INTERCEPT CORRECTIONS AND STRUCTURAL-CHANGE

Analyses of forecasting that assume a constant, time-invariant data ge nerating process (DGP), and so implicitly rule out structural change o r regime shifts in the economy, ignore an aspect of the real world res ponsible for some of the more dramatic historical episodes of predicti ve failure. Some models may offer greater protection against unforesee n structural breaks than others, and various tricks may be employed to robustify forecasts to change. We show that in certain states of natu re, vector autoregressions in the differences of the variables (in the spirit of Box-Jenkins time-series modelling), can outperform vector ' equilibrium-correction' mechanisms. However, appropriate intercept cor rections can enhance the performance of the latter, albeit that reduct ions in forecast bias may only be achieved at the cost of inflated for ecast error variances.