Merchant cash advance underwriting: the complete guide

Merchant cash advance underwriting is the process of evaluating a business’s bank statements — its deposit volume, daily balances, existing financing obligations, and cash-flow patterns — to decide whether to purchase a portion of its future receivables, and on what terms. Unlike traditional loan underwriting, which leans on credit scores, tax returns, and collateral, MCA underwriting treats the bank statement as the primary evidence. Repayment is collected daily or weekly from the merchant’s receipts, so the only question that really matters is whether the cash flow can carry another fixed remittance on top of everything the business already pays.

That difference shapes everything about the discipline. An MCA is structured as a purchase of future receivables rather than a loan, repaid through a fixed daily or weekly debit or a holdback percentage of receipts. There is no amortization schedule to stress-test and usually no hard collateral to fall back on. The underwriter’s protection is the quality of the cash flow itself: how much real revenue comes in, how consistently it arrives, how much of it is already committed to other funders, and whether the statements being reviewed are genuine.

It is also fast by design. Where a bank loan can take weeks, MCA underwriting typically compresses into hours or a day or two, working from three to six months of business bank statements. That speed is a competitive requirement — merchants shopping for capital take the first reasonable offer — but it puts enormous pressure on the accuracy of bank statement analysis. A missed position or an inflated revenue figure at hour one becomes a defaulted advance at month three.

Every funder weights these differently, but virtually every MCA underwriting shop evaluates the same six dimensions.

Deposit volume and consistency. Monthly deposit volume sets the ceiling for offer size, but consistency matters as much as the total. A merchant depositing $60,000 every month is a stronger file than one swinging between $20,000 and $110,000, even at the same average. Underwriters also look past gross deposits to external revenue — money genuinely earned from customers and processors — because transfers between the merchant’s own accounts and proceeds from prior advances inflate the gross number without representing a dollar of income.

NSF activity and negative-balance days. Non-sufficient-funds events and days the account ran below zero are the clearest signals of cash-flow strain. A merchant who can’t cover today’s obligations will not reliably cover a new daily remittance. Many funders tolerate a handful of NSFs across three months; a pattern of them points to a business already living past its cash.

Existing positions. Each active advance the merchant is repaying — each position — claims a fixed slice of daily or weekly cash flow before the new advance sees a cent. Position count and combined debt service are the single most decisive inputs in most MCA underwriting models.

Time in business. Newer businesses fail more often, and their statements offer less history to read. Most funders want six to twelve months minimum; stronger programs and better pricing typically start at two or more years.

Industry risk. Some industries carry structurally volatile or seasonal cash flow — restaurants, trucking, construction, retail — while others, like medical practices and professional services, tend to show steadier deposits. Industry doesn’t decide a file on its own, but it shifts where the bar sits on everything else.

Paper grades A through D. The industry compresses all of the above into a letter grade. A-paper is the strongest file: multiple years in business, high and consistent external revenue, minimal NSF activity, zero or one existing position — it earns the largest advances, the lowest factor rates, and the longest terms. B-paper is solid with a blemish: good revenue but a second position, or some NSF history. C-paper carries real risk — multiple positions, meaningful NSF or negative-day activity, shorter time in business — and pays for it with higher factor rates and shorter terms. D-paper is high-risk: heavily stacked, strained cash flow, often only fundable in small amounts at aggressive pricing. The grade isn’t a formality; it is the pricing.

A bank statement is a few hundred rows of noise with a handful of decisive signals buried inside. Experienced underwriters read for four things.

Recurring daily or weekly ACH debits. Existing advances announce themselves as equal debits on a fixed rhythm — the same amount pulled every banking day, or the same amount every week. Cadence is the key discriminator: daily or weekly debits to a lender-like payee indicate an active MCA position, while monthly debits indicate a term loan or installment. No real MCA collects once a month. Conflating the two inflates a merchant’s apparent stack.

Known funder descriptors. ACH debits carry company names and descriptor fields, and repeat funder signatures become recognizable. Grouping debits by payee — rather than scanning row by row — is how positions are actually found, and it is how MetrikData assembles them automatically, with each position tied back to its source transactions.

Revenue versus transfers. Credits from payment processors, card settlements, and customers are revenue. Credits from the merchant’s own savings account, reversed payments, and advance fundings are not. Separating the two is what turns gross deposits into external revenue — the only honest denominator for burden math.

Low-balance and negative days. The daily balance column tells you whether the merchant has any cushion. An account that dips near zero before every deposit cycle will miss remittances the first slow week.

A worked example ties these together. Consider a merchant showing $85,000 in monthly deposits. Line-by-line review finds $11,500 in transfers from the owner’s savings account and $1,200 in reversed payments — so external revenue is $72,300, not $85,000. Grouping the debits by payee reveals two existing positions: Funder A pulling $740 every banking day and Funder B pulling $500 — a combined $1,240 per day.

• Gross deposits: $85,000 — external revenue after removing $12,700 in transfers and reversals: $72,300
• Existing positions: Funder A at $740/day + Funder B at $500/day = $1,240/day
• Monthly debt service: $1,240 × 21 banking days = $26,040
• MCA burden: $26,040 ÷ $72,300 = 36.0% of external revenue
• Same math against gross deposits: $26,040 ÷ $85,000 = 30.6% — understating the real pressure by more than five points
• Add a proposed third position at $600/day: $38,640 ÷ $72,300 = 53.4% — past the 50% threshold most underwriters treat as significant repayment pressure

MCA stacking is a merchant taking on a new advance while one or more existing advances are still being repaid, layering multiple daily or weekly remittances on top of each other. It is the risk MCA underwriting exists to catch, for a simple structural reason: every funder in the stack priced their advance against cash flow that the next funder quietly claims a piece of. The first position may have been prudently sized at 25% of revenue; by the third position, half the merchant’s real income can be leaving the account before payroll, rent, or suppliers see it.

Stacking hides in plain sight because no single row in a statement says “stack.” It looks like several distinct lender-like payees, each on its own rhythm. Spotting it means grouping every debit by payee across the full statement period, applying the cadence rule to separate true MCA positions from term loans, and watching for advance proceeds landing mid-period — a large lender credit followed by new recurring debits means a position opened during the statement window. New payees with only one or two debits deserve flagging rather than counting, since a rhythm isn’t established yet.

This is exactly the analysis MetrikData’s stacking detection automates: transactions are grouped by payee to identify recurring funder withdrawals, MCAs are separated from term loans so the stack count isn’t inflated, and ambiguous payees are surfaced for review instead of silently dropped or counted. The result is a position table with estimated daily and weekly debt service — every entry backed by the source transactions.

Because the statement is the collateral in merchant cash advance underwriting, the statement is also the fraud surface. Altered and fabricated PDFs are a persistent problem, and the tells fall into a few repeatable categories.

Document-level tampering signals. A genuine bank-issued PDF is generated once by the bank’s systems. A doctored one usually betrays itself in the file: editing-software traces in the metadata, a modified timestamp later than the created timestamp, incremental updates layered into the file, or fonts that don’t match across the document — a sign that individual numbers were replaced after the fact.

Balance discontinuities. Arithmetic doesn’t lie. Beginning balance plus total credits minus total debits must reconcile to the ending balance, and one month’s ending balance must equal the next month’s beginning balance. When a merchant inflates a deposit or deletes an NSF fee, the chain breaks somewhere.

Behavioral red flags. Some fraud is staged rather than edited: round-number deposits appearing with no processor descriptor, especially clustered in the weeks before application; deposit volume that spikes sharply in the most recent month; missing pages or missing months in the statement set; and heavy transfer activity dressing up gross deposits.

None of these signals is proof on its own — a re-saved file or a legitimate owner injection can trigger one — which is why verification should flag for review rather than accuse. MetrikData’s statement verification runs these checks on every upload, with each flag pointing back to the document itself so the underwriter can inspect the evidence and make the call.