How DSCR Requirements Differ Across CRE Lender Categories

Debt Service Coverage Ratio math looks the same everywhere: net operating income divided by annual debt service, with 1.25x meaning the property throws off 25% more cash than the loan payment requires. What changes, and changes a lot, is where each lender sets the floor and what else it demands alongside that number. A sponsor who walks into a bank expecting the debt yield logic of a CMBS conduit, or approaches a life company thinking HUD's leverage is on the table, wastes weeks discovering the mismatch. This piece walks through what each lender category actually requires and why, so the target is clear before the first call gets made.

Three constraints work at once on every deal: DSCR, loan-to-value, and debt yield. Lenders size to whichever one binds first, so clearing the DSCR floor doesn't guarantee the loan amount a sponsor has in mind. That interaction matters more now than it has in years. An industry trade group for mortgage lenders counted $875 billion of commercial and multifamily mortgage debt maturing in 2026, about 17% of the $5.0 trillion outstanding, and total CRE mortgage originations were up 47% year-to-date through the third quarter of 2025. Refinancing volume is climbing straight into a market where lenders have quietly moved their floors higher. The floors are moving. Knowing which floor applies to which lender is the whole game now.

How banks underwrite DSCR and what global cash flow analysis adds to the picture

Bank portfolio lenders typically want 1.20x to 1.35x on stabilized assets, with the exact number shifting by property type and by how well the bank knows the borrower. That range sounds close to everyone else's, until the global cash flow layer gets added on top of it.

Banks, along with SBA lenders and credit unions, don't stop at the property. They pull in every income source the borrower has, salary, other real estate holdings, business income, and weigh it against every personal debt obligation on the books: the mortgage, the car loan, the credit cards. A sponsor can walk in with a property DSCR of 1.35x and still get declined because personal debt service drags the combined ratio under the bank's floor. The reverse happens too. A property that limps in at 1.15x can still get funded if the sponsor's other income covers the gap.

Documentation follows the same logic. Banks want profit and loss statements across multiple years, not just the trailing twelve months, and they'll flag volatility in that history even when the current-year DSCR clears the bar. A property that swung from 1.10x to 1.40x and back over three years draws more scrutiny than one that sat steady at a middling coverage ratio the whole time.

What banks offer in exchange for that scrutiny is flexibility no other category matches: relationship lending. A borrower with ten years of deposit and loan history at the same institution can negotiate structure in ways a first-time CMBS borrower never will. Average loan-to-value across the market is around 63.3%, per CBRE, a useful anchor for judging how conservative a given bank's terms actually are.

Banks look at the whole borrower. CMBS conduits go the other way entirely, and mostly ignore the person signing the loan.

CMBS conduit and SASB underwriting: where the property does all the qualifying work

Conduit lenders set the floor at 1.20x to 1.25x, computed on stabilized NOI divided by amortizing debt service at the note rate. Replacement reserves and TI/LC reserves usually come out of NOI before that ratio gets calculated, which quietly tightens the real coverage below the headline number.

Single-asset, single-borrower deals run higher, 1.25x to 1.35x, though some recent SASB pools have negotiated down to 1.20x as convention loosens at the institutional end of the market. The defining feature of conduit and SASB underwriting isn't the number, though. It's what the number replaces. Unlike agency, bank, and HUD programs, CMBS lenders put little weight on borrower net worth or real estate track record. The property's cash flow coverage carries the underwriting almost by itself.

Hospitality breaks from that pattern hard. Flagged hotels typically need 1.40x, and unflagged hospitality often needs at least 1.50x, because income volatility from competitive supply and seasonality makes hotel cash flow a different animal from apartment or office rent rolls.

Debt yield can override the DSCR math. CMBS conduits generally require 8.0% to 10.0% debt yield, and hotel lenders often want 10% to 12%. A hotel generating $1,000,000 of NOI against a 12% floor caps out near $8.3 million of debt, full stop, regardless of what the DSCR calculation would otherwise permit. Sponsors who don't ask about the debt yield floor early find that out at term sheet, which is the wrong time to find it out.

The broader market context: 73% of 2025 CMBS conduit maturities paid off at or before maturity, according to Principal Real Estate's analysis. SASB has been the busiest sub-segment through 2024 and 2025, and CRE CLO issuance hit $30.5 billion in 2025, a notably elevated total. Inside that CLO universe, the share of refinance loans climbed to 68.2% in 2024, up from 37.9% the year before. That jump says something plain: a lot of borrowers are using the CLO market to work their way out of coverage that's gotten too thin for anyone else to touch.

Life insurance company standards: the highest floors paired with the lowest rates

Life companies want 1.30x to 1.50x, with 1.25x as the minimum floor, and they calculate it on current income only. No pro forma bumps, no projected rent growth folded into the numerator.

The accompanying requirements make the category self-selecting before DSCR even enters the conversation: 50% to 65% LTV, stabilized occupancy at 90% or better, strong locations in primary and secondary markets, and sponsors with a track record. Value-add plays don't get in the door.

What a sponsor gets for clearing those bars is pricing nobody else in the stack can match. Life companies typically price 25 to 75 basis points below banks and 50 to 150 basis points below CMBS on comparable deals. In early 2026, top-quality assets are pricing in the 5.50% to 6.25% range through this channel. Life companies don't run public-facing loan programs, so access is the catch, and reaching one usually means going through a commercial mortgage broker who already has the relationship.

The conservative DSCR buffer isn't arbitrary caution. Life insurers hold these loans on balance sheet for long terms, matching long-duration liabilities to long-duration assets, so the extra coverage functions as an actuarial cushion against a decade or more of income variance nobody can forecast precisely today.

Agency lenders (Fannie Mae and Freddie Mac): multifamily-specific floors with a stressed-rate wrinkle

Fannie Mae typically wants 1.25x. Freddie Mac allows minimums as low as 1.20x depending on market. Both apply only to multifamily, full stop, so a sponsor with an office or retail deal can skip this section.

The number that actually decides most agency deals isn't the base DSCR floor. It's the stressed-rate test layered on top of it, where the agency runs the loan against a hypothetical higher interest rate before signing off. That test binds tighter than the nominal floor more often than sponsors expect. A deal that clears 1.20x comfortably at the note rate can still fail once the agency applies its stress rate, and that failure appears late in the process if nobody ran the math beforehand.

Debt yield adds a second constraint, agency multifamily programs typically require 7.5% to 9.0%, which can bind before DSCR does in a market where rates sit high and cap rates stay flat, a combination that's described 2024 through 2026 reasonably well.

Sponsors put up with the extra hurdle because the payoff includes the lowest rates and longest terms available anywhere in the multifamily stack, at scale. Agencies also weigh borrower net worth and real estate experience, unlike CMBS, so the deal and the sponsor both have to clear the bar. The practical move is to run both DSCR calculations, base and stressed, before submitting anything. Finding out about a stress-test failure at the term sheet stage burns time that could have gone toward a lender who was never going to be a fit anyway.

HUD/FHA multifamily programs: the lowest DSCR floors in the market and why the structure makes it possible

HUD's 221(d)(4) program, under Mortgagee Letter 2025-03, sets the floor at 1.15x for market-rate properties and 1.11x for affordable housing and rental assistance properties. Those are the lowest DSCR requirements in commercial real estate lending, by a wide margin, and they come paired with leverage no conventional lender offers: market-rate deals at 1.15x DSCR can reach 87% LTV, and affordable deals at 1.11x can reach 90%.

That math produces a loophole not because it hides one but because HUD amortization runs 35 to 40 years, far longer than the shorter schedules typical of conventional lenders, which cuts the annual debt service payment enough that thinner NOI can still cover it. It's a structural feature of the program, built in on purpose.

Policy has moved in the sponsor's favor recently, too. Following advocacy from an industry trade group for mortgage lenders, HUD issued two mortgagee letters in January 2025 that rolled back earlier tightening on DSCR and LTV for some FHA multifamily programs. Separately, mortgage insurance premiums got standardized at 0.25% across FHA multifamily programs for applications submitted or amended on or after October 1, 2025, with Section 232 residential care facilities carved out under their own premium schedule.

HUD tends to enter the conversation at a specific moment: when a deal can't clear agency DSCR requirements, or when the maximum LTV Fannie or Freddie will offer still leaves a gap in the capital stack. At that point, HUD's combination of a low floor and high leverage often becomes the only permanent financing path left standing.

SBA 7(a) and 504 programs: government-backed floors that also require global cash flow

SBA 504 and 7(a) loans run DSCR floors of 1.15x to 1.25x, lower than conventional lending across the board, which lines up with the programs' stated mission of expanding access to capital for borrowers who wouldn't otherwise get funded.

That access comes with the same global cash flow test banks and credit unions run. SBA lenders look at the borrower's full financial picture, not just the property's, so a sponsor with strong in-place NOI but heavy personal debt can still get turned down. The programs exist mainly for acquisition, renovation, or construction financing with long repayment terms, and they're built for owner-operators, not investment sponsors chasing a cap rate spread.

How these programs are designed states the trade-off: lower DSCR floors widen access for borrowers with thinner cash flow, and that same widening raises the risk sitting on the lender's book. SBA isn't a fallback option for any commercial deal that can't get done elsewhere. It's built for owner-occupied or owner-operated properties specifically, and trying to force a pure investment deal through an SBA underwriter creates eligibility problems that don't go away with better paperwork.

Bridge lenders and debt funds: the lowest floors in the market, with a different underwriting logic entirely

Bridge lenders and debt funds generally run lower DSCR floors than conventional lenders, and some will accept minimal or no in-place coverage when the business plan shows a credible path to stabilization. On paper, that's the loosest DSCR standard anywhere in commercial lending.

The number stops mattering much on construction and lease-up deals, where current DSCR often doesn't exist at all because there's no in-place income to measure yet. Underwriting shifts instead to the projected stabilized picture and whether the business plan holds together: can this sponsor actually finish the project and lease it up on the timeline the pro forma assumes?

That's the real underwriting question at every debt fund: not what the ratio says today, but whether the business plan, the sponsor's execution history, and the exit assumptions hold together. CRE debt funds captured 14% of total lending in the first half of 2025, well above their roughly 9% post-2008 average, as banks kept rationalizing balance-sheet CRE exposure, according to Principal Real Estate's analysis citing MBA data.

Sponsors financing at thin or sub-1.0x coverage through bridge debt are operating in territory where any income shortfall, a vacancy, a rent concession, a slower lease-up than modeled, can quickly attract lender attention. Bridge capital is built for a specific window in a deal's life. Holding it past that window, or using it beyond the phase it was priced for, is exactly where the risk piles up.

How property type moves the DSCR target across all lender categories

Multifamily is at the low end almost everywhere, 1.20x to 1.25x across most lender categories, because income from dozens or hundreds of individual leases diversifies risk in a way single-tenant assets can't.

Office and retail push higher, often 1.30x or above, to build in a cushion against rollover risk and the kind of submarket softness that's hit both property types unevenly over the past several years. Hotels split further still: flagged hotels typically need 1.40x, and unflagged hospitality often needs at least 1.50x, driven by the volatility that competition, seasonality, and economic sensitivity bake into hotel income. Self-storage can also draw elevated requirements from lenders, because month-to-month leases mean occupancy can swing fast.

NNN-leased properties with national tenants on long-term leases go the other direction, where the stability of a credit tenant on a long-term lease removes much of the income volatility the DSCR floor exists to guard against in the first place.

Geography adds its own layer on top of asset type. In New York City, Local Law 97 compliance costs are cutting into NOI on older multifamily and office buildings, and lenders are responding by pushing DSCR requirements above 1.30x to compensate. In Los Angeles, rising insurance premiums are squeezing NOI on retail and multifamily assets, nudging lender requirements upward for the same reason. In Austin, rent growth has cooled sharply after the rapid run-up of prior years, and lenders have shifted to more conservative rent assumptions, with DSCR thresholds typically landing between 1.25x and 1.30x.

A handful of standardized inputs that almost every lender applies before the DSCR number ever gets calculated produce all of it: market vacancy assumptions in the rough range of 5% to 10%, and management fee assumptions layered into the expense side of NOI. Those inputs shape the denominator and numerator alike, long before a sponsor ever sees the ratio a lender is willing to quote.