A HUD underwriter looks at a project's NOI from two angles: one NOI is used against HUD's debt service coverage ratio to determine the DSCR loan amount, and one is based on the appraiser's direct capitalization value (NOI divided by the market cap rate) — which determines the LTV loan amount. The two NOIs have different standards the underwriter must follow. The DSCR NOI is meant to mirror the property's reality while staying inside HUD's underwriting rules; the appraiser's Market NOI is what a hypothetical investor would expect the property to earn. The two usually are close, but they can differ on three lines — income, vacancy, and expenses — and having a lender know how to interpret the differences can produce materially more mortgage proceeds.
Everything that follows describes the 223(f), where the loan is sized on both value and coverage — so the appraiser's Market NOI and the DSCR NOI each drive a separate loan amount. The other common programs only put a DSCR NOI in play: the 221(d)(4) is sized to cost rather than to an appraised value, so the Market NOI doesn't drive its proceeds, and the 223(a)(7) streamlined refinance uses no new appraisal at all.
How HUD Builds the DSCR NOI
For the DSCR NOI, the underwriter starts from the property's current rent collection. Normally, the method is the latest rent roll's actual rent on occupied units plus the market rent on vacant units; that sum is the Gross Residential Rent Potential, and a vacancy factor comes off it. HUD's minimum vacancy depends on how much affordability the project carries — for a market-rate project it's 7%, including collection loss. Where the property's actual vacancy runs higher, the underwriter will normally use the actual rate. The underwritten vacancy is the greater of the minimum below and the property's own rate.
| Minimum Vacancy & Collection Loss | Property Type |
|---|---|
| 3% | Properties with HAP contracts covering 90% or more of units |
| 5% | LIHTC properties meeting the minimum set-aside requirements and with attainable tax-credit rents at least 10% below market |
| 7% | Market-rate properties, or other affordable projects not meeting the 3% or 5% rules |
Worth watching. HUD has signaled interest in lowering the market-rate floor below 7% for properties with a sustained occupancy record. Nothing has been issued and nothing has changed. We will update this paper if it does. (Noted August 2026.)
HUD's 7% floor applies only to the DSCR NOI. Fannie's equivalent — the greater of 5% of gross potential rent or the gap between gross potential rent and annualized trailing-three-month collections — sits inside the underwritten net cash flow that drives both its coverage test and its value conclusion. On a clean market-rate property Fannie's 5% is the lower number, so on coverage alone Fannie underwrites less vacancy than HUD does. The difference is that Fannie's floor travels into value and HUD's does not.
Two other differences worth knowing. Fannie's treatment of declining collections is formulaic: the underwriter tests the trailing three months against the trailing six and twelve and marks income down when the decline exceeds two percent. HUD has no equivalent formula, which is not the same as no exposure: if collections are trending down, HUD can question whether the current rent roll represents sustainable income and require the Rent Potential be underwritten lower. What you can compute in advance on a Fannie deal, you have to defend on a HUD deal. And on a property still leasing up, HUD will underwrite to the revenue actually being collected once the project holds programmatic coverage for three consecutive months, where Fannie requires stabilized occupancy. From there the DSCR NOI runs through HUD's DSCR coverage mortgage amount test; our 223(f) loan sizing guide walks the full set of sizing tests.
Where the Appraiser Can Go Higher
The appraiser builds the Market NOI a little differently. Income comes from comparable projects' rents compared against the property's actual, current rents. If the property runs very low vacancy with rents below market, the appraiser can set income above the figure used in the DSCR NOI analysis.
The appraiser can also use a vacancy below the DSCR table. Where the DSCR analysis is locked at 7% for a market-rate project, the appraiser might conclude the market's vacancy is 4% and underwrite to 4%. That vacancy difference alone can add millions in mortgage proceeds if the LTV is constraining the mortgage amount, because the Market NOI drives appraised value and the LTV ceiling moves with it.
Where the Expense Lines Differ
Expenses can vary between the two NOIs the same way. The DSCR side wants expenses that reflect the property's reality; the appraiser wants expenses that reflect the marketplace, and the two get reconciled where they differ. Easy examples are an above-market management fee paid to a related entity, or a property insurance premium that runs unusually low.
The most common difference is the annual deposit to replacement reserves. The appraiser often carries $250 per unit; while the DSCR NOI must use the figure the Property Capital Needs Assessment (PCNA) sets during underwriting. Larger gaps in the total expenses can come from the cost of operating under affordability restrictions and from real estate taxes under abatements, TIFs, and similar arrangements.
What a Small Gap Is Worth
The two NOIs can finish close and still produce very different loan amounts. Say the appraiser concludes the market's management fee is $10,000 below the property's actual. At a 5% cap rate, that's roughly $160,000 of additional value — and additional LTV-constrained proceeds — off a single expense line.
None of that is gaming the rules. It's building each NOI to the standard HUD assigns it: reality on the DSCR side, the market on the appraisal side. On the right deal, that reconciliation is the difference between an adequate loan and the largest one the property can defensibly support.
