DSCR Is a Lender Constraint, Not Just an Analyst Metric
Debt Service Coverage Ratio comes up early in every CRE acquisition conversation, but it's worth being precise about what it is and what it isn't. DSCR is fundamentally a lender sizing constraint — a minimum threshold below which a loan request will be declined or reduced in proceeds. Acquisition analysts model DSCR not because it defines whether a deal is good, but because it defines how much debt is available, which defines how much equity you need, which defines your levered returns. Every piece of the deal structure flows downstream from that number.
The formula is straightforward: DSCR = Net Operating Income / Annual Debt Service. But the inputs to that formula — what the lender uses for NOI, what stress they apply to the interest rate, and what amortization period they assume — vary meaningfully across lender types, and an acquisition team that doesn't understand those differences will mis-model their debt proceeds in ways that alter the deal economics significantly.
Lender DSCR Floors by Asset Class and Lender Type
Not all DSCR floors are equal, and not all are published. Agency lenders — Fannie Mae and Freddie Mac multifamily programs — have the most standardized and publicly referenced thresholds. Agency multifamily typically requires a minimum 1.25x DSCR on a 30-year amortization schedule, using the lender's underwritten NOI (which is often not the borrower's underwritten NOI). Some agency programs that allow interest-only periods will test DSCR on both the IO payment and the amortizing payment to ensure the loan doesn't violate minimums when the IO period burns off.
For office and retail assets, lenders have been requiring higher floors — 1.35x to 1.45x is common from CMBS and life company lenders on post-2022 transactions, particularly where rollover risk is elevated. Industrial construction and bridge lenders pricing in risk premium may permit a 1.20x minimum on stabilized cash flow, but they'll stress-test it at a higher rate and run sensitivity on absorption assumptions.
The critical distinction for acquisition modeling: the lender's DSCR calculation uses the lender's underwritten NOI, which is not the same as the acquisition model's NOI. Lenders will apply their own vacancy assumptions (typically 5-10% even on stabilized assets), use market rents capped at some percentage of in-place rents, and apply their own management fee and reserve assumptions. An acquisition team that models a 1.30x DSCR using their own assumptions should budget for the lender to come back at 1.22x — and structure the deal capitalization to handle both scenarios.
The Interest-Only Period Distortion
Many bridge and CMBS loan structures include an initial interest-only period — commonly 12 to 36 months — during which the borrower pays only interest rather than principal plus interest. This meaningfully lowers the annual debt service figure during the IO period, which inflates the DSCR calculation.
Consider a $20 million loan at a 6.5% note rate. On an IO basis, annual debt service is $1.3M. On a 30-year amortization schedule, annual debt service on the same loan is approximately $1.52M. An NOI of $1.6M produces a 1.23x DSCR on the IO payment but only 1.05x on the amortizing payment. If the loan flips to amortizing before the property has reached its stabilized income target, the borrower is in a difficult position — either the lender covenant trips or refinance proceeds are constrained.
We're not saying IO periods are a problem to be avoided — they're a legitimate tool for assets in lease-up or value-add transition. But acquisition models should test DSCR on the amortizing payment, not just the IO payment, and should show the IC committee both figures explicitly. The IO period buys time; it doesn't eliminate the coverage requirement.
Debt Yield as the Secondary Constraint
Many lenders — particularly conduit/CMBS and larger balance sheet lenders — have added a debt yield floor alongside DSCR. Debt yield is simply NOI divided by loan amount. A loan of $20M with an NOI of $1.6M has a debt yield of 8.0%.
Debt yield doesn't depend on interest rate assumptions, which is precisely why lenders like it. In a rising-rate environment, DSCR can be manipulated by assuming shorter amortization periods or lower rate scenarios. Debt yield is a pure income-to-leverage ratio — it doesn't care about the note rate, only about how much income the property generates relative to how much debt is secured against it.
Typical debt yield floors in the 2023-2024 lending environment: 8-9% for stabilized multifamily, 9-10% for industrial and retail, 10-12% for office where lenders are still willing to transact. These floors constrain maximum loan proceeds independently of the DSCR calculation — the binding constraint determines the actual loan size.
Modeling DSCR from Day One of the Underwriting Process
One of the more common workflow errors in acquisition analysis is treating debt sizing as a back-end calculation — figuring out the purchase price and returns first, then checking whether the debt works. This sequence creates deals where the sponsor has mentally committed to a price before testing the financing constraint, which leads to creative accounting when the DSCR doesn't clear.
The better approach: run DSCR and debt yield simultaneously with the income model from the first draft of the underwriting. Set up a sensitivity table that shows maximum loan proceeds at various NOI scenarios and at both the DSCR floor and the debt yield floor. The binding constraint changes depending on the rate environment and the lender type — in a high-rate environment, DSCR tends to bind first; in a low-rate environment, debt yield may bind on lower-leverage deals.
A practical scenario illustrating this: a 120,000 SF suburban office acquisition in a mid-Atlantic market, 78% leased, asking price of $18.5M. In-place NOI: $1.42M. Stabilized NOI (pro forma at 92% occupancy): $1.85M. At current cap rates for the submarket, the acquisition model prices the deal at a 7.7% going-in cap.
Running DSCR at a 6.75% note rate and 25-year amortization, the maximum proceeds at a 1.30x in-place DSCR floor are approximately $10.2M — a 55% LTV. On a 9.5% debt yield floor using in-place NOI, maximum proceeds are $14.9M. DSCR binds first. That $4.7M difference in maximum debt proceeds is equity — and it materially changes whether the levered return clears the hurdle rate.
The NOI That Goes Into the DSCR Calculation
What NOI figure should the acquisition team use when modeling DSCR? There are at least three reasonable candidates: T-12 trailing NOI, current in-place NOI based on the rent roll, and stabilized pro forma NOI based on market assumptions. Each tells a different story, and different lenders will use different ones.
Construction and bridge lenders working on a value-add deal will typically underwrite to a stabilized NOI with a haircut — they want to see the path to stabilization and have a view on whether the stabilized NOI assumption is achievable. Permanent lenders — life companies, CMBS at stabilization — will underwrite to in-place or T-12 with their own adjustments, not to the sponsor's stabilized pro forma.
For acquisition analysis, the practice that produces the most defensible model is to run all three NOI figures through the DSCR and debt yield calculation, show all three to the IC, and use the lender-underwritten NOI (your best estimate of what the lender will produce) as the basis for the debt sizing assumption. The spread between your underwritten NOI and the lender's underwritten NOI is a known unknown that should be sized, not ignored.
Refinance Risk and the Coverage Test at Refi
DSCR matters not just at acquisition but at every refinance event over the hold period. A 3-year bridge loan that reaches maturity with NOI below the refinance lender's DSCR floor is a loan that can't be taken out on schedule — the sponsor either needs to contribute additional equity, accept higher-rate extension terms, or sell at a distressed timeline.
Modeling refinance risk requires projecting NOI at the planned refi date, assuming a plausible refinance note rate (which may be significantly different from the original rate if the rate environment has changed), and testing whether the projected DSCR clears the expected permanent lender floor. This is particularly relevant for value-add multifamily deals underwritten in 2020-2021 with bridge financing that assumed a 3-5% refi rate and are now approaching maturity in a 6-7% rate environment. The refinance math in those cases often doesn't work at the original leverage level, which is showing up in bridge extension volumes across the industry.
