Then use Exercise4SearchSequence to more quickly check
alternative sequences of regimes. The search is done by checking if the ZLB is
violated, if credit policy is negative, or if the lagrange multipliers for the
slack conditions are negative. Each time that the SearchSequence code is run
tables with checks and a plot with evolution of some of these are shown to
inform on how to change the candidate sequence of regimes.
IRFPlotCompareExercise4.m
Plot the IRFs with and without credit policy. There are several options
available.
Exercise5.m
Exercise5SearchSequence.m
Generate solution and IRF with and without credit policy, when interest rate
follows a simple Taylor rule. First need to run Exercise5 to generate base
results under a given sequence of regimes (with and without credit policy and
with and without ZLB binding). Then use Exercise5SearchSequence to more
quickly check alternative sequences of regimes. The search is done by checking
if the ZLB is violated, if credit policy is negative, or if the lagrange
multipliers for the slack conditions are negative. Each time that the
SearchSequence code is run tables with checks and a plot with evolution of
some of these are shown to inform on how to change the candidate sequence of
regimes.
IRFPlotCompareExercise5.m
Finally, IRFPlotCompareExercise5 is used to plot the IRFs with and without
credit policy.
Exercise6.m
Exercise6SearchSequence.m
Generate solution and IRF under simple credit policy rule, when interest rate
follows a simple Taylor rule. First need to run Exercise6 to generate base
results under a given sequence of regimes (with and without ZLB binding). Then
use Exercise6SearchSequence to more quickly check alternative sequences of
regimes. The search is done by checking if the ZLB is violated. Each time that
the SearchSequence code is run tables with checks and a plot with evolution
of some of these are shown to inform on how to change the candidate sequence of
regimes.
IRFPlotCompareExercise6.m
Plot the IRFs under alternative credit policy rule coefficients. These figures
are not shown in the JME version of the paper but are shown in
the NBER working paper version.
BindBp controls how much higher CB intervention cost is relative to
the steady state critical level for no credit policy lending in steady
state, measured in basis points, where a value of zero implies that it is at
the critical level and a value of 10 is 10bp higher, as shown in figure 8,
for example.
dSP controls the size of the shock to the spread, with 4 representing 4% in
annualized terms.
PersSP controls the persistency of the shock, with 90 corresponding to an
AR(1) coefficient of 0.90.
Because there are many options and switches available, we also include scripts that generate the individual figures of the JME and NBER papers. Each of these runs all necessary pre-requirements and tweaks needed for each figure.
MakeFig4.m
MakeFig5.m
MakeFig6.m
MakeFig7.m
MakeFig8.m
MakeFig9.m
MakeFig10.m
MakeFig11.m
MakeFig12.m
MakeFigNBER5.m
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent