Use the follwing command in R:

> load("SmallNumberPerm.RData)

Containded items:
- t.caar
- perm.test.caar
- ar.est.scandal
- ar.est.announce
- ar.ev.scandal
- ar.ev.announce

perm.test.caar is the function that carries out the permuatin test; t.caar is the function that computes the underlying (alternative) test statistic. Both functions have minimal comments; to see them, at the command line type the name of the function without parentheses and then hit return.

The other files contain the two data sets, two files per data set (ARs in estimation window and ARs in event window).

To replicate the results of Table 3, use the following commands:
> perm.test.caar(ar.est.scandal, ar.ev.scandal[11, ], w.sizes = c(1, 1), B = 99999)
> perm.test.caar(ar.est.scandal, ar.ev.scandal[10:12, ], w.sizes = c(1, 1), B = 99999)
> perm.test.caar(ar.est.scandal, ar.ev.scandal[9:13, ], w.sizes = c(1, 1), B = 99999)
> perm.test.caar(ar.est.scandal, ar.ev.scandal[8:14, ], w.sizes = c(1, 1), B = 99999)

To replicate the results of Table 4, use the following commands:
> perm.test.caar(ar.est.announce, ar.ev.announce[11, ], w.sizes = c(5, 4), B = 99999)
> perm.test.caar(ar.est.announce, ar.ev.announce[10:12, ], w.sizes = c(5, 4), B = 99999)
> perm.test.caar(ar.est.announce, ar.ev.announce[9:13, ], w.sizes = c(5, 4), B = 99999)
> perm.test.caar(ar.est.announce, ar.ev.announce[8:14, ], w.sizes = c(5, 4), B = 99999)


Michael Wolf
Zurich, April 2026
