DSCR lending has become one of the largest segments of the non-QM market because it solves a real problem — a professional real-estate investor with ten properties and a tax return engineered to show very little personal income can't qualify on debt-to-income the way a W-2 borrower does, but the properties themselves throw off cash, and a debt-service-coverage loan underwrites the cash rather than the borrower. I came at this from the consumer-credit side, building Pagaya's origination platform at roughly $10 billion a year of flow, and now at Fundable we build the AI platform that fund and asset managers use to underwrite and diligence exactly these loans. This post explains how a DSCR loan is actually underwritten, where the math quietly breaks, and why two loans that both show a 1.25 DSCR can carry very different risk.
What the ratio measures
The debt-service-coverage ratio is the property's net operating income divided by its debt service, and in residential DSCR lending it's usually simplified to monthly rent divided by the monthly payment of principal, interest, taxes, insurance, and any association dues — the PITIA. A ratio of 1.0 means the rent exactly covers the payment, above 1.0 means the property generates a cushion, and below 1.0 means the rent doesn't fully cover the debt and the borrower is feeding the property out of pocket. Most lenders set a floor somewhere around 1.0 to 1.25 for their best pricing, will go down to break-even with pricing adjustments, and offer "no-ratio" or sub-1.0 programs at progressively worse terms. The appeal to the originator is that the documentation burden collapses — instead of tax returns, pay stubs, and a debt-to-income calculation, you need a rent figure, a payment, and an appraisal — which is precisely what makes DSCR fast to originate and fast to get wrong.
Where the rent number comes from — and where it breaks
The entire loan hinges on the rent figure, and the rent figure is the softest input in the file. On a tenant-occupied property the underwriter uses the lower of the actual lease and the appraiser's market-rent estimate, which is reasonably defensible. The trouble starts on a vacant property or a short-term rental, where there is no in-place lease and the rent is an estimate — the appraiser's opinion of market rent on a residential rent schedule, which can be generous, stale, or based on comparables that don't reflect the subject's actual condition. A property underwritten at a 1.2 DSCR on an optimistic market-rent estimate can be a sub-1.0 property the day a real tenant signs a real lease, and the loan looks identical on the tape either way.
Short-term-rental DSCR is its own hazard, because the income used to qualify is annualized from seasonal, volatile, platform-dependent revenue, and a regulatory change — a city restricting short-term rentals, which has happened in market after market — can take the rentable income to near zero overnight on a property that was underwritten at a comfortable coverage ratio. The ratio didn't change. The world underneath it did.
Why two loans at the same DSCR aren't the same risk
This is the part that statistical, tape-level analysis misses. Imagine two loans, both at a 1.25 DSCR, both at 70% LTV, both on single-family rentals. The first is a long-term tenant on a two-year lease in a stable metro, with the rent corroborated by twelve months of bank deposits and a market-rent estimate that lines up with the lease. The second is a vacant short-term rental in a beach market, qualified on an annualized projection, with no in-place tenant and a city council that just put a short-term-rental ordinance on the agenda. Same ratio, radically different probability of the rent actually showing up. The DSCR is a single number that compresses tenant quality, lease durability, income verifiability, market regulation, and property condition into one figure, and the entire job of good DSCR underwriting is reconstructing the things the number threw away.
That reconstruction is exactly what gets skipped at scale, because it's labor. Pulling the bank statements to corroborate that the rent is actually being deposited, checking whether the lease term outlasts the loan's near-term horizon, verifying the property isn't in a jurisdiction that just restricted the use it was underwritten on — each of those is a manual step, and when an originator is pushing volume or a buyer is diligencing a thousand-loan pool, the manual steps are the first thing to go.
The fields a DSCR diligence pass should reconcile
Rent, against evidence. Corroborate the qualifying rent against an actual lease and against deposit activity in the bank statements, rather than accepting the appraiser's market-rent estimate at face value, especially on vacant or short-term-rental properties.
PITIA, against the real obligations. Confirm the taxes and insurance in the payment are current figures, not origination-era estimates, because a tax reassessment or an insurance-premium spike — both common in the markets where DSCR volume concentrates — can quietly push a 1.1 DSCR below break-even.
Occupancy and use, against the public record and the platform. Verify the property is being used the way it was underwritten, since a property qualified as a long-term rental and operated as a short-term one carries different income volatility and, increasingly, different legal risk.
Borrower experience and the rest of the portfolio. A DSCR borrower with twenty properties cross-collateralized across several lenders is a different credit than a first-time investor at the same ratio, and the tape rarely shows the rest of the borrower's book.
What this segment does in a downturn
DSCR is where the surprises in non-QM tend to concentrate, precisely because the underwriting was already non-standard and the income is a property-level estimate rather than a verified borrower wage. A DSCR loan on a short-term rental where the underlying income evaporates can move from current to seriously delinquent in two months — far faster than a prime residential loan secured by a borrower's primary residence and W-2 income, where the borrower will fight to keep the home. Because of that, the secondary-market bid on DSCR paper is tighter, and the buyer prices in higher loss severity from day one. The way to be on the right side of that pricing, whether you originate or buy, is to know which loans in the pool are the durable-income loans and which are the optimistic-estimate loans — a distinction the DSCR field on the tape will never tell you, and one that only shows up when you reconcile the ratio back to the evidence.
The takeaway
The debt-service-coverage ratio is a genuinely useful underwriting tool, because it lets a lender finance professional real-estate investors quickly without drowning in personal-income documentation, but its compression is also its danger — it reduces a tangle of tenant quality, lease durability, regulatory exposure, and verifiable income to a single decimal, and that decimal looks identical whether the income is rock-solid or a hopeful projection. Good DSCR underwriting, and good DSCR diligence, is the discipline of putting back what the ratio left out, loan by loan. When that work is done by hand, it's the first thing cut under volume pressure — which is the entire reason the surprises cluster in this segment, and the reason we built Fundable to do that reconciliation on every loan instead of a sample.