Selling in summer when you broke ground in winter, the carry math is never what you budgeted

The seasonal pricing risk is almost never a named line, and this case shows exactly why it should be.

The assumption doing the most work in that pro forma was not the construction cost buffer. It was the implicit claim that exit price is independent of exit date. The 7,500 in extra carry is visible and calculable after the fact, but the 18k price cut was invisible at underwriting because the model had no mechanism for it. The builder budgeted a contingency percentage on costs and assumed the revenue side would hold wherever the calendar landed.

One way to structure this differently: model two exit scenarios explicitly, a peak-season close and a trough-season close, with separate assumed sale prices for each. The gap between those numbers becomes a named line in sensitivity analysis rather than a surprise at day 60 on market. In a market where February-to-April drives competition and summer stalls it, that gap might be 4 to 6 percent of price, which on a 380k home is 15k to 23k. That range is material relative to a 42k projected profit margin. It is not a contingency on costs; it is a separate risk class entirely.

The other mechanism worth modeling is the carry cost of a delayed sale independent of a delayed build. Even finishing on schedule, a spec home that closes 60 days after list adds roughly 5k at that loan rate. A price cut to accelerate velocity is sometimes cheaper than holding, but the decision needs the numbers in front of you before you are staring at an offer.

The risk this thread did not mention is that weather delay and seasonal penalty compounded here. Either alone was manageable. Together they moved the exit from the best window to one of the weakest. A builder who breaks ground in winter needs to underwrite the probability that both risks arrive simultaneously, not just the construction schedule in isolation.

What does your pro forma currently use as the exit price assumption, a comp from peak season or a trailing average across all closed months?

Nobody I've watched do this builds the calendar backwards from the best close window and then asks what groundbreak date that requires. They build forward from when they can start and hope the exit lands somewhere reasonable. That's the whole problem in the post right there, and the 25k variance on 42k of profit is a clean illustration of it. The weather delay didn't kill the deal, the market it landed in did.

What I'd want to know before I felt confident modeling this at all is whether the market in question has a hard seasonal cliff or just a softer one. The example reads like a pretty sharp February through April window with July genuinely dead, and those markets exist, but I've also seen markets where summer inventory drops enough that a well-priced house still gets multiple looks in August. The 60-day hold and 18k cut suggests this one was the hard-cliff kind, which means the only real protection was a groundbreak date that guaranteed a spring CO. And if winter weather delays are a known variable in that region, the builder probably needed a six-week buffer baked into the schedule before the first shovel moved, not a contingency dollar figure added after the fact.