The three weeks you spent lifting your NOI probably bought you nothing
A borrower gets a quote back smaller than they wanted. They do the sensible thing. They go back into the operating statement, find some expense to trim, argue a vacancy assumption, and come back three weeks later with a better net operating income and a better coverage ratio.
The loan amount does not move.
This happens constantly, and the reason is that almost nobody outside a credit department knows how a loan gets sized. A lender does not run one test. They run three, independently, and then they lend the smallest of the three answers. Improving your performance on a test that was not binding changes nothing at all.
The three tests
Every permanent CRE loan is sized by some version of these, run in parallel:
The coverage test. Net operating income divided by annual debt service must clear a minimum, conventionally quoted around 1.25x. This is the DSCR, and it is the one everybody talks about. What it is and how it is calculated, in full, is here.
The leverage test. Loan divided by appraised value must sit under a cap. Call it 65 percent for a stabilised asset.
The debt yield test. Net operating income divided by the loan amount must clear a floor, often quoted around 10 percent. This one is deliberately blind to both the interest rate and the appraisal, which is exactly why lenders trust it.
Three tests. Three different loan amounts. The lender funds the lowest one, and that test is called the binding constraint.
Watch which one binds
Same building throughout. Net operating income of $600,000. Rate of 7.01 percent, which is the 10-year Treasury at 5.01 percent on 16 September plus an assumed 200 basis point spread. Thirty-year amortisation. Coverage minimum 1.25x, leverage cap 65 percent, debt yield floor 10 percent.
The only thing that changes down the table is the cap rate the property is valued at.
| Cap rate | Value | Coverage allows | Leverage allows | Debt yield allows | Binding test |
|---|---|---|---|---|---|
| 5.0% | $12,000,000 | $6,006,238 | $7,800,000 | $6,000,000 | Debt yield |
| 5.5% | $10,909,091 | $6,006,238 | $7,090,909 | $6,000,000 | Debt yield |
| 6.0% | $10,000,000 | $6,006,238 | $6,500,000 | $6,000,000 | Debt yield |
| 6.5% | $9,230,769 | $6,006,238 | $6,000,000 | $6,000,000 | Leverage |
| 7.0% | $8,571,429 | $6,006,238 | $5,571,429 | $6,000,000 | Leverage |
| 7.5% | $8,000,000 | $6,006,238 | $5,200,000 | $6,000,000 | Leverage |
| 8.0% | $7,500,000 | $6,006,238 | $4,875,000 | $6,000,000 | Leverage |
Read the middle column. The coverage test allows the same $6,006,238 in every single row, and it is never the smallest number. Not once, anywhere in that range.
At tight cap rates the debt yield binds, because the building is worth a lot relative to what it earns and the lender refuses to lend against the valuation. At wider cap rates the leverage cap binds, because the value has come down to meet the income. The DSCR sits there, quietly clearing, in every scenario.
That is the whole argument. The borrower in the first paragraph spent three weeks improving the one number that was not deciding anything.
When DSCR does bind
It is not a useless test, and the table above is one set of assumptions rather than a law. Coverage becomes the binding constraint in three situations, and they are all worth recognising:
When the rate moves against you. Coverage is the only one of the three tests that contains the interest rate. Hold everything else and push the rate up, and the coverage test is the one that tightens while the other two sit still.
When the amortisation shortens. Thirty-year amortisation allows $6,006,238 on this deal. Twenty-five-year allows $5,654,372. $351,866, decided by a schedule. We have worked that mechanism through in full, including why an interest-only structure supports around twenty percent more loan at the same coverage and the same coupon.
When the coverage minimum itself is higher. A 1.35x requirement on a less-liked asset class bites where 1.25x did not.
So the honest version is not "ignore DSCR." It is: DSCR is a rate-and-structure test wearing a property-metric costume, and you should work on it when the rate or the structure is what is hurting you, not reflexively.
The number that is doing the most work is the spread
One more run of the same arithmetic. Same $600,000 NOI, same 30-year amortisation, same 10-year Treasury at 5.01 percent. The only variable is the spread the lender adds on top.
| Spread | All-in rate | DSCR at $6M | Max loan at 1.25x |
|---|---|---|---|
| +150 bp | 6.51% | 1.32 | $6,321,854 |
| +200 bp | 7.01% | 1.25 | $6,006,238 |
| +250 bp | 7.51% | 1.19 | $5,715,107 |
| +300 bp | 8.01% | 1.13 | $5,446,165 |
$875,689 of proceeds sits between the best and worst spread in that table. Same building. Same income. Same benchmark. The difference is which lender is reading the file and how much they want the asset this quarter.
You cannot negotiate the Treasury. You can absolutely change who is competing for the deal, and that is the largest single lever on the page.
What to actually do
- Ask the lender which test is binding. It is one question, they know the answer immediately, and it tells you where any effort you spend will land. If the answer is leverage, a better operating statement will not help you and a better appraisal might.
- Negotiate the amortisation, not only the rate. The table above puts $351,866 on that conversation and almost nobody opens it.
- Run the three tests yourself before the call. The commercial mortgage calculator gives you debt service, coverage and the constant in four fields and requires no account. The other two are arithmetic you can do on a phone.
- Get more than one spread. It is the widest variable on the page and the only way to see it is to put the same file in front of several lenders at once.
The uncomfortable summary
Three tests size your loan and the lender funds the smallest answer. The one everybody optimises is frequently not the one that is binding, the one nobody negotiates is the amortisation schedule, and the one with the most money attached is a spread that varies by who is reading the file.
Find out which number is actually deciding your deal before you spend three weeks improving a different one.
Frequently asked questions
How do lenders size a commercial real estate loan?
Lenders run three tests independently and lend the smallest resulting amount: a debt service coverage test, a loan-to-value test, and a debt yield test. The test producing the lowest loan amount is called the binding constraint. Improving performance on a test that is not binding does not increase proceeds.
What is the binding constraint on a CRE loan?
The binding constraint is whichever of the coverage, leverage and debt yield tests produces the smallest loan amount. It varies with the cap rate, the interest rate and the amortisation. Borrowers can simply ask the lender which test is binding before deciding where to focus.
Does improving NOI always increase my loan amount?
No. Improving net operating income raises the amount allowed by the coverage and debt yield tests, but if the loan-to-value cap is the binding constraint, a higher NOI changes nothing until the appraised value changes with it.
Does amortisation affect DSCR?
Yes, substantially. On a $6,000,000 loan at 7.01 percent with $600,000 of net operating income, a thirty-year amortisation produces a DSCR of 1.25 while a twenty-five-year amortisation produces 1.18. At a 1.25x minimum that is a difference of roughly $352,000 in allowable proceeds, with no change to the property.
Is DSCR or debt yield more important to lenders?
Both are used, and they answer different questions. DSCR includes the interest rate and amortisation, so it moves when financing terms move. Debt yield deliberately excludes both the rate and the appraisal, which is why lenders use it as a valuation-independent check. Which one binds depends on the deal.
Find out which test is binding before you spend three weeks on the wrong one.
Underwrite the deal, build the deal memo, and put the same file in front of the lenders who write your asset class, then compare the quotes side by side.
Start for free