You fix a price with a buyer at contract. You buy the materials to build that home six to nine months later. Lumber, steel, copper and gypsum do not hold still for six to nine months, and nobody in the business owns the gap.
It does not show up as a loss. It shows up as a margin that came in under the pro forma, explained after the fact by whichever cost code moved most, on a home that was priced before anyone knew.
The reason it stays invisible is that the exposure is created on the sales floor and paid in purchasing. By the time it is a purchase order it is a price, not a decision, and the decision that mattered — when to lock — was made months earlier by nobody in particular.
The model derives the lag per home, weights a commodity basket to your own bill of materials rather than to a published index, and splits the exposure into what a lock decision could have prevented and what the market simply did.
The gap between contract signature and material commitment. The lock date is a decision somebody makes, or fails to make.
Four commodities move enough to be worth covering and none of them is covered. A treasury decision more than a procurement one.
Real price change on material already committed. No model prevents it, and it is not claimed as recoverable.
A different product at a different price is not escalation. It is a design decision and it stays.
The rest of this model — the data sets it runs on, the cause split, the working behind the reference figure and a sample output — goes out by email. Two fields, nothing else.