After downloading extract the files to a directory and remember where you put it to.
The necessesary subdirectories will be created automatically.

For running the quartalisation with internal validation and replication of the paper results in appendix B do the following:

Open DisAggMainAnybody.OX with a text editor (e.g. OxEdit) and adjust line 56 according to where you put the files.

Run the program with OX. The default input reproduces table 5 in the appendix B to the article. Columns 6--13 are the estimates obtained with the 
improved estimation approach, cols 1--5 are estimated with classic Chow-Lin.

You may do other disaggregations by carefully adjusting the input variables.
	- adjust the model specification section in lines 31-36 (vary the choice of trend, seasonal dummies etc.)
	- inform OX about the properties of the input data in lines 68-73
	- adjust the Data input section in lines 84-84
	- load other date files (make sure you use exactly the same file layout as in the examples)

The input files are provided in Microsoft Excel and in ASCII format. Column headers correspond to the variable names in the published article.

The output will be found in the subdirectory OUTPUT. You can choose graphical output too. Regarding the text output the
results listed in the paper are written in columns six and the following. The first five columns replicate the
Chow-Lin-estimator and calculates a ersricted AR(1) model. The test statistic given in column chi^2(1).

For replicating the simulation exercise and running further simulations open NonlARMA.ox and follow the instructions given in the header. Note that the output will be saved to a file.

If you have any queries regarding the output, do write to me: mail@cmueller.ch

Good luck!