If you have a loan maturing this year, you have probably been told to be patient. Rates will ease, the refinance will pencil, the gap closes.
Rates did ease. The gap did not close.
In Q2 2026 the average commercial mortgage rate came in at 5.7%, down from 5.9% a year earlier. Spreads tightened too, to 204 basis points on commercial loans and 162 on multifamily, both narrower than the year before. Lenders are competing on price.
And over exactly that period, the average deal was underwritten to a 1.43x debt service coverage ratio, up from 1.34x. Average debt yield rose from 9.7% to 10.2%. Average LTV slipped from 60.8% to 59.6%, and multifamily LTV from 65.8% to 63.3%.
Read those two paragraphs together and you have the whole 2026 refinancing story. The price of debt fell. The amount of debt fell further.
The number that actually sizes your loan
Most borrowers think of a commercial loan as a rate applied to a value. Lenders do not size loans that way.
A lender runs your property through two or three tests and lends the smallest number that comes out.
The DSCR test asks how much annual debt service your net operating income can cover at the required ratio. At a 1.43x coverage requirement, a property with $500,000 of NOI can carry roughly $350,000 of annual debt service. Convert that to a loan at today's rate and amortization and you have one ceiling.
The debt yield test asks what your NOI is as a percentage of the loan, with no reference to rate or amortization at all. At a 10.2% debt yield requirement, that same $500,000 of NOI supports roughly $4.9 million of debt. Full stop. It does not matter what the rate is.
The LTV test caps the loan against appraised value.
Your loan is the lowest of the three. And notice what the debt yield test is: a number that a rate cut cannot move. That is the point of it. Lenders adopted debt yield precisely because it strips out the rate and the amortization schedule, the two variables a borrower can dress up.
Run the actual math
Take a stabilized property throwing off $500,000 of NOI. Underwrite it twice, once to last year's average standards and once to this year's, using CBRE's reported averages both times.
A year ago. Debt yield at 9.7% gives you about $5.15 million. The DSCR test at 1.34x on a 5.9% rate over 30 years gives you about $5.24 million. The lower number binds, so you are looking at roughly $5.15 million.
Today. Debt yield at 10.2% gives you about $4.90 million. The DSCR test at 1.43x on a better 5.7% rate gives you about $5.02 million. The lower number binds again, so you are looking at roughly $4.90 million.
Same property. Same income. A rate that improved by 20 basis points. And roughly $250,000 less in proceeds, about 5%.
Now put a maturing loan on the other side of it. If you are refinancing $5.15 million of existing debt, you are not refinancing. You are writing a cheque for a quarter of a million dollars, or finding someone who will take the gap.
(Illustrative, using CBRE's Q2 2026 and Q2 2025 reported averages applied to a hypothetical property. Your lender's actual tests will differ.)
Why the coverage test moved when rates did not
This is not lenders being difficult. It is lenders pricing a risk that showed up in their own portfolios.
CMBS delinquency reached 7.86% in July 2026, up 51 basis points in a single month, on $6.0 billion of newly delinquent balances. Two thirds of that new delinquency was non-performing matured balloons. And Trepp's read on why those loans transferred is the sentence the whole market should be reading: most of them moved because of refinancing challenges rather than property performance.
Those are not broken buildings. Those are working buildings that could not clear the refinancing test. A lender watching that happen across a portfolio does not respond by cutting the coverage requirement.
The rate environment is not helping either. The Fed held at 3.50% to 3.75% on July 29, 2026, and the vote was 9 to 3 with all three dissenters wanting a hike, not a cut. The 10-year has risen roughly 70 basis points since the outbreak of the Iran war, touching 4.74%. Anyone still underwriting to a rate-cut assumption is underwriting to a scenario the Fed's own committee is arguing against.
Multifamily is the sharpest version of this
If you own apartments, the squeeze is tighter than the averages suggest, and it is worth understanding why.
Operations are fine. Better than fine, actually. Net absorption hit 167,500 units in Q2 2026, vacancy fell to 4.3%, and absorption outpaced new supply in 45 markets, up from 3 in Q4 2025. That is a genuine inflection.
What is not fine is rent growth as a refinancing input. Advertised rents rose 0.2% year over year in July 2026. Your NOI is roughly where it was. The coverage test is not.
Meanwhile multifamily is carrying the most aggressive underwriting in this year's CMBS vintage: the highest LTV at 68.4% and the lowest DSCR at 1.33x of any major property type. Trepp's own assessment is that this is "not a performance problem today, but it is the smallest margin for error at refinancing if rates hold." Multifamily CMBS delinquency rose 46 basis points in July to 7.69%.
And the timing is now. MSCI expects 60% of 2021 and 2022 vintage apartment loans to mature in the second half of 2026. Those loans were written at a leverage point that does not exist anymore.
Occupied, performing, rent-collecting apartment buildings are going to fail refinancing tests this year. Not because anything went wrong with the asset. Because the test changed underneath it.
So what do you actually do
Three things, in order.
Underwrite your own deal to today's tests before a lender does. Take your trailing twelve NOI, divide it by 0.102, and look at that number honestly. Then run the coverage test at 1.43x. The lower of those two is roughly what the market will lend you. If it is below your existing balance, you have a gap, and you have it today whether or not anyone has told you.
Solve the gap as a capital question, not a rate question. More equity, a partner, mezzanine, preferred, a partial paydown, or a smaller loan against a partial sale. Every one of those takes months to arrange. Waiting for a rate cut to close a debt yield gap is waiting for the wrong variable, because debt yield does not contain the rate.
Then widen the search, because the tests are not uniform. Alternative lenders wrote 38% of non-agency closings in Q2 2026, up from 34% a year earlier, and banks 30%, up from 24%. Banks eased standards on nonfarm nonresidential, multifamily and construction lending in the Q2 Senior Loan Officer survey. There is real dispersion in who will stretch and on what. Calling the three lenders you already know is the most common unforced error in this market, and it is more expensive in a year when the binding constraint varies by lender.
The part nobody says out loud
A property that only refinances at 2021 leverage is not a financing problem. It is a capital structure that has not been marked to the present.
That is uncomfortable, and it is also the most useful thing an owner can hear right now, because it moves the conversation from waiting to acting. Capital is available. Originations rose 16% year over year in Q2 2026, with retail up 61% and office up 47%. Spreads are tighter than a year ago. Banks are lending again.
The money is there. It is just sized to a coverage test, not to your maturity date.
FAQ
What DSCR do commercial lenders require in 2026?
Lenders underwrote to an average debt service coverage ratio of 1.43x in Q2 2026, up from 1.34x a year earlier, according to CBRE. Requirements vary by property type, lender and asset quality, and stabilized multifamily often prices tighter, but 1.25x is now closer to a floor than a norm on most commercial product.
What is debt yield and why does it matter more than the rate?
Debt yield is net operating income divided by the loan amount, expressed as a percentage. It contains no rate and no amortization schedule, so it does not improve when rates fall. Average debt yield rose to 10.2% in Q2 2026 from 9.7% a year earlier. Because lenders lend the lowest number their tests produce, debt yield frequently binds before the DSCR or LTV test does.
How much can I borrow against my commercial property?
Lenders run a DSCR test, a debt yield test and an LTV test, and lend the smallest result. On Q2 2026 averages, a property with $500,000 of NOI supports roughly $4.9 million on a 10.2% debt yield test and roughly $5.0 million on a 1.43x coverage test at a 5.7% rate over 30 years, so the debt yield test binds at about $4.9 million.
Why did my loan get smaller when rates went down?
Because rate is only one input. In Q2 2026 average commercial mortgage rates fell to 5.7% from 5.9%, but the average underwritten DSCR rose to 1.43x from 1.34x, debt yield to 10.2% from 9.7%, and LTV fell to 59.6% from 60.8%. Tighter coverage and debt yield requirements reduced proceeds by more than the lower rate increased them.
Will rate cuts fix my 2026 refinancing gap?
Not reliably, for two reasons. The debt yield test contains no rate, so it does not move when rates do. And the rate path is not assured: the Fed held at 3.50% to 3.75% on July 29, 2026, with three of twelve voters dissenting in favour of a hike rather than a cut.
Is multifamily harder to refinance than other property types in 2026?
In several respects, yes. Multifamily carries the highest LTV at 68.4% and the lowest DSCR at 1.33x in this year's CMBS vintage, multifamily CMBS delinquency rose 46 basis points in July 2026 to 7.69%, and MSCI expects 60% of 2021 and 2022 vintage apartment loans to mature in the second half of 2026. Operations have improved, with vacancy at 4.3%, but rent growth of 0.2% year over year has not lifted NOI enough to offset tighter coverage tests.