Main Content

Simulate Stationary Processes

Simulate AR Process

This example shows how to simulate sample paths from a stationary AR(2) process without specifying presample observations.

Create Model

Specify the AR(2) model

yt=0.5+0.7yt-1+0.25yt-2+εt,

where the innovation process is Gaussian with variance 0.1.

Mdl = arima('Constant',0.5,'AR',{0.7 0.25},'Variance',.1);

Generate One Sample Path

Generate one sample path (with 50 observations) from the specified model, and plot.

rng(5)
Y = simulate(Mdl,50);

figure
plot(Y)
xlim([0,50])
title('Simulated AR(2) Process')

Figure contains an axes object. The axes object with title Simulated AR(2) Process contains an object of type line.

Because presample data was not specified, simulate sets the two required presample observations equal to the unconditional mean of the process

μ=c(1-ϕ1-ϕ2).

The unconditional variance of the process is

σ2=(1-ϕ2)(1+ϕ2)σε2(1-ϕ2)2-ϕ12.

Compute the unconditional mean and variance of the process.

c = Mdl.Constant;
phi = Mdl.AR;
sigmaEps2 = Mdl.Variance;
mu = c/(1-phi{1}-phi{2})
mu = 10.0000
sigma2 = (1-phi{2})/(1+phi{2})*sigmaEps2/((1-phi{2})^2 - phi{1}^2)
sigma2 = 0.8276

Generate Many Sample Paths

Generate 1000 sample paths, each with 50 observations.

Y = simulate(Mdl,50,'NumPaths',1000);

figure
subplot(2,1,1)
plot(Y,'Color',[.85,.85,.85])
title('Simulated AR(2) Process')
hold on
h = plot(mean(Y,2),'k','LineWidth',2);
legend(h,'Simulation Mean','Location','NorthWest')
hold off
subplot(2,1,2)
plot(var(Y,0,2),'r','LineWidth',2)
title('Process Variance')
hold on
plot(1:50,sigma2*ones(50,1),'k--','LineWidth',1.5)
legend('Simulation','Theoretical',...
       'Location','SouthEast')
hold off

Figure contains 2 axes objects. Axes object 1 with title Simulated AR(2) Process contains 1001 objects of type line. This object represents Simulation Mean. Axes object 2 with title Process Variance contains 2 objects of type line. These objects represent Simulation, Theoretical.

The simulation mean is constant over time. This is consistent with the definition of a stationary process. The process variance is not constant over time, however. There are transient effects at the beginning of the simulation due to the absence of presample data.

Around observation 50, the simulated variance approaches the theoretical variance.

Oversample to Reduce Transient Effects

To reduce transient effects, one option is to oversample the process. For example, to sample 50 observations, you can generate paths with more than 50 observations, and discard all but the last 50 observations as burn-in. Here, simulate paths of length 150, and discard the first 100 observations.

Y = simulate(Mdl,150,'NumPaths',1000);
Y = Y(101:end,:);

figure
subplot(2,1,1)
plot(Y,'Color',[.85,.85,.85])
title('Simulated AR(2) Process')
hold on
h = plot(mean(Y,2),'k','LineWidth',2);
legend(h,'Simulation Mean','Location','NorthWest')
hold off
subplot(2,1,2)
plot(var(Y,0,2),'r','LineWidth',2)
xlim([0,50])
title('Process Variance')
hold on
plot(1:50,sigma2*ones(50,1),'k--','LineWidth',1.5)
legend('Simulation','Theoretical',...
       'Location','SouthEast')
hold off

Figure contains 2 axes objects. Axes object 1 with title Simulated AR(2) Process contains 1001 objects of type line. This object represents Simulation Mean. Axes object 2 with title Process Variance contains 2 objects of type line. These objects represent Simulation, Theoretical.

The realizations now look like draws from a stationary stochastic process. The simulation variance fluctuates (due to Monte Carlo error) around the theoretical variance.

Simulate MA Process

This example shows how to simulate sample paths from a stationary MA(12) process without specifying presample observations.

Create Model

Specify the MA(12) model

yt=0.5+εt+0.8εt-1+0.2εt-12,

where the innovation distribution is Gaussian with variance 0.2.

Mdl = arima('Constant',0.5,'MA',{0.8,0.2},...
    'MALags',[1,12],'Variance',0.2);

Generate Sample Paths

Generate 200 sample paths, each with 60 observations.

rng(5)
Y = simulate(Mdl,60,'NumPaths',200);

figure
plot(Y,'Color',[.85,.85,.85])
hold on
h = plot(mean(Y,2),'k','LineWidth',2);
legend(h,'Simulation Mean','Location','NorthWest')
title('MA(12) Process')
hold off

Figure contains an axes object. The axes object with title MA(12) Process contains 201 objects of type line. This object represents Simulation Mean.

For an MA process, the constant term is the unconditional mean. The simulation mean is approximately c = 0.5.

Plot Simulation Variance

The unconditional variance of the process is

σ2=(1+θ12+θ122)σε2.

Compute the unconditional variance.

theta = cell2mat(Mdl.MA);
sigmaEps2 = Mdl.Variance;
sigma2 = (1+sum(theta.^2))*sigmaEps2
sigma2 = 0.3360

Because the model is stationary, the unconditional variance should be constant across all times. Plot the simulation variance, and compare it to the theoretical variance.

figure
plot(var(Y,0,2),'Color',[.75,.75,.75],'LineWidth',1.5)
xlim([0,60])
title('Unconditional Variance')
hold on
plot(1:60,sigma2*ones(60,1),'k--','LineWidth',2)
legend('Simulation','Theoretical',...
       'Location','SouthEast')
hold off

Figure contains an axes object. The axes object with title Unconditional Variance contains 2 objects of type line. These objects represent Simulation, Theoretical.

There appears to be a short burn-in period at the beginning of the simulation. During this time, the simulation variance is lower than expected. Afterwards, the simulation variance fluctuates around the theoretical variance.

Generate Many Sample Paths

Simulate 10,000 paths from the model, each with length 1000. Plot the simulation variance.

YM = simulate(Mdl,1000,'NumPaths',10000);
figure
plot(var(YM,0,2),'Color',[.75,.75,.75],'LineWidth',1.5)
ylim([0.3,0.36])
title('Unconditional Variance')
hold on
plot(1:1000,sigma2*ones(1000,1),'k--','LineWidth',2)
legend('Simulation','Theoretical',...
       'Location','SouthEast')
hold off

Figure contains an axes object. The axes object with title Unconditional Variance contains 2 objects of type line. These objects represent Simulation, Theoretical.

The Monte Carlo error is reduced when more realizations are generated. There is much less variability in the simulation variance, which tightly fluctuates around the theoretical variance.

See Also

|

Related Examples

More About