README.md

In vst/mgarch: Tools for simulating, estimating and diagnosing MGARCH processes

mgarch - R Package for MGARCH Model Simulation and Estimation

mgarch is an R package for simulating and estimating multivariate GARCH models. Currently, only BEKK, GJR and DCC specifications are implemented.

Basic Usage

After you install the package, you can load the library:

> library(mgarch)

Probably, you are interested in the following functions:

1. mvBEKK.sim: Simulates an MGARCH-BEKK process
2. mvBEKK.est: Estimates an MGARCH-BEKK model
3. mvBEKK.diag: Provides diagnostics for an estimated MGARCH-BEKK model

You can get help from the embedded documentation by issuing:

> ?mvBEKK.sim
> ?mvBEKK.est
> ?mvBEKK.diag

As an example, you can simulate an MGARCH-BEKK process as follows:

> sim <- mvBEKK.sim(series.count=2, T=1200)

This will give you two vectors of 2500 observations along with many other objects. Checkout the help for further details.

Now create a data frame with the given two vectors:

> eps <- data.frame(s1=sim$eps[[1]], s2=sim$eps[[2]])

The first and last few rows of the resulting data frame looks like as follows:

> head(eps)
       s1      s2
1 -0.7302 -0.9066
2 -1.0601 -0.9100
3 -2.0181  0.8561
4  1.1678  0.9071
5 -0.1927  0.2441
6  0.9193 -0.0993

> tail(eps)
          s1      s2
2495 -1.0160 -0.9626
2496  1.5960 -0.5020
2497 -1.6413  1.1664
2498  2.0670 -1.9811
2499  0.6587  0.1838
2500  1.0310  1.2592

When you need to supply your own data, make sure that it follows the same convention as above.

Now, we will estimate a model:

> est <- mvBEKK.est(eps)

The estimation may take longer depending on the number of series and the length of series. The result contains quite a lot of information:

> names(est)
[1] "eps"               "series.length"
[3] "estimation.time"   "total.time"
[5] "order"             "estimation"
[7] "aic"               "asy.se.coef"
[9] "est.params"        "cor"
[11] "sd"                "H.estimated"
[13] "eigenvalues"       "uncond.cov.matrix"
[15] "residuals"

Check the documentation for details.

You can see a summary by using the mvBEKK.diag command:

> mvBEKK.diag(est)

    Number of estimated series :  2
    Length of estimated series :  1200
    Estimation Time            :  3.971
    Total Time                 :  4.788
    BEKK order                 :  1 1
    Eigenvalues                :  0.9371 0.5384 0.5016 0.3784
    aic                        :  5592
    unconditional cov. matrix  :  3.821 -0.5761 -0.5761 11.3
    var(resid 1 )                :  0.9996
    mean(resid 1 )               :  0.0005716
    var(resid 2 )                :  0.9922
    mean(resid 2 )               :  0.04112
    Estimated parameters       :

    C estimates:
     [,1]   [,2]
[1,] 1.16 0.2905
[2,] 0.00 0.9404

    ARCH estimates:
        [,1]     [,2]
[1,] -0.5225  0.14426
[2,]  0.3576 -0.09822

    GARCH estimates:
        [,1]     [,2]
[1,] -0.5358 -0.07182
[2,] -0.2370  0.90391

    asy.se.coef                :

    C estimates, standard errors:
       [,1]   [,2]
[1,] 0.1767 0.7308
[2,] 0.0000 0.2853

    ARCH estimates, standard errors:
        [,1]    [,2]
[1,] 0.03618 0.03759
[2,] 0.03326 0.03972

    GARCH estimates, standard errors:
       [,1]    [,2]
[1,] 0.0519 0.09403
[2,] 0.1336 0.03051
Called from: mvBEKK.diag(est)
Browse[1]>

This command will also produce some helpful plots.

Copyrights and Licensing

This software is licensed under the GPLv3::

Copyright (C) 2005-2014 Harald Schmidbauer <[e-mail]>,
    Angi Roesch <[e-mail]>,
    Vehbi Sinan Tunalioglu <[e-mail]>

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program.  If not, see <http://www.gnu.org/licenses/>.

vst/mgarch documentation built on May 3, 2019, 7:09 p.m.