Commercial loan sizing is not a single calculation. It is three tests run in parallel, and the lender funds the smallest number the three produce. That is the mechanic borrowers miss most often. They optimize one ratio, watch it clear comfortably, and assume the loan is theirs, only to have a second test cap the proceeds far below what the first allowed. A lender sizes to the most restrictive of loan-to-value, debt service coverage ratio, and debt yield. Whichever constraint yields the smallest supportable loan wins, and which one wins changes with the interest rate, the appraised value, and the loan structure.
The thesis: the three ratios are not redundant. Each measures a different failure mode, and the deal is only as large as its weakest test allows.
Key Takeaways
Commercial loan sizing computes a maximum loan under each of three tests separately, then funds the minimum of the three: LTV-based, DSCR-based, and debt-yield-based.
LTV sizes to appraised value, DSCR sizes to the current payment, and debt yield sizes to NOI alone. Only one binds on any given deal, and it is the smallest of the three outputs.
Most 2026 commercial lenders require a minimum DSCR of 1.20x to 1.35x, maximum LTV of 65 to 80 percent, and a minimum debt yield of roughly 8 to 12 percent, per lender guidance surveyed by CRE Daily and Commercial Loan Direct.
Which test binds moves with the rate. In low-rate markets debt yield usually binds; as rates rise DSCR tightens and can take over; LTV binds when values are aggressive relative to income.
A borrower who sizes only to the ratio that flatters the deal will consistently over-estimate proceeds. The lender's number is the floor of three, not the ceiling of one.
How do lenders size a commercial loan across three ratios?
Lenders size a commercial loan by calculating a maximum loan amount under each of three independent tests, then lending the smallest of the three. The LTV test caps the loan against appraised value, the DSCR test caps it against the current debt payment, and the debt yield test caps it against NOI. The binding constraint is whichever produces the lowest number.
Wall Street Prep and CRE Daily both describe the same underwriting sequence: maximum supportable loan equals the minimum of the LTV-based maximum, the DSCR-based maximum, and the debt-yield-based maximum. The three are not averaged and not blended. They are a floor function. A deal that clears LTV at 20 million dollars, DSCR at 18 million, and debt yield at 15 million is a 15 million dollar loan, full stop.
Each test isolates a different question. LTV asks how much equity cushion exists if the property must be sold. DSCR asks whether current cash flow covers the current payment. Debt yield asks what unleveraged return the loan balance earns on NOI, independent of rate and amortization. See the loan-to-value ratio, debt service coverage ratio, and debt yield glossary entries for each formula in full.
Test | Formula for maximum loan | What it protects against |
LTV | Appraised value x max LTV | Value decline; too little equity below the loan |
DSCR | Annual NOI / (min DSCR x debt constant) | Cash flow failing to cover the payment |
Debt yield | NOI / min debt yield | Weak asset income relative to loan size |
The debt constant in the DSCR formula is annual debt service divided by loan balance, so it folds the interest rate and amortization schedule into one number. See the debt constant entry for the mechanics. That single term is why DSCR-based proceeds swing so hard when rates move, while debt yield does not move at all.
Which constraint binds, and why does it change with interest rates?
The binding constraint is simply the test that produces the smallest loan on a given deal, and it shifts with rates because only two of the three ratios respond to rate changes. Debt yield ignores rates entirely. LTV responds only through value. DSCR moves directly with the payment, so as rates rise DSCR tightens and can become the constraint that caps proceeds.
The pattern reported by Tactica RES and others runs like this. In a low-rate market, cheap debt makes DSCR easy to satisfy, so debt yield usually becomes the binding constraint. As rates rise, the debt constant climbs, DSCR-based proceeds fall, and DSCR can overtake debt yield as the tightest test. LTV binds separately, when the appraised value is high relative to income, letting a borrower push value-based proceeds past what the income tests support.
A worked example makes the floor function concrete. Take a stabilized property with 1,000,000 dollars of NOI, an appraised value of 15,000,000 dollars, and a lender that requires a maximum LTV of 70 percent, a minimum DSCR of 1.25x, and a minimum debt yield of 10 percent. Assume a 7 percent interest-only loan, so the debt constant equals the 7 percent rate.
Test | Calculation | Maximum loan |
LTV | 15,000,000 x 0.70 | 10,500,000 |
DSCR | 1,000,000 / (1.25 x 0.07) | 11,428,571 |
Debt yield | 1,000,000 / 0.10 | 10,000,000 |
Here debt yield binds at 10,000,000 dollars, the smallest of the three, so that is the loan. Verify the DSCR line: annual debt service on a 10,000,000 dollar interest-only loan at 7 percent is 700,000 dollars, and 1,000,000 divided by 700,000 is 1.43x, comfortably above the 1.25x minimum. DSCR is not the constraint at this rate. Now push the rate to 9 percent. The DSCR-based maximum falls to 1,000,000 divided by (1.25 x 0.09), or 8,888,889 dollars, and DSCR takes over as the binding test, cutting proceeds below the debt yield ceiling. Same property, same value, same NOI, and the constraint moved purely because the rate moved.
As one lender framing puts it: the borrower optimizes the ratio they understand, and the lender funds the one they forgot.
What do lenders require for DSCR, LTV, and debt yield in 2026?
In 2026, most commercial lenders require a minimum DSCR of roughly 1.20x to 1.35x, a maximum LTV of 65 to 80 percent depending on asset and sponsor, and a minimum debt yield in the range of 8 to 12 percent. These are underwriting floors, not targets. A stronger asset earns the low end of each range; a transitional or higher-risk deal is pushed to the strict end.
Per DSCR guidance compiled by Commercial Loan Direct and lender memos surveyed for 2026, thresholds have tightened. Reporting notes that at least one top-tier lender raised its minimum stabilized DSCR from 1.20x to 1.25x entering 2026, with transitional deals pushed to 1.30x or higher. Terrydale Capital and CRE Daily place typical maximum LTV at 65 to 80 percent, with investment property often landing at 75 to 80 percent. Debt yield floors cluster near 10 percent for stabilized assets, higher for hospitality, where lenders commonly require 10 to 12 percent.
Ratio | Common 2026 range | Direction under stress |
DSCR | 1.20x to 1.35x minimum | Rising; some lenders moved 1.20x to 1.25x |
LTV | 65 to 80 percent maximum | Falling for transitional and higher-risk assets |
Debt yield | 8 to 12 percent minimum | Rising; often binds in refinances |
The recalibration runs deeper than the ratios themselves. Underwriters are also haircutting the NOI that feeds all three tests, applying conservative vacancy, management, and reserve assumptions rather than accepting a broker's stabilized number. A lower NOI shrinks every test at once, because NOI is the numerator in DSCR and debt yield and drives the value that sets LTV. This is why the same deal that sized cleanly in 2021 can come back materially smaller in 2026 even before the rate change is counted. For more on how proceeds compress at refinance, see why debt yield beats DSCR.
Frequently Asked Questions
Do lenders average DSCR, LTV, and debt yield to size a loan?
No. Lenders do not average the three tests. They calculate a maximum loan under each one separately and then fund the smallest of the three outputs. The result is a floor function, so a deal that clears two tests generously is still capped by whichever test produces the lowest number.
Which loan sizing constraint usually binds?
The binding constraint depends on rates and value. In low-rate markets debt yield usually binds because cheap debt makes DSCR easy to clear. As rates rise, the debt payment climbs and DSCR can become the tightest test. LTV binds separately when appraised value is high relative to the property's income.
Can maximizing NOI increase every part of the loan size?
Yes. Because NOI is the numerator in both DSCR and debt yield and drives the appraised value behind LTV, a higher, defensible NOI lifts all three maximum-loan calculations at once. That is why lenders scrutinize NOI so heavily and why an inflated NOI is the fastest way to have proceeds cut at closing.
Conclusion
Commercial loan sizing rewards the borrower who underwrites all three tests, not the one who optimizes the ratio that reads best. LTV, DSCR, and debt yield each measure a different failure mode, and the lender funds the smallest loan the three allow. The constraint that binds is not fixed. It moves with the rate, the value, and the structure, which means a deal sized comfortably under one test can be capped hard by another the borrower never modeled. The operator lesson is to run the floor function yourself before the lender does. Compute the maximum loan under each ratio, take the minimum, and size the equity to that number. The gap between the ratio you optimized and the one that binds is exactly the gap that shows up as a surprise at closing.