To implement our proposed tests for predictability
load your chosen returns and predictor data into Gauss as
vectors named "ret" and "pred", respectively, and run
the programme "RUNME - Confidence Intervals Beta.gp".

The returns and predictor data of Campbell and Yogo
(2006) is included in the spreadsheet "data.xlxs".

This programme returns:

1) The estimate of the correlation parameter delta.
2) The DF-OLS test statistic applied to the predictor.
3) The DF-GLS test statistic applied to the predictor.
4) The standard predictive regression t-test.
5) An estimate of |theta| as defined in (20).
6) The Dickey Fuller normalised bias test statistic.
7) The Dickey Fuller normalised bias critical value.
8) The upper and lower bound of the confidence interval
for beta from the t-OLS, t-GLS, Q-OLS, Q-GLS, U^hyb,
W_1^hyb and W_2^hyb test procedures.


