CRE Financial Modeling Inputs Lenders Scrutinize Most
Lenders are rebuilding NOI from scratch and stress-testing every assumption.
Sponsors and brokers entering the 2026 capital market are working inside an underwriting climate that bears little resemblance to the one that produced many of the loans now coming due. Lenders this year are prioritizing credit quality, stable cash flow, and execution certainty over volume-driven origination, a direct response to the optimistic rent-growth assumptions and thin cash-flow cushions baked into deals from 2021. The scale of the problem is forcing the issue: $875 billion in commercial and multifamily mortgage debt matures in 2026, and that volume of refinancing pressure concentrates lender attention on a narrower question than it used to ask, which deals actually hold up under current conditions. Credit committees are emphasizing downside protection and long-term asset performance rather than chasing origination targets, and that priority becomes visible fast at the term sheet stage: borrowers who lack transparency, lean on unrealistic assumptions, or can't point to relevant experience are seeing delays or reduced leverage. None of that is arbitrary. It traces back to a small set of inputs in the financial model that lenders now interrogate line by line, and how well a sponsor can explain each one shapes whether a deal closes on schedule or stalls in the credit room.
Why Lenders Do Not Accept Your NOI at Face Value
Every number a lender produces, the loan amount, the rate, the required equity, starts from one figure: net operating income. It is the single most consequential input in CRE underwriting because it drives valuation, loan sizing, and the return the deal can plausibly deliver. What sponsors miss is that lenders do not take the NOI on the offering memo and move forward with it. They rebuild it from the ground up, and the version that comes out the other side is frequently smaller than the version that went in.
That gap matters because it isn't cosmetic. Specific adjustments that lenders apply as a matter of course produce the gap: real estate taxes trued up to current assessments rather than trailing bills, insurance repriced to the market a property actually faces today, management fees calculated at market rates rather than whatever a sponsor's affiliate happens to charge, replacement reserves added whether or not the sponsor budgeted for them, vacancy built in even where current occupancy looks strong, and any income or expense line that looks one-time or non-recurring stripped out. Each of those adjustments is a place where a sponsor's pro forma and a lender's credit memo diverge, and sponsors routinely understate or omit the very items a lender treats as non-negotiable.
The broader shift reinforces this. Trended rent growth, the practice of underwriting to where rents are expected to go rather than where they sit today, has largely disappeared from serious credit conversations. Underwriting now anchors to current, verifiable cash flow, and lenders have gotten more exacting about expenses too, scrutinizing taxes, insurance, and payroll with particular care and declining to simply roll forward historical averages when forward-looking costs are clearly higher. For a sponsor, the useful move is to anticipate the adjustment: building a model that shows the lender's likely adjusted NOI next to the sponsor's own figure, with every adjustment spelled out, functions as a credibility signal. When a credit team sees its own math already done, it tends to trust the rest of the package more.
DSCR at today's coupons, not yesterday's rates
Once the lender has its own NOI, the next question is whether that cash flow covers the debt, and that question, not the loan-to-value ratio, is what sizes most deals in 2026. Debt service coverage has displaced LTV as the primary constraint, with leverage now treated as a secondary check. The formula itself hasn't changed: DSCR is lender-underwritten NOI divided by annual debt service. What has changed is the denominator. Lenders are running that calculation against today's coupon, not the rate a sponsor locked in years ago, and they're frequently layering in flat or declining rent assumptions on top of that higher cost of debt. If the resulting ratio doesn't clear the lender's minimum, the refinance conversation tends to stop there, and it doesn't move toward a structure negotiation.
Minimum DSCR thresholds aren't uniform across property types. Agency multifamily lenders tend to underwrite to looser minimums, because residential income is relatively stable, but office and hospitality lenders demand more cushion to cover the income volatility those asset classes carry. Deals originated in 2021 face the sharpest version of this test: banks and CMBS shops have raised their DSCR benchmarks since then and are far less willing to give credit for future rent growth, so a loan that comfortably cleared its coverage test at origination can fail the same test today purely because the numerator shrank and the denominator grew.
Credit teams also run a second check alongside DSCR, measured independently of interest rates: debt yield, which divides NOI by the loan amount itself. It matters most when a deal has an interest-only period or an unusually favorable coupon that makes the DSCR number look stronger than the underlying economics support. Debt yield strips out the rate assumption entirely and asks a more basic question, whether the loan proceeds are justified by the property's earning power on their own terms. A sponsor who shows up with a model built on projected rent growth should expect the lender to re-run it at flat rents before the meeting is over. If you arrive with that stress-tested version already prepared, a potential retrade turns into a straightforward conversation.
How lenders read a rent roll
The rent roll is often the first document a lender reads closely, and a high occupancy percentage at the top of the page settles very little. Lenders prioritize in-place performance over any projection. The real work of the review happens in the lease-by-lease detail: who's paying what, for how long, and what happens if any one of them leaves.
Concentration risk is the first thing a credit team looks for. When a single tenant carries a large share of a property's revenue and its lease expires in the near term, a lender doesn't read that as a stable income source. It reads as a future NOI reduction, so the model gets stress-tested, and the lender asks directly what the cash flow looks like if that tenant doesn't renew at the current rent. Above-market rents create a related problem. When in-place rent sits above what the market will currently bear, a lease rolling over amounts to a scheduled cut to income, and lenders re-underwrite those units or suites at market rent once the existing lease expires.
A deeper issue occurs when the documents themselves don't agree. A rent roll, a trailing-twelve-month income statement, and an appraisal report can each show a different occupancy or income figure for the same property, and when they do, every calculation built on top of them, DSCR, LTV, the lender's internal risk grade, inherits that discrepancy. Lenders catch this quickly, and the burden of reconciling it falls on the sponsor, not the credit team. This is part of why document verification has become one of the clearest uses of AI tools inside lending shops: cross-checking a submitted rent roll against leases and financials for internal consistency is a task machines do faster than a human underwriter working line by line, and it means the kind of inconsistency a sponsor might hope goes unnoticed gets caught earlier than it used to.
Vacancy assumptions and expense loads that survive lender scrutiny
Lenders treat a sponsor's vacancy and expense assumptions as a starting offer in a negotiation, not as settled fact, and they apply their own stress test on top of it. A common approach knocks 5 to 10 percent off underwritten NOI for tenant rollover, unexpected vacancy, or softening rental rates, and lenders apply it no matter how strong current performance looks.
Vacancy gets a specific treatment even when a property runs near full occupancy today. Lenders add a stabilized vacancy figure that reflects long-run market norms rather than the current snapshot, on the view that today's occupancy may simply be above trend and won't hold through the loan term. Alongside that sits a credit loss line, a separate deduction for tenants who are technically in occupancy but may not reliably pay, a line that lender models include far more often than sponsor projections do. Replacement reserves follow a similar pattern: lenders apply a per-unit or per-square-foot reserve for future capital expenditures even when a sponsor argues the property is in excellent physical condition. That reserve is a formulaic line item in most lender models, not a judgment call open to negotiation, so arguing against it rarely moves the number.
Value-add multifamily deals face the heaviest version of this scrutiny. Transitional assets get tighter underwriting across the board, with lenders applying conservative renovation assumptions and requiring structured reserves tied to the business plan rather than the sponsor's general assurance that the capital improvements will pay for themselves on schedule.
What lenders look for in LTV
Loan-to-value limits haven't disappeared from the underwriting process, but if a deal clears LTV comfortably and fails DSCR, it doesn't get done. That ordering is the practical correction many sponsors still need to make to their expectations: LTV is a ceiling, not a starting point, and it's applied after cash flow coverage, not before it.
The valuation side of LTV carries its own discipline. Lenders underwrite to the lower of purchase price or appraised value, so if a fresh appraisal comes in above the purchase price, it doesn't open up additional proceeds, it just confirms the deal was priced reasonably. Maximum LTV thresholds also shift meaningfully by asset class. Industrial properties tend to get more favorable leverage treatment, but office faces the most conservative constraints of any major property type, because uncertainty about long-term demand for that space continues. Even a property with strong, well-documented cash flow still runs into a leverage ceiling, because DSCR and LTV are two independent tests that both have to be satisfied at the same time, not two ways of asking the same question. Appraisal review has become a standard part of how lending shops operate for exactly this reason: a lender's downside case depends on the collateral holding its value under stress, so the appraisal gets checked with the same skepticism applied to the rent roll, not accepted as a given simply because a licensed appraiser signed it.
Sponsor equity and sponsorship strength as underwriting inputs, not soft factors
A technically sound financial model doesn't underwrite itself. The sponsor standing behind it is now a quantitative input in the credit decision, not a qualitative impression formed somewhere in the background. Lenders apply specific net-worth and liquidity thresholds alongside DSCR and LTV, and those thresholds are tested the same way the financial ratios are, formulaically, not as a matter of feel.
The standard benchmark many lenders apply requires sponsor net worth to at least equal the loan amount, plus post-closing liquidity that can cover several months of debt service. These are conditions a deal either meets or doesn't, and a sponsor who can't show the balance sheet to back them up will see that reflected in structure, in pricing, or in whether the deal gets approved. The stakes run further than approval, too: a sponsor who can't or won't recapitalize a struggling asset may face a forced discounted sale, or in distressed cases may end up handing the property to the lender, which makes equity capacity a question about whether the deal survives, not only whether it gets approved.
Value-add and transitional assets raise the bar further still. Lenders want to see improvements that are actually funded, not promised, leasing timelines that hold up against current market absorption rather than optimistic projections, and reserves sized to the real cost of the business plan. The capital plan behind a transitional asset gets read with the same scrutiny applied to the rent roll and the expense schedule, because it shows whether the sponsor can execute what the model assumes.
Stress testing and exit scenarios lenders run before approving a deal
Passing underwriting at today's snapshot isn't the same as passing underwriting. Lenders run a deal forward under a standard set of adverse scenarios before signing off, and a model that only works under current conditions is a model that fails this stage. The scenarios are consistent across most credit shops: higher refinance rates at the loan's maturity, cap rate expansion at exit that compresses the property's residual value, and NOI reduction driven by vacancy or rising expenses, frequently modeled together.
Sponsors who get ahead of this rather than react to it improve their position by prioritizing financial transparency, conservative projections, and a business plan that's been thought through. A realistic model accounts for market variability, for lease-up timelines that might run longer than hoped, and for expense pressure that doesn't ease on the sponsor's schedule, and that kind of model moves faster through approval because it reads as credible on first pass. Office assets and anything carrying bridge financing face a sharper version of this test: a credible, funded renovation or leasing plan is a prerequisite for getting new term debt, and lenders expect detailed leasing cost and tenant improvement projections before they'll engage seriously. Exit strategy gets tested just as directly. A deal that only pencils out at an optimistic exit cap rate reads to a lender as a deal whose success depends on market conditions the sponsor doesn't control, and that's precisely the risk a credit committee is pricing, or declining to take on, at the finish line.
Presenting a lender-adjusted NOI next to the sponsor's own, with every adjustment made explicit, is the discipline that separates a credible package from one that invites a retrade. Platforms built specifically for commercial real estate deal workflows, Lev among them, now help sponsors and brokers automate that reconciliation between sponsor and lender assumptions, so the package accounts for the gap between the two figures upfront, before a credit team finds it on its own. In a market sorting $875 billion in maturities by which deals can withstand today's coupons and tomorrow's stress tests, that kind of preparation is what gets a deal to the closing table.
