Everything rots to the same place eventually, and the only question is whether you bought before or after it showed up

The carry doing more damage than the repair cost is the part nobody believes until it lands on them. Run numbers on a 1940s duplex in a market like Central Falls and the line that keeps moving the pro forma is the rate on the construction loan accruing while a subcontractor sits three weeks out, because every GC in Providence County is booked until fall. The piece worth adding, and it rarely gets said clearly: the six-week slip probably did not start when the rot was found. It started the moment the scope had to be repermitted or restructured, because a load-bearing wall changes the permit class in most Rhode Island municipalities and that triggers a new review cycle. The gap between discovery and revised permit approval is where the weeks actually go, ahead of any physical work. Anything touching structure in Pawtucket specifically tends to run four to six weeks on the amended permit alone, before a single nail. So the contingency math needs a time buffer that is completely decoupled from the dollar contingency, because those two do not move together.

The blind-price-it-in move is where this usually lands, and the number matters a lot. On a pre-1970 three-unit in a wet market, the open question is whether 15k flat or something closer to 8 to 12 per unit is defensible as a structural rot assumption before anyone has seen the interior. What I would want to know is how the 38k broke down between materials and labor, because if the joists and subfloor alone were 20k of it, that changes how you size contingency on similar buildings.

The one thing that never seems to work is the tiered contingency. People build 10 percent for cosmetic unknowns and tell themselves they have a separate mental bucket for structural. There is no separate bucket when you are standing in that opening. Pre-1970 in a wet New England market, the sound underwriting is a flat 20k structural assumption on top of whatever the inspection says, before you even know what you are buying. Pawtucket, Providence, same story. Six weeks of carry at hard money rates is the number that moves a minimum margin requirement, and it moves it more than the repair itself does.

The 15 percent contingency convention is worth questioning, because people set that number off purchase price and that has always looked like the wrong base. Fifteen percent of 112k is 16,800, which was never going to cover 38k even with carry softening the gap. Running contingency off estimated hard construction cost instead of acquisition shifts the number meaningfully on anything pre-1960. For a gut on three units in a wet market, 22 to 25 percent of the construction budget sitting liquid before closing is the more defensible figure, rather than a percentage of what was paid for the dirt and the rot together. What actually costs people on hidden structural work like yours is the decision lag inside the project more than the repair dollars. You found it, you stopped, you waited on a structural engineer, you waited on the revised scope, and your subs shuffled to other jobs while that played out, so you either paid a premium to get them back or waited again. That is where the six weeks came from, well beyond the actual repair time. A 10-day discovery turning into a 7-week delay is common purely because nobody had a structural engineer on retainer and the good ones in the area were booked 3 weeks out. Pre-identifying who takes that call before a wall ever gets opened is the piece of prep most often skipped on a first wet-climate gut.

The part that never shows up in the underwriting conversation is which unit the rot is in. A second-floor bathroom draining into a load-bearing wall is close to worst-case anatomy for spread, because gravity pulls the water exactly where you do not want it. A ground-floor bathroom of the same vintage and the same neglect usually stays more contained. The sound response is to differentiate contingency by floor and by whether the bathrooms are stacked. Pre-1970 with stacked plumbing and second-floor baths warrants 22 to 25 percent where 15 would once have looked prudent. Six extra weeks of carry at hard money rates on a 112k acquisition probably hurt worse than most people would admit out loud.

On pre-1970 in a wet climate I just add a flat 20k to the rehab number before I even do a scope, and I don't apologize for it in my offers. Passed on a few deals because of it, but the ones I closed I wasn't surprised when the walls opened up. The six-week carry is the thing that actually changes the math more than people admit going in, because it compounds against everything else that slipped.

Thirty-eight extra on a 112k purchase is a third of the acquisition price, and that ratio is what actually matters here. The carry hurting more than the repair is a symptom of the real problem, which is that your contingency was sized as a percentage of construction budget rather than a percentage of what the building actually was. The same failure shows up constantly in older stock. Take a 1958 duplex where an operator adds 22k of contingency against a 140k rehab estimate and feels responsible about it, then needs 47k more because a chimney has been weeping into the framing since somewhere around the Reagan administration. The contingency math looks fine right up until the building shows what it actually needs. The better approach on anything pre-1965 and coastal is a separate line item called structural unknown. You set it as a flat dollar figure by asking what the worst plausible version of this building costs to fix, and you fund half of that before you close.

What was the exit number before the 38k hit, and did the deal still make money? If it did, the lesson may not be about contingency sizing at all. It may just be that 112k in Pawtucket with a 15 percent cushion was a good enough basis to absorb a bad surprise and still work. That changes how the whole situation should be read going forward.