Flat 10% contingency is a superstition. What do you actually use?

Question

Provocation, mildly held: flat 10% contingency treats a 1962 pier-and-beam and a 2005 slab build identically, which is obviously wrong — but nearly everyone (me included, for years) does it anyway because the alternative takes effort.

What our estimate-vs-actual data suggests: contingency should scale with property age, foundation style, scope share hidden behind drywall, and trade count. For context in our local RDU market:

2000s–2010s Slab (Apex/Cary/North Raleigh infill): 5%–7% is usually plenty. You’re mostly dealing with cosmetic updates, non-load-bearing framing shifts, and predictable mechanical swaps.
1980s–1990s Frame on Crawl (Garner/Wake Forest/Clayton): 8%–12%. You start seeing polybutylene pipe swaps, aging HVAC ducting in tight crawlspaces, and minor moisture/joist remediation.
Pre-1970 Pier-and-Beam / Vented Crawl (Inside the Beltline / Downtown Durham): 15%–22%. Between unreinforced masonry footings, chronic Piedmont clay drainage issues, sub-floor rot, outdated galvanized plumbing, knob-and-tube remnants, and RRP lead rules, 10% gets wiped out before the demo crew even leaves.

On our projects, the right number ranged from 8% to 25% depending strictly on those variables. So what do you actually do — flat %, line-item-specific, age-based, or 'my number is my number'? And has anyone's method survived contact with a pre-1970 house?"

Provocation, mildly held: flat 10% contingency treats a 1962 pier-and-beam and a 2005 slab build identically, which is obviously wrong — but nearly everyone (me included, for years) does it anyway because the alternative takes effort.

What our estimate-vs-actual data suggests: contingency should scale with property age, foundation style, scope share hidden behind drywall, and trade count. For context in our local RDU market:

• 2000s–2010s Slab (Apex/Cary/North Raleigh infill): 5%–7% is usually plenty. You’re mostly dealing with cosmetic updates, non-load-bearing framing shifts, and predictable mechanical swaps.
• 1980s–1990s Frame on Crawl (Garner/Wake Forest/Clayton): 8%–12%. You start seeing polybutylene pipe swaps, aging HVAC ducting in tight crawlspaces, and minor moisture/joist remediation.
• Pre-1970 Pier-and-Beam / Vented Crawl (Inside the Beltline / Downtown Durham): 15%–22%. Between unreinforced masonry footings, chronic Piedmont clay drainage issues, sub-floor rot, outdated galvanized plumbing, knob-and-tube remnants, and RRP lead rules, 10% gets wiped out before the demo crew even leaves.

2000s–2010s Slab (Apex/Cary/North Raleigh infill): 5%–7% is usually plenty. You’re mostly dealing with cosmetic updates, non-load-bearing framing shifts, and predictable mechanical swaps.

1980s–1990s Frame on Crawl (Garner/Wake Forest/Clayton): 8%–12%. You start seeing polybutylene pipe swaps, aging HVAC ducting in tight crawlspaces, and minor moisture/joist remediation.

Pre-1970 Pier-and-Beam / Vented Crawl (Inside the Beltline / Downtown Durham): 15%–22%. Between unreinforced masonry footings, chronic Piedmont clay drainage issues, sub-floor rot, outdated galvanized plumbing, knob-and-tube remnants, and RRP lead rules, 10% gets wiped out before the demo crew even leaves.

On our projects, the right number ranged from 8% to 25% depending strictly on those variables. So what do you actually do — flat %, line-item-specific, age-based, or 'my number is my number'? And has anyone's method survived contact with a pre-1970 house?"

Ashish Acharya · Answer

Alexis, I’m with you on this. A flat 10% contingency is easy to use, but it can give a false sense of precision when the actual risk profile of two rehabs is completely different.

I’d rather size contingency around the things most likely to create surprises: property age, foundation type, how much work is hidden behind walls, condition of plumbing/electrical/HVAC, permit complexity, and how many trades are involved.

For a newer slab property with mostly cosmetic work, a smaller contingency may be reasonable. On an older pier-and-beam or crawlspace house where you’re opening walls and touching multiple systems, I’d rather underwrite more conservatively from the beginning than pretend 10% covers everything.

I also like separating known-but-not-yet-priced items from true contingency. If you already suspect sewer, foundation, electrical, or drainage issues, those shouldn’t live inside the contingency bucket. They should be estimated as actual scope items, with contingency sitting on top of that.

From the tax side, I’d keep the rehab costs detailed by project and component. If this is a flip, the activity is generally active business activity, and many of the acquisition, rehab, and carrying costs may need to be capitalized into the project rather than deducted immediately.

If someone is flipping consistently and profitably, I’d also evaluate whether an S-Corp makes sense based on profit level, activity volume, payroll, and reasonable compensation.

And if the investor is doing both flips and rentals, there can be a powerful planning opportunity. Depending on participation, depreciation, entity structure, and whether the rental losses are usable, those losses may sometimes offset active real estate income. In the right fact pattern, taxable income can potentially be reduced very significantly, even to zero, but it has to be planned correctly.

Feel free to DM me, I’d be happy to send over a few resources that might help with flip underwriting and downside planning.

Maurizio Pisciotta · Answer

I think this is much closer to reality than using a flat 10% across every deal.

For me, contingency should be tied to how much uncertainty is actually in the scope. Age matters, but I’d also look heavily at how much of the property is still “unknown” before closing. A 1960s house that has already had plumbing, electrical, roof, HVAC, and foundation work documented may actually be less risky than a 1990s house with poor maintenance and limited inspection access (rare, but it happens!).

I’d probably underwrite it in two layers: price the known work as accurately as possible line by line, then apply a higher contingency specifically to the parts of the project with hidden conditions (anything behind walls, under floors, in crawlspaces, structural work, old plumbing/electrical...)

With pre-1970 properties, I think the biggest mistake is assuming the contingency is just there for “a few surprises.” On some of those houses, uncovering problems is basically part of the scope. At that point, 15–20%+ doesn’t feel overly conservative...

Mark Updegraff · Answer

I’m with you on the principle. Ten percent is useful as a placeholder, not as a risk model.

With older Rochester housing, I care more about how much of the scope is still hidden. A cosmetic turn where we can see the systems is one thing. Once we’re opening walls or dealing with old electrical, plumbing, structure or years of questionable prior work, the reserve has to move.

I also separate known scope from unknown-condition reserve. Otherwise a predictable big-ticket item gets called a “surprise” and burns the same contingency you actually needed for unknowns.

Age matters, but prior workmanship and how invasive the rehab is probably matter just as much.

Flat 10% contingency is a superstition. What do you actually use?

Most Popular Reply