Historical-ratio bridge

A way to estimate a number you can't measure directly by anchoring it to one you can, through a relationship that has held steady over time — and labeling the result an estimate, not a measurement. If the relationship isn't stable, you don't bridge.

Why it matters

It's how you get an honest number when no direct one exists — without inventing it. The Cost Index builds a derived reading this way: when an item has no public wholesale series, it bridges from a related public series whose historical ratio has held, draws it as a dashed estimate with a confidence band, and labels it — never dressed up as a measurement. If the ratio isn't stable, it doesn't bridge; the item goes absent with the reason.

Frequently asked

What is a historical-ratio bridge?

It's a general estimation principle: when you can't measure something directly, you anchor it to something you can measure through a relationship between the two that has held steady over time, and you disclose that the result is an estimate. If a wholesale price has no public series of its own but moves in a stable relationship with one that does, you can read the one you have and bridge across to the one you don't — as a labeled estimate, never as a measurement.

When does the bridge refuse to give an answer?

When the relationship isn't stable. The whole method rests on the ratio holding, so it's only used where the two series are tested as genuinely cointegrated — the gap between them mean-reverts — rather than two lines that happen to drift together. If the relationship fails that test, or there's no real overlap window to measure it on, or there's no plausible reason the two should be linked, the bridge yields no number. In the Cost Index that ingredient goes absent with the reason, instead of being shown a made-up price.

How does the Muntin Cost Index use a historical-ratio bridge?

It's how the index builds a derived reading. When no direct public wholesale series exists for an item, the index bridges from a related public series using a historical ratio that has held steady, and tags the result derived — drawn with a dashed line, a confidence band sized to the relationship's own error, and an estimate badge. A bridge never overrides a directly measured level, and it's re-fit as more overlapping history accumulates.

What is cointegration in forecasting?

Cointegration is when two series move together in a relationship that stays stable over time — the gap between them keeps returning to its average rather than wandering off. The Cost Index uses it as the test for a historical-ratio bridge: it only estimates a missing price from a related series when the two are genuinely cointegrated, not just drifting the same way by coincidence.