Peter Cundill on forecasting

From There's Always Something to Do:

I think that intelligent forecasting (company revenues, earnings, etc.) should not seek to predict what will in fact happen in the future. Its purpose ought to be to illuminate the road, to point out obstacles and potential pitfalls and so assist management to tailor events and to bend them in a desired direction. Forecasting should be used as a device to put both problems and opportunities into perspective. It is a management tool, but it can never be a substitute for strategy, nor should it ever be used as the primary basis for portfolio investment decisions.