The most common cause of a repriced or dead file is a coverage ratio built on the borrower’s own income statement. Deriving cash flow independently moves the argument to before the term sheet, where it belongs.
Key Figures
$141.8K
Illustrative proceeds gap between pro forma and derived NOI
3
Ways a pro forma overstates income without anyone lying
7
Line items independently sourced on every file
1.20x
Coverage column published on every valuation
Every experienced sponsor has lived through the retrade. Terms are issued, third-party reports are ordered, money is spent, and then — two weeks before closing — the proceeds come down. The explanation is always some version of the same sentence: our underwriter took another look at the operating statement.
That sentence describes a process failure, not a credit event. The information that caused the reduction was available on day one. It was simply not examined until day thirty, because the initial sizing was run on the number the borrower supplied.
1. Three Ways a Pro Forma Overstates Income
Start with the assumption that the sponsor is honest, because in the overwhelming majority of cases they are. A pro forma still drifts high, structurally, for three reasons that have nothing to do with intent.
Omission of Non-Cash and Deferred Costs
An owner-operator’s actual statement reflects what they spent. It does not reflect what a lender must assume: a management fee even when the owner self-manages, replacement reserves even when nothing was replaced last year, and a vacancy factor reflecting market conditions rather than the owner’s own excellent tenant retention.
Stale Insurance
This is now the single largest source of NOI variance we see, particularly in coastal and Sun Belt markets. The binder in the file is last year’s. The quote the new owner will actually receive, on a new entity with a new loan, is frequently forty to eighty percent higher.
Nobody misrepresented anything; the market moved.
Pre-Sale Tax Basis
A property assessed years ago at a value far below the contract price will be reassessed after transfer in most jurisdictions. Underwriting to the seller’s tax bill is underwriting to a number that will not exist by the second year of the loan.
2. What the Independent Build-Up Looks Like
The alternative is straightforward and, once automated, nearly free: the lender constructs the income statement itself, line by line, from sources that do not belong to either party.
Consider a twelve-unit property valued at $1,500,000, where the sponsor is requesting $1,175,000 of cash-out proceeds.
Exhibit 1 — Pro Forma vs. Derived NOI
Same property, same rent roll, two methods of arriving at net operating income.
| Line | Borrower Pro Forma | Independently Derived | Source Used |
|---|---|---|---|
| Gross potential rent | $178,800 | $174,000 | Market rent comparables and FMR schedule |
| Vacancy & credit loss | (3.0%) $5,364 | (7.0%) $12,180 | Submarket vacancy, program floor |
| Real estate taxes | $14,000 | $21,500 | Assessor record, post-transfer basis |
| Insurance | $8,400 | $13,800 | Current carrier quote, subject entity |
| Management | $0 | $9,708 | 6% of effective income, program minimum |
| Repairs & maintenance | $6,600 | $8,400 | $700 per unit floor |
| Replacement reserves | $0 | $3,000 | $250 per unit |
| Administrative & turnover | $28,700 | $3,600 | Reallocated from blended line |
| Net operating income | $115,736 | $101,812 | Delta: $13,924 |
| Supportable loan at 1.20x | $1,178,300 | $1,036,500 | Proceeds gap: $141,800 |
Illustrative: Assumes 7.25% interest, 30-year amortization, and 1.20x minimum coverage. Not an offer of terms; actual sizing depends on program, market, asset condition, and credit profile.
A $13,900 difference in annual cash flow — roughly twelve percent — produces a $141,800 difference in proceeds.
That is the entire retrade, visible on day one, expressed in a table the sponsor can argue with line by line while there is still time to do something about it.
The disagreement is not eliminated. It is relocated to the beginning of the process, where it costs nothing.
3. The Data Waterfall, and What Happens When a Source Is Silent
Independence is only credible if the sourcing is disciplined.
Each line follows a documented waterfall:
- Primary commercial data provider first
- Public record second
- Modeled estimate as a last resort
Critically, the class of source actually used is stored on the field itself, not assumed.
That last point does more work than it appears to.
When an underwriter can see that a coverage ratio rests on two modeled inputs rather than two verified ones, they treat it differently, condition it differently, and price it differently.
A silent fallback — where a system quietly substitutes a weaker source and presents the result with the same confidence as a verified one — is among the more dangerous defects a valuation stack can carry.
It is invisible unless provenance is a stored field.
See the Derivation, Not Just the Answer
Every line item in an AIVAA cash flow carries its source class and its operands. Nothing is asserted without a lineage.
4. The Coverage Matrix
One output convention changes the tenor of the entire conversation.
Rather than publishing a single supportable loan amount, we publish three columns on every file.
Current
Coverage at the requested proceeds and quoted terms. This tells the sponsor where they actually stand today.
1.00x
The proceeds at which the property exactly services its debt. This is the break-even, and it is the number that matters if rents soften.
1.20x
The proceeds at program coverage — the sizing the file will clear without an exception.
Three numbers instead of one converts an opaque decision into a negotiation with visible parameters.
A sponsor who sees that the gap between requested and supportable is $141,800 can:
- Bring $141,800
- Reduce the request
- Add a rate buydown
- Make the case for an exception with a mitigant
A sponsor who sees only a reduced number has nothing to work with but frustration.
4b. When the Sponsor Is Right and the Model Is Wrong
An independently derived number is not automatically a correct one, and a lender that treats its own output as unchallengeable has replaced one bad habit with a worse one.
There are specific, recurring situations where a conservative derivation understates a property’s real economics, and every one of them is curable with evidence.
Below-Market Leases With Near-Term Expirations
A property carrying legacy rents thirty percent under market is not a weak asset; it is an asset with a schedule.
The cure is a lease-by-lease expiration table with market comparables attached, so the derived rent can be weighted toward the achievable rather than the historical.
Recent Capital Expenditure
A roof, boiler, or full turn completed last year legitimately suppresses forward maintenance.
Invoices and permits move the repairs line; an assertion does not.
Expense Structure Changes
Sub-metering utilities, converting to a triple-net structure, or bringing a property onto a portfolio insurance policy all change the operating basis in ways a comparable-based estimate cannot see.
Documentation of the change — not the intention to make it — is what moves the number.
Short-Term and Mid-Term Rental Income
Where the strategy is documented, the platform history is verifiable, and the jurisdiction permits it, the income is real.
It is also more volatile, so it carries a different vacancy assumption rather than a rejection.
Insurance Quoted Below the Modeled Figure
If a sponsor has a bindable quote below our estimate, the quote wins.
This happens regularly and we adjust every time.
The operating principle is symmetry.
If we require evidence to underwrite a sponsor’s optimism, we owe the same responsiveness when their evidence beats our estimate.
A derivation that only ever moves in one direction is not independence; it is a haircut with better documentation.
The exception register captures both directions, and reviewing where our derived figures are consistently rebutted is one of the more useful signals we get about where the model needs work.
5. How to Make the Two Numbers Agree
Sponsors ask what they can do to narrow the gap.
The honest answer is that most of it is documentation, and it is all available before submission.
1. Get a Current Insurance Quote
Get a current insurance quote on the acquiring entity, not a copy of the seller’s binder.
This is the largest single variance and the easiest to remove.
2. Model the Post-Transfer Tax Assessment
Pull the assessor record and estimate the post-transfer reassessment yourself.
If your jurisdiction reassesses on sale, model it.
3. Present Rent at the Unit Level
Present rent at the unit level with lease copies, not as a blended figure.
Verified rent survives underwriting; asserted rent gets discounted.
4. Separate Operating Expenses
Separate your operating expenses into standard categories rather than a single administrative line.
Blended lines get reallocated conservatively, and conservative reallocation always costs you.
5. Underwrite a Management Fee
If you self-manage, underwrite a management fee anyway.
Every lender will impute one; better that you see the deal the way the credit does.
None of that is sophisticated.
It is simply the work of arriving at the lender’s number before the lender does — which is, in the end, the whole discipline.
The sponsors who consistently close on their original terms are not the ones with the best pro formas.
They are the ones whose pro formas were never going to be surprised.
About the Author
Robert S. Stewart Jr. is Founder and Chief Executive Officer of CR Equity AI, Inc., an AI-native specialty real estate private credit and commercial lending platform.
A U.S. military veteran and licensed real estate professional in Florida and Virginia, he founded CR Equity AI in 2021 and leads development of AIVAA™, the firm’s proprietary underwriting and valuation engine.
Sources & Further Reading
- U.S. Department of Housing and Urban Development — Fair Market Rent schedules.
- Cotality (CoreLogic) — Single-Family Rent Index and related rent series.
- County assessor and tax authority records; carrier-issued insurance binders.
- Bloomberg — property insurance cost escalation coverage.
- Yahoo Finance and Forbes — rent growth data and operator commentary on expense creep.
Get an Independent Read on Your Property’s Cash Flow
AIVAA derives net operating income from third-party rent, tax, and insurance data, then publishes the coverage matrix at current, 1.00x, and 1.20x — so you see the gap before the appraisal does.
Click for more information
Disclaimer
This publication is for educational and informational purposes only and does not constitute financial, legal, tax, or investment advice. It is not a commitment to lend. All loan products, rates, terms, and leverage are subject to underwriting, credit approval, property eligibility, and change without notice. Business-purpose lending only. Worked examples are illustrative and do not represent an offer of terms. Figures cited reflect data available as of the publication date and are drawn from the sources listed. Third-party publications and organizations referenced are not affiliated with, and do not endorse, CR Equity AI, Inc.