Value-weighted forecast loss

A technically strong model can still mislead if validation leaks future information. This brief explores value-weighted forecast loss and the checks needed for credible comparison.

Practitioner lens

For value-weighted forecast loss, validation must respect the information set available at each forecast origin. Features published after the decision date create leakage and overstate likely performance.

First experiment

Check calibration and decision impact separately; sharp intervals are useful only when they cover at the promised rate.

This is an original educational brief. It is a starting point for discussion, not a claim that one technique fits every supply chain.