Achim Zeileis, Friedrich Leisch, Christian Kleiber and Kurt Hornik,
"Monitoring Structural Change in Dynamic Econometric Models", Journal of
Applied Econometrics, Vol. 20, No. 1, 2005, pp. 99-121.


SOFTWARE

All the results in this paper have been computed using the R system for
statistical computing (http://www.R-project.org/) and in particular the R
package strucchange (see Zeileis et al., 2002) for testing for structural
change in linear regression relationships. Both the R system and the
strucchange package are freely available at no cost under the terms of the
GNU General Public Licence (GPL) from the Comprehensive R Archive Network at
http://CRAN.R-project.org/

The results in the paper have been obtained using strucchange 1.1-1
which is also available in this data archive as source code
(strucchange_1.1-1.tar.gz) and as a compiled binary for Windows
(strucchange_1.1-1.zip). Current versions of strucchange can be
obtained from CRAN as described above.
The R version used to obtain the results in the paper was 1.8.0.


DATA

Three data sets are analyzed in this paper:

  o German M1 money demand (GermanM1),
  o U.S. labor productivity (durab),
  o S&P 500 stock returns (SP500).

For each data set, the following files are available:

  o .dat: a tab-separated ASCII table containing the data
  o .rda: the data (with some further markup) in a binary R data file
  o .Rt: an R session transcript reproducing the results from the paper

Each data set is described in more detail below.


GERMAN M1 MONEY DEMAND

This data set was analyzed in Luetkepohl et al. (1999) and the data werr
obtained from the JAE data archive at
  
  http://qed.econ.queensu.ca/jae/1999-v14.5/lutkepohl-terasvirta-wolters/

To facilitate the fitting of dynamic models in R some transformations
(lags, differences) were computed explicitely. The ASCII table (.tab) has
the following columns

  year - year
  quarter - quarter
  m - Logarithm of real M1 per capita,
  p - Logarithm of a price index,
  y - Logarithm of real per capita gross national product,
  R - Long-run interest rate,
  dm - First differences of 'm',
  dy2 - First differences of lag 2 of 'y',
  dR - First differences of 'R',
  dR1 - First differences of lag 1 of 'R',
  dp - First differences of 'p',
  m1 - Lag 1 of 'm',
  y1 - Lag 1 of 'y',
  R1 - Lag 1 of 'R',
  ecm.res - the OLS residuals of the Luetkepohl et al. (1999)
            ECM model fitted in the history period.

The data in the binary file (.rda) look very similar but contain some
markup (a "data.frame" with objects of class "ts" and "factor").
Furthermore, the data has been split explicitly into the history sample
(historyM1) and the monitoring sample (monitorM1).
This data is already available in strucchange. If the package is
installed it is available via data(GermanM1), more information can
be found on help(GermanM1).

All three files are zipped in GermanM1.zip. The .dat file is in DOS format.


U.S. LABOR PRODUCTIVITY

The data set was analyzed by Hansen (2001) and is available online from
Bruce Hansen's web page at

  http://www.ssc.wisc.edu/~bhansen/

To facilitate the fitting of dynamic models in R some transformations
(lags, differences) were computed explicitely. The ASCII table (.tab) has
the following columns

  year - year
  month - month
  y - growth rate of the Industrial Production Index to average weekly
      labor hours in the manufacturing/durables sector
  lag - Lag 1 of 'y'

The data in the binary file (.rda) look very similar but contain some
markup (a multiple time series of class "mts").
This data is already available in strucchange. If the package is
installed it is available via data(durab), more information can
be found on help(durab).

All three files are zipped in durab.zip. The .dat file is in DOS format.


S&P 500 STOCK RETURNS

The S&P 500 stock returns have been obtained from Yahoo! Finance at

  http://finance.yahoo.com/

an online portal quoting data provided by Reuters. The data set has
501 columns: The first column gives the day of the observation (in
ISO format) and the remaining columns have the abbreviated name of
the quote as used by Yahoo! Finance.

The data can also be easily obtained from within R using the function
get.hist.quote() from the tseries package. A session transcript which
shows how to do this is available in the file getSP500.Rt.
A convenience plotting function for this data set is stored in the
file plot.irefp.R.

All five files are zipped in SP500.zip. The .dat file is in DOS format.