What We See in Your Bank Statements (So You Can See It First)

Four months of bank statements is the universal admission ticket to working capital underwriting. Operators submit them constantly. Most never know what an underwriter is actually scanning for.

The list isn't long. There are nine signals. They get read in roughly the same order every time, by every desk worth its salt, and the order itself reveals the actual risk model behind the underwriting decision.

Here it is, exactly as we walk a file.

1. Daily ending balance trend

Not the average. The trend.

Open the four most recent statement periods side by side. Look at the daily ending balance graph at the top of each statement (most banks include this; if yours doesn't, plot it from the daily detail in five minutes). The question is whether the line is trending up, flat, or down across the four months.

A flat or rising daily balance line, even a low one, is fundable. A descending one — even from a high starting point — is a yellow flag that opens a different conversation. The trend is read first because it's the single best predictor of where the business is heading, and underwriting is fundamentally a prediction exercise.

2. NSF and overdraft frequency

Count them. Not categorize them — count them.

Three NSFs across four months is normal. Six is a yellow flag. Ten is a hard pass. The reason isn't moral; it's predictive. Operators who run overdrafts repeatedly are operators whose ACH debits will eventually fail, and a failed debit is the single most expensive event in a portfolio.

What underwriters are not counting: NSF letters that were reversed, single-day clusters from a known event, or fees from the bank's own servicing. Real underwriters discount these correctly. Algorithmic ones don't.

3. Negative-day count

The number of business days the account ended below zero. Different from NSF count — you can be negative without triggering an NSF if you have overdraft protection.

Same scoring: 0–3 is clean, 4–8 is monitored, 9+ is a problem. The frame is identical: predictive signal of repayment friction, not a moral judgment.

4. Average daily balance vs. requested funding amount

The single most common reason an applicant gets a counter-offer instead of a full approval.

The rule of thumb at most desks is that average daily balance should be at least 8–12% of the requested funding amount. Below that, the deal gets countered to a smaller amount that the cash position can support, regardless of revenue.

A business doing $80K/month in revenue with a $4K average daily balance is asking the cash position to do more work than it can. The same business with a $9K average daily balance opens up a meaningfully larger approved amount. The revenue is the same. The cash floor is what changed.

5. Deposit consistency

Three signals here, all from the same data:

Number of deposit days per week. A business with deposits on 5–6 of 5 business days is high-confidence. A business with deposits on 1–2 days is more variable and gets sized smaller.

Deposit count vs. month length. A 30-day month with 8 deposits is choppy revenue. A 30-day month with 22 deposits is high-volume operations.

Largest single deposit as a % of monthly volume. When a single deposit is more than 30% of monthly volume, the file looks more like an invoice business than a daily-receipts business. That's not bad — it's a different shape — but it changes the underwriting model.

6. Outflow patterns

Now look at debits, not credits.

What's the recurring weekly debit pattern? Payroll on the 1st and 15th plus rent on the 5th plus utilities mid-month is a textbook operating business. Random large debits with no pattern is either a very young business or one with cash control issues.

The specific number underwriters compute: fixed monthly outflows ÷ average monthly revenue. Below 60% is healthy. 60–80% is monitored. Above 80% means the business has very little discretionary cash to absorb a weekly funding debit.

7. Existing funding stack

The statements will show every existing funder, every day, by line item. There's no hiding it.

Underwriters total up the daily and weekly debits to existing funders and ask: with the proposed new advance added, what is total weekly debt service as a percent of average weekly revenue?

Above 25% is the universal pause line. Above 35% gets the file declined regardless of revenue. The reason isn't that we don't want to compete; it's that adding a fifth weekly debit to a stack of four is operationally fragile. Any one bad week and the whole stack fails.

8. Card-vs-non-card revenue mix

Read three lines of the deposit detail and you can usually tell. Card processors deposit nightly with their unmistakable naming patterns (TSYS, Worldpay, Square, Toast, Stripe). Everything else is ACH from invoiced customers, wire-ins, or counter cash.

The mix tells you what kind of business this is for product-fit purposes. High card mix = MCA territory. Mixed = working capital. High invoice/ACH = factoring or term loan territory.

9. Anomalies in the most recent 30 days

Last. Always last.

A clean three-month picture with a recent 30-day cluster of NSFs, large unexplained transfers, or a sudden drop in deposit consistency triggers a recent change concern. The conversation shifts from "is this business fundable" to "what changed in the last month and is it still happening."

Sometimes the answer is benign — a known seasonal dip, a one-time business interruption, an owner away on a closing. Sometimes it's the leading edge of a real deterioration. Either way, recent anomalies always get a conversation.

What this gets you as an applicant

Two things.

First: stop sending statements with the expectation that "the numbers speak for themselves." They speak — but they're being read in a specific order, against specific thresholds, for specific predictive reasons. If you know in advance that one of the nine signals will look weak, write a one-paragraph note and attach it. Underwriters universally weight context they were given upfront over context they had to dig for.

Second: the nine signals above are the entire decision. Anything not on this list is decoration. If your file is clean on the list, you're fundable. If it's not, no amount of business-narrative slides will move the file.