irf

Generate vector autoregression (VAR) model impulse responses

Syntax

Response = irf(Mdl)

Response = irf(Mdl,Name,Value)

[Response,Lower,Upper] = irf(___)

Description

The irf function returns the dynamic response, or the impulse response function (IRF), to a one-standard-deviation shock to each variable in a VAR(p) model. A fully specified varm model object characterizes the VAR model.

To estimate or plot the IRF of a dynamic linear model characterized by structural, autoregression, or moving average coefficient matrices, see armairf.

IRFs trace the effects of an innovation shock to one variable on the response of all variables in the system. In contrast, the forecast error variance decomposition (FEVD) provides information about the relative importance of each innovation in affecting all variables in the system. To estimate the FEVD of a VAR model characterized by a varm model object, see fevd.

Response = irf(Mdl) returns the 20-period, orthogonalized IRF of the response variables that compose the VAR(p) model Mdl, characterized by a fully specified varm model object. irf shocks variables at time 0, and returns the IRF for times 0 through 19.

Response = irf(Mdl,Name,Value) uses additional options specified by one or more name-value pair arguments. For example, 'NumObs',10,'Method',"generalized" specifies estimating a generalized IRF for 10 time points starting at time 0, during which irf applies the shock, and ending at period 9.

[Response,Lower,Upper] = irf(___) uses any of the input argument combinations in the previous syntaxes and returns lower and upper 95% confidence bounds for each period and variable in the IRF:

• If you specify series of residuals by using the E name-value pair argument, then irf estimates the confidence bounds by bootstrapping the specified residuals.

• Otherwise, irf estimates confidence bounds by conducting Monte Carlo simulation.

If Mdl is a custom varm model object (an object not returned by estimate or modified after estimation), irf might require a sample size for the simulation SampleSize or presample responses Y0.

Examples

Estimate and Plot VAR Model IRF

Fit a 4-D VAR(2) model to Danish money and income rate series. Then, estimate and plot the orthogonalized IRF from the estimated model.

Load the Danish money and income data set.

load Data_JDanish

The data set includes four time series in the table DataTable. For more details on the data set, enter Description at the command line.

Assuming that the series are stationary, create a varm model object that represents a 4-D VAR(2) model. Specify the variable names.

Mdl = varm(4,2);
Mdl.SeriesNames = DataTable.Properties.VariableNames;

Mdl is a varm model object specifying the structure of a 4-D VAR(2) model; it is a template for estimation.

Fit the VAR(2) model to the data set.

Mdl = estimate(Mdl,DataTable.Series);

Mdl is a fully specified varm model object representing an estimated 4-D VAR(2) model.

Estimate the orthogonalized IRF from the estimated VAR(2) model.

Response = irf(Mdl);

Response is a 20-by-4-by-4 array representing the IRF of Mdl. Rows correspond to consecutive time points from time 0 to 19, columns correspond to variables receiving a one-standard-deviation innovation shock at time 0, and pages correspond to responses of variables to the variable being shocked. Mdl.SeriesNames specifies the variable order.

Display the IRF of the bond rate (variable 3, IB) when the log of real income (variable 2, Y) is shocked at time 0.

Response(:,2,3)

ans = 20×1

    0.0018
    0.0048
    0.0054
    0.0051
    0.0040
    0.0029
    0.0019
    0.0011
    0.0006
    0.0003
      ⋮

Plot the IRFs of all series on separate plots by passing the estimated AR coefficient matrices and innovations covariance matrix of Mdl to armairf.

armairf(Mdl.AR,[],"InnovCov",Mdl.Covariance);

Each plot shows the four IRFs of a variable when all other variables are shocked at time 0. Mdl.SeriesNames specifies the variable order.

Estimate Generalized IRF of VAR Model

Consider the 4-D VAR(2) model in Estimate and Plot VAR Model IRF. Estimate the generalized IRF of the system for 50 periods.

Load the Danish money and income data set, then estimate the VAR(2) model.

load Data_JDanish

Mdl = varm(4,2);
Mdl.SeriesNames = DataTable.Properties.VariableNames;
Mdl = estimate(Mdl,DataTable.Series);

Estimate the generalized IRF from the estimated VAR(2) model.

Response = irf(Mdl,"Method","generalized","NumObs",50);

Response is a 50-by-4-by-4 array representing the generalized IRF of Mdl.

Plot the generalized IRF of the bond rate when real income is shocked at time 0.

figure;
plot(0:49,Response(:,2,3))
title("IRF of IB When Y Is Shocked")
xlabel("Observation Time")
ylabel("Response")
grid on

The bond rate fades slowly when real income is shocked at time 0.

Monte Carlo Confidence Intervals on True IRF

Consider the 4-D VAR(2) model in Estimate and Plot VAR Model IRF. Estimate and plot its orthogonalized IRF and 95% Monte Carlo confidence intervals on the true IRF.

Load the Danish money and income data set, then estimate the VAR(2) model.

load Data_JDanish

Mdl = varm(4,2);
Mdl.SeriesNames = DataTable.Properties.VariableNames;
Mdl = estimate(Mdl,DataTable.Series);

Estimate the IRF and corresponding 95% Monte Carlo confidence intervals from the estimated VAR(2) model.

rng(1); % For reproducibility
[Response,Lower,Upper] = irf(Mdl);

Response, Lower, and Upper are 20-by-4-by-4 arrays representing the orthogonalized IRF of Mdl and corresponding lower and upper bounds of the confidence intervals. For all arrays, rows correspond to consecutive time points from time 0 to 19, columns correspond to variables receiving a one-standard-deviation innovation shock at time 0, and pages correspond to responses of variables to the variable being shocked. Mdl.SeriesNames specifies the variable order.

Plot the orthogonalized IRF with its confidence bounds of the bond rate when real income is shocked at time 0.

irfshock2resp3 = Response(:,2,3);
IRFCIShock2Resp3 = [Lower(:,2,3) Upper(:,2,3)];

figure;
h1 = plot(0:19,irfshock2resp3);
hold on
h2 = plot(0:19,IRFCIShock2Resp3,'r--');
legend([h1 h2(1)],["IRF" "95% Confidence Interval"])
xlabel("Time Index");
ylabel("Response");
title("IRF of IB When Y Is Shocked");
grid on
hold off

The effect of the impulse to real income on the bond rate fades after 10 periods.

Bootstrap Confidence Intervals on True IRF

Consider the 4-D VAR(2) model in Estimate and Plot VAR Model IRF. Estimate and plot its orthogonalized IRF and 90% bootstrap confidence intervals on the true IRF.

Load the Danish money and income data set, then estimate the VAR(2) model. Return the residuals from model estimation.

load Data_JDanish

Mdl = varm(4,2);
Mdl.SeriesNames = DataTable.Properties.VariableNames;
[Mdl,~,~,E] = estimate(Mdl,DataTable.Series);
T = size(DataTable,1) % Total sample size

T = 55

n = size(E,1)         % Effective sample size

n = 53

E is a 53-by-4 array of residuals. Columns correspond to the variables in Mdl.SeriesNames. The estimate function requires Mdl.P = 2 observations to initialize a VAR(2) model for estimation. Because presample data (Y0) is unspecified, estimate takes the first two observations in the specified response data to initialize the model. Therefore, the resulting effective sample size is T – Mdl.P = 53, and rows of E correspond to the observation indices 3 through T.

Estimate the orthogonalized IRF and corresponding 90% bootstrap confidence intervals from the estimated VAR(2) model. Draw 500 paths of length n from the series of residuals.

rng(1); % For reproducibility
[Response,Lower,Upper] = irf(Mdl,"E",E,"NumPaths",500,...
    "Confidence",0.9);

Plot the orthogonalized IRF with its confidence bounds of the bond rate when real income is shocked at time 0.

irfshock2resp3 = Response(:,2,3);
IRFCIShock2Resp3 = [Lower(:,2,3) Upper(:,2,3)];

figure;
h1 = plot(0:19,irfshock2resp3);
hold on
h2 = plot(0:19,IRFCIShock2Resp3,'r--');
legend([h1 h2(1)],["IRF" "90% Confidence Interval"])
xlabel("Time Index");
ylabel("Response");
title("IRF of IB When Y Is Shocked");
grid on
hold off

The effect of the impulse to real income on the bond rate fades after 10 periods.

Input Arguments

Mdl — VAR model

varm model object

VAR model, specified as a varm model object created by varm or estimate. Mdl must be fully specified.

Name-Value Arguments

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Before R2021a, use commas to separate each name and value, and enclose Name in quotes.

Example: 'NumObs',10,'Method',"generalized" specifies estimating a generalized IRF for 10 time points starting at time 0, during which irf applies the shock, and ending at period 9.

Options for All IRFs

NumObs — Number of periods

20 (default) | positive integer

Number of periods for which irf computes the IRF, specified as the comma-separated pair consisting of 'NumObs' and a positive integer. NumObs specifies the number of observations to include in the IRF (the number of rows in Response).

Example: 'NumObs',10 specifies the inclusion of 10 time points in the IRF starting at time 0, during which irf applies the shock, and ending at period 9.

Data Types: double

Method — IRF computation method

"orthogonalized" (default) | "generalized"

IRF computation method, specified as the comma-separated pair consisting of 'Method' and a value in this table.

Value	Description
"orthogonalized"	Compute impulse responses using orthogonalized, one-standard-deviation innovation shocks. irf uses the Cholesky factorization of Mdl.Covariance for orthogonalization.
"generalized"	Compute impulse responses using one-standard-deviation innovation shocks.

Example: 'Method',"generalized"

Data Types: string

Options for Confidence Bound Estimation

NumPaths — Number of sample paths

100 (default) | positive integer

Number of sample paths (trials) to generate, specified as the comma-separated pair consisting of 'NumPaths' and a positive integer.

Example: 'NumPaths',1000 generates 1000 sample paths from which the software derives the confidence bounds.

Data Types: double

SampleSize — Number of observations for Monte Carlo simulation or bootstrap per sample path

positive integer

Number of observations for the Monte Carlo simulation or bootstrap per sample path, specified as the comma-separated pair consisting of 'SampleSize' and a positive integer.

• If Mdl is an estimated varm model object (an object returned by estimate and unmodified thereafter), then the default is the sample size of the data to which the model is fit (see summarize).

• If irf estimates confidence bounds by conducting a Monte Carlo simulation (for details, see E), you must specify SampleSize.

• If irf estimates confidence bounds by bootstrapping residuals, the default is the length of the specified series of residuals (size(E,1)).

Example: If you specify 'SampleSize',100 and do not specify the 'E' name-value pair argument, the software estimates confidence bounds from NumPaths random paths of length 100 from Mdl.

Example: If you specify 'SampleSize',100,'E',E, the software resamples, with replacement, 100 observations (rows) from E to form a sample path of innovations to filter through Mdl. The software forms NumPaths random sample paths from which it derives confidence bounds.

Data Types: double

Y0 — Presample response data

numeric matrix

Presample response data that provides initial values for model estimation during the simulation, specified as the comma-separated pair consisting of 'Y0' and a numpreobs-by-numseries numeric matrix.

Rows of Y0 correspond to periods in the presample, and the last row contains the latest presample response. numpreobs is the number of specified presample responses and it must be at least Mdl.P. If numpreobs exceeds Mdl.P, then irf uses only the latest Mdl.P rows.

numseries is the dimensionality of the input VAR model Mdl.NumSeries. Columns must correspond to the response variables in Mdl.SeriesNames.

• If Mdl is an estimated varm model object (an object returned by estimate and unmodified thereafter), irf sets Y0 to the presample response data used for estimation by default (see 'Y0').

• Otherwise, you must specify Y0.

Data Types: double

X — Predictor data

numeric matrix

Predictor data for estimating the model regression component during the simulation, specified as the comma-separated pair consisting of 'X' and a numeric matrix containing numpreds columns.

numpreds is the number of predictor variables (size(Mdl.Beta,2)).

Rows correspond to observations. X must have at least SampleSize rows. If you supply more rows than necessary, irf uses only the latest SampleSize observations. The last row contains the latest observation.

Columns correspond to individual predictor variables. All predictor variables are present in the regression component of each response equation.

To maintain model consistency when irf estimates the confidence bounds, a good practice is to specify X when Mdl has a regression component. If Mdl is an estimated model, specify the predictor data used during model estimation (see 'X').

By default, irf excludes the regression component from confidence bound estimation, regardless of its presence in Mdl.

Data Types: double

E — Series of residuals from which to draw bootstrap samples

numeric matrix

Series of residuals from which to draw bootstrap samples, specified as the comma-separated pair consisting of 'E' and a numeric matrix containing numseries columns. irf assumes that E is free of serial correlation.

Columns contain the residual series corresponding to the response series names in Mdl.SeriesNames.

If Mdl is an estimated varm model object (an object returned by estimate), you can specify E as the inferred residuals from estimation (see E or infer).

By default, irf derives confidence bounds by conducting a Monte Carlo simulation.

Data Types: double

Confidence — Confidence level

0.95 (default) | numeric scalar in [0,1]

Confidence level for the confidence bounds, specified as a numeric scalar in the interval [0,1].

For each period, randomly drawn confidence intervals cover the true response 100*Confidence% of the time.

The default value is 0.95, which implies that the confidence bounds represent 95% confidence intervals.

Data Types: double

Output Arguments

Response — IRF

numeric array

IRF, returned as a numobs-by-numseries-by-numseries numeric array. numobs is the value of NumObs. Columns and pages correspond to the response variables in Mdl.SeriesNames.

Response(t + 1,j,k) is the impulse response of variable k to a one-standard-deviation innovation shock to variable j at time 0, for t = 0, 1, ..., numObs – 1, j = 1,2,...,numseries, and k = 1,2,...,numseries.

Lower — Lower confidence bounds

numeric array

Lower confidence bounds, returned as a numobs-by-numseries-by-numseries numeric array. Elements of Lower correspond to elements of Response.

Lower(t + 1,j,k) is the lower bound of the 100*Confidence% percentile interval on the true impulse response of variable k to a one-standard-deviation innovation shock to variable j at time 0.

Upper — Upper confidence bounds

numeric array

Upper confidence bounds, returned as a numobs-by-numseries-by-numseries numeric array. Elements of Upper correspond to elements of Response.

Upper(t + 1,j,k) is the upper bound of the 100*Confidence% percentile interval on the true impulse response of variable k to a one-standard-deviation innovation shock to variable j at time 0.

More About

Impulse Response Function

An impulse response function (IRF) of a time series model (or dynamic response of the system) measures the changes in the future responses of all variables in the system when a variable is shocked by an impulse. In other words, the IRF at time t is the derivative of the responses at time t with respect to an innovation at time t₀ (the time that innovation was shocked), t ≥ t₀.

Consider a numseries-D VAR(p) model for the multivariate response variable y_t. In lag operator notation, the infinite lag MA representation of y_t is:

$$\begin{array}{c}{y}_t={\Phi }^{-1}(L)\left(c+\beta x_t+\delta t\right)+{\Phi }^{-1}(L){\epsilon }_t\\ =\Omega (L)\left(c+\beta x_t+\delta t\right)+\Omega (L){\epsilon }_t.\end{array}$$

The general form of the IRF of y_t shocked by an impulse to variable j by one standard deviation of its innovation m periods into the future is:

$${\psi }_j(m)=C_m{e}_j.$$

• e_j is a selection vector of length numseries containing a 1 in element j and zeros elsewhere.

• For the orthogonalized IRF, $$C_m={\Omega }_{m}P,$$ where P is the lower triangular factor in the Cholesky factorization of Σ, and Ω_m is the lag m coefficient of Ω(L).

• For the generalized IRF, $$C_m={\sigma }_j^{-1}{\Omega }_m\Sigma ,$$ where σ_j is the standard deviation of innovation j.

• The IRF is free of the model constant, regression component, and time trend.

Vector Autoregression Model

A vector autoregression (VAR) model is a stationary multivariate time series model consisting of a system of m equations of m distinct response variables as linear functions of lagged responses and other terms.

A VAR(p) model in difference-equation notation and in reduced form is

$$y_t=c+{\Phi }_1{y}_{t-1}+{\Phi }_2{y}_{t-2}+\mathrm{...}+{\Phi }_p{y}_{t-p}+\beta x_t+\delta t+{\epsilon }_t.$$

• y_t is a numseries-by-1 vector of values corresponding to numseries response variables at time t, where t = 1,...,T. The structural coefficient is the identity matrix.

• c is a numseries-by-1 vector of constants.

• Φ_j is a numseries-by-numseries matrix of autoregressive coefficients, where j = 1,...,p and Φ_p is not a matrix containing only zeros.

• x_t is a numpreds-by-1 vector of values corresponding to numpreds exogenous predictor variables.

• β is a numseries-by-numpreds matrix of regression coefficients.

• δ is a numseries-by-1 vector of linear time-trend values.

• ε_t is a numseries-by-1 vector of random Gaussian innovations, each with a mean of 0 and collectively a numseries-by-numseries covariance matrix Σ. For t ≠ s, ε_t and ε_s are independent.

Condensed and in lag operator notation, the system is

$$\Phi (L)y_t=c+\beta x_t+\delta t+{\epsilon }_t,$$

where $$\Phi (L)=I-{\Phi }_{1}L-{\Phi }_2{L}^2-\mathrm{...}-{\Phi }_p{L}^p$$, Φ(L)y_t is the multivariate autoregressive polynomial, and I is the numseries-by-numseries identity matrix.

Algorithms

• NaN values in Y0, X, and E indicate missing data. irf removes missing data from these arguments by list-wise deletion. Each argument, if a row contains at least one NaN, then irf removes the entire row.

  List-wise deletion reduces the sample size, can create irregular time series, and can cause E and X to be unsynchronized.

• If Method is "orthogonalized", then the resulting IRF depends on the order of the variables in the time series model. If Method is "generalized", then the resulting IRF is invariant to the order of the variables. Therefore, the two methods generally produce different results.

• If Mdl.Covariance is a diagonal matrix, then the resulting generalized and orthogonalized IRFs are identical. Otherwise, the resulting generalized and orthogonalized IRFs are identical only when the first variable shocks all variables (that is, all else being the same, both methods yield the same value of Response(:,1,:)).

• The predictor data X represents a single path of exogenous multivariate time series. If you specify X and the VAR model Mdl has a regression component (Mdl.Beta is not an empty array), irf applies the same exogenous data to all paths used for confidence interval estimation.

• irf conducts a simulation to estimate the confidence bounds Lower and Upper.

  • If you do not specify residuals E, then irf conducts a Monte Carlo simulation by following this procedure:

    1. Simulate NumPaths response paths of length SampleSize from Mdl.

    2. Fit NumPaths models that have the same structure as Mdl to the simulated response paths. If Mdl contains a regression component and you specify X, then irf fits the NumPaths models to the simulated response paths and X (the same predictor data for all paths).

    3. Estimate NumPaths IRFs from the NumPaths estimated models.

    4. For each time point t = 0,…,NumObs, estimate the confidence intervals by computing 1 – Confidence and Confidence quantiles (upper and lower bounds, respectively).

  • If you specify residuals E, then irf conducts a nonparametric bootstrap by following this procedure:

    1. Resample, with replacement, SampleSize residuals from E. Perform this step NumPaths times to obtain NumPaths paths.

    2. Center each path of bootstrapped residuals.

    3. Filter each path of centered, bootstrapped residuals through Mdl to obtain NumPaths bootstrapped response paths of length SampleSize.

    4. Complete steps 2 through 4 of the Monte Carlo simulation, but replace the simulated response paths with the bootstrapped response paths.

References

[1] Hamilton, James D. Time Series Analysis. Princeton, NJ: Princeton University Press, 1994.

[2] Lütkepohl, Helmut. New Introduction to Multiple Time Series Analysis. New York, NY: Springer-Verlag, 2007.

[3] Pesaran, H. H., and Y. Shin. "Generalized Impulse Response Analysis in Linear Multivariate Models." Economic Letters. Vol. 58, 1998, pp. 17–29.

Version History

Introduced in R2019a