You need to enable JavaScript to run this app.
By Travis Yule — CEO & Founder, Full Send Funding
Published 2022-02-16 · Updated 2026-09-23
Revenue is the entry ticket. What sets the offer is average daily balance, deposit shape, negative days and existing positions — and most of it is improvable in 30 days.
In one sentence: What an underwriter derives from four months of statements is rarely the figure the statement prints — revenue is deposits net of transfers between the business's own accounts and of funding proceeds, and the balance that decides the offer is a calendar-day mean rather than the peak an owner remembers.
When funding is underwritten from bank statements rather than tax returns, those four months of PDFs are the application. Everything the decision rests on is in them, and almost none of it is the number an owner thinks it is.
Owners tend to assume the file is scored on revenue. Revenue is the entry ticket. What separates an approval from a decline, and a good offer from a thin one, is what the account looks like between the deposits.
An underwriter opens the statements and derives a handful of figures. They are not published, they are not identical between funders, and they are not complicated.
Each of those is improvable, and what follows takes them one at a time, with the arithmetic shown.
Each figure comes from a particular part of the statement, and knowing which explains why some defects are expensive and others merely untidy.
The first pass is not analytical: the account title against the applying entity, the four statement periods against each other for a gap, the "Page 1 of N" line against the pages supplied. A file that fails it never reaches anyone's arithmetic, and the mechanical defects that stall one are worth twenty minutes before sending.
Most business bank statements print a daily balance summary, usually near the end. Two of the seven figures above — average daily balance and negative days — come out of that table in one pass, without reading a transaction.
Which is why the last pages are the ones to check. Any missing page stops the file — it waits until the window is whole, because a statement with a hole in it cannot be read — but the pages are not equal. A middle page costs some transactions; the balance table costs the two figures that most often decide the size of the offer, and without it negative days have to be rebuilt from the register one transaction at a time.
The bank's totals lump every debit together, the same way its summary box lumps every credit. Positions, overdraft fees, tax debits, garnishments and chargebacks all live on the withdrawals side, and no printed total says which of them are there. So the register is scanned for repetition — same amount, same originator, same weekday rhythm — because repetition is what a fixed obligation looks like, and fixed obligations are what a new remittance has to fit beside.
Monthly revenue is the headline — ours starts at $10,000 a month — but the deposit total is reconciled before it becomes the figure an offer is built on.
Take one month on an ordinary file. The statement's summary box reports total deposits and other credits of $62,000. Inside that number: $54,000 across 41 customer payments and card settlements, and $8,000 transferred in from the business's own savings account. Take the transfer out and verifiable revenue for the month is $54,000 — $8,000 below the number printed at the top of the statement.
Two classes of credit come out by rule, each for its own reason, and they are the two the definition of verifiable revenue names:
Two other things the summary box counts sit outside that netting. A customer check credited and then returned unpaid comes back out as its own returned-item line in the withdrawals column, where section 3 picks it up. A one-off receipt — an insurance settlement, a tax refund, a sold vehicle — is real money but not trade, and reads as a one-off, as the single large wire below does. Either is worth a line of explanation in the application.
The owner is not lying when he says the business did $62,000 in March; he is quoting the figure his bank printed. The offer is built on $54,000. What the lower figure then does — three separate computations rather than one — is worked out in the three numbers an offer is the smallest of.
Forty-one deposits across twenty-one banking days reads as durable, recurring business from many customers. The same $54,000 arriving as one wire on the 30th reads as a single customer, or a one-off, and is discounted accordingly. The second business is not worse; it is less legible, and legibility is what a four-month window buys.
Count is not the whole of it. If $19,500 of that $54,000 came from one payer across three payments, that payer is 36% of the month, and the file becomes partly an assessment of that customer. We publish no deposit-concentration threshold, because none survives a real file: 36% from a municipality on a signed three-year contract and 36% from one private customer with nothing in writing are different risks at the same percentage. Either way it is visible whether or not it is mentioned, and the version you explain beats the version an underwriter infers.
How many days money lands on also decides which payment rhythm fits: deposits on most banking days fit a daily debit and deposits on four or five days fit a weekly one. A business splitting revenue across two accounts halves its apparent deposit count as well as its apparent revenue, and reads as lumpier than it is on both.
This is the number most owners have never computed and the one that most often sets the size of the offer.
The mean runs over every calendar day, banking day or not.
Take a 30-day month in which the account closes at $14,500 on the three days after the largest customer payments land, at $4,300 on five days, and at $950 on the remaining twenty-two. The average daily balance is (3 × $14,500 + 5 × $4,300 + 22 × $950) ÷ 30: $43,500 plus $21,500 plus $20,900 is $85,900, and $85,900 ÷ 30 is $2,863.
Two figures an owner would be far likelier to quote are in that same month, and neither is the one used. The high balance was $14,500. The closing balance was $950. The number the offer is sized against is $2,863, and it sits much nearer the low one for a structural reason: there are twenty-two low days and three high ones, and the mean does not care which felt more representative.
The calendar-day part matters more than it sounds. About nine days of a thirty-day month are not banking days, and the balance does not move on them — a Friday that closes at $950 contributes $950 three times, once each for Friday, Saturday and Sunday. A business that clears its payables on Friday afternoon carries its lowest balance of the week through three days of the mean, every week of the window. That alone separates two files with identical revenue. The same arithmetic is why a childcare center whose tuition arrives once a month and whose payroll runs the account close to zero in between sizes smaller than one with the same revenue and a float left in it.
The balance exists, for underwriting purposes, to pay the remittance out of. So it is read as a ratio rather than as a level.
Stay with the same business: $54,000 a month of verifiable revenue across 21 banking days is $2,571.43 a day of deposits. Suppose an offer at 100% of monthly revenue — $54,000 at a 1.28 factor, which is a 28% cost of capital on the amount funded rather than an annual rate — delivered over a nine-month term of 189 banking days. Total payback is $69,120, and the daily remittance is $69,120 ÷ 189 = $365.71.
Set that against deposits and it is 14.2% of daily receipts, a normal, fundable share. Set it against the balance and the answer changes: $2,863 ÷ $365.71 is 7.8 banking days of cover, a thin cushion for a daily debit. How both measurements are judged, and what happens when either comes in badly, is set out in the fourteen reasons applications are declined. Same business, same month, two tests, and it is the balance that binds — which is the usual answer to "why did I get less than my competitor".
The largest lever most owners have on that number is not how much they take out of the business. It is when.
Take a $9,000 owner draw. Debited on the 3rd, it is out of the account for twenty-eight of the month's thirty days. Debited on the 28th, for three. The twenty-five-day difference, spread across a thirty-day mean, is $9,000 × 25 ÷ 30 = $7,500 added to the average daily balance — same money, same month, same closing balance. On the file above that moves the balance from $2,863 to $10,363 and days of cover from 7.8 to 28.3.
Two honest qualifications, because this is where preparation can shade into presentation.
It is not window dressing. The cash genuinely was in the account on those twenty-five days, a remittance debited on any of them genuinely would have cleared, and the draw is not hidden either — a $9,000 transfer to a personal account is a line in the register like any other. What moving it changes is the cushion that gets measured, not what anyone can see. The version that does not work is the sweep: a draw that takes every dollar above zero on the last day of the month is read as an account with no cushion, as the article on why applications are declined sets out, and sized against whatever the mean says. A fixed draw moved later helps only while the account keeps its own floor after it goes out, as the $950 here does.
But it cuts the other way. If the draw has to go out on the 3rd because a personal obligation lands on the 5th, the cushion is genuinely absent and the thinner file is the accurate one. Moving the date is only available to a business that can afford to, and one that cannot should not arrange a single clean window and then live with a remittance sized to a cushion it does not have.
Days below zero are counted one by one across the window, and the pattern matters more than the count. Three isolated negative days across four months, each explained by a timing collision, is normal. Overdrafts in the same week of every month say the business is structurally short before any new obligation is added, and no responsible funder should put a daily debit on top of that.
Two mechanical notes worth more than the principle.
You can count them yourself. Overdraft fees are their own index: each date carrying one is a day the bank paid an item into a negative balance. A returned-item (NSF) fee is a different line, meaning the bank refused a payment the balance could not cover, and it is read beside the negative days rather than as one of them. Both carry the bank's own descriptor, so they take minutes to find; a balance left below zero for several days adds negative days without new fees, and the daily balance summary shows those.
Which side the return is on changes everything. A customer's check that bounces is a lost receipt rather than a mark against the business: the bank reverses the deposit it came in on, though the reversal lowers the balance like any debit and on a thin day can cause a negative day of its own. A returned debit to another funder is read as a default in progress, and it is one of the few single lines in a statement capable of ending a file on its own.
A fixed debit from another funder is recognised in seconds, and stacking is the reason this matters so much.
An operating account is full of things that repeat: rent, insurance, software, equipment leases, payroll, the utility bill. The tell is not that a remittance repeats — it is a combination of four properties, no one of which is enough alone.
What the four cannot always resolve is a reconciled remittance — one that moves with receipts under a reconciliation clause, which reads as a variable debit rather than a fixed one — or a percentage holdback taken at a card processor, which never reaches the debit column at all and shows only as card settlements smaller than the sales behind them. Both are among the few things an underwriter asks about rather than infers.
The other half of the disclosure sits on the credit side, which is why the netting above and this section are one operation read twice.
Advance proceeds land as a deposit. A large credit carrying a funder's name, followed a few banking days later by a fixed weekday debit running to the end of the window, is a complete account of a position — its approximate size, its funding date and its remittance — assembled without anybody filling in a field. That line is removed from revenue and added to debt service, and it is the only line in a statement that does both.
Underwriters are not offended that positions exist. They need the true picture to size anything safely, and a disclosed position is simply an input. What ends an application is an undisclosed one discovered in the statements, because at that point the file is no longer a question about cash flow — it is a question about whether the rest of the application is accurate. Disclosure is free.
Four months is enough to see a slope, and the slope is measured on net deposits per month rather than gross credits — so a transfer or a funding deposit landing in month three can invent a trend that is not there. The most recent month carries the most weight, because it is the closest thing in the file to a forecast.
Growing beats flat, flat beats declining, but the slope is read in context, and context is what a human underwriter can price around. A February dip at a landscaping company is not a decline; it is February. The harder problem is that four months is shorter than a season, so no position of the window shows a seasonal year honestly — one contains the trough and sizes to it, the next contains the fall off the peak and reads as decline. Seasonal businesses need the structure built around the shape of the year, and the document that resolves it is last year's same four months, which turns volatility into a fact.
The corollary: if your window catches your trough, say so in the application rather than hoping it goes unnoticed. It will not go unnoticed, and an explanation offered up front is worth considerably more than the same one extracted later.
Recurring tax-authority debits, wage-garnishment entries and processor chargeback patterns — none is automatically disqualifying, and each changes how an offer is sized.
The distinction that does the work is between an obligation being managed and one that is not. A payment plan with a taxing authority, visible as a regular debit of a consistent amount, is a positive fact: somebody negotiated it and somebody is honouring it. The same balance with no plan and no debits is a claim on the account that has not started yet and will. A garnishment against the business usually follows a judgment, and an active judgment is usually a conversation rather than a decline, decided file by file, where a payment plan the statements show helps — the decline list explains how it is read; one withheld from an employee's pay and passed on is payroll, not the business's own debt.
One line of arithmetic explains more offers than anything else in this article.
Suppose a file is sized at 100% of monthly revenue: revenue R, delivered over n months of 21 banking days at a factor f. The daily remittance is R × f ÷ (n × 21) and daily deposits are R ÷ 21. Divide the first by the second and both R and the banking days cancel: the remittance share is simply the factor divided by the number of months.
So at the factor this site's worked examples use for each term (actual pricing is set per deal), an offer at a full month of revenue takes 30.0% of daily deposits at 1.20 over a four-month term, 20.7% at 1.24 over six months, 14.2% at 1.28 over nine and 11.0% at 1.32 over twelve — and the size of the business never enters it. The factor rises with the term, but the share still falls, because the months grow faster than the factor. The first two are heavy shares for any account to carry on their own, which is why a request for a full month of revenue on a short term so often fails the capacity test however healthy the account is; how that share is judged, and what happens to a file that carries too much, is the first entry in the reasons a file is declined.
That is why term is not a detail settled at the end of a conversation. It decides whether the amount you want fits the account you have, and it is legible in the statements before anybody has written an offer. This is a point about how an offer is priced and sized while it is being built, not about what happens to an agreement already signed.
In rough order of how much each moves the file.
Your personal credit score is not the decision. It is pulled, softly, and it is one input among many — revenue-based underwriting exists precisely because the score is a poor predictor for a business with real cash flow, which is why, on deals up to $2 million, we approve 90% of complete applications that meet our requirements. What a soft pull is and what it is not covers the mechanics.
Profitability, as an accounting concept, is not visible here. The account shows cash, not accrual. A profitable business collecting in ninety days can be less able to carry a daily debit than a break-even one paid at the till, and the statements show the second thing rather than the first.
Old history is not the focus. The most recent months carry the weight, which is why a business that fixes three things and waits a window is very often a different file.
The amount you asked for is read; it is not one of the three numbers the offer is the smallest of, and a business plan, a projection or a résumé is not an input at all. What is actually worth preparing before you apply is a short list, and the only writing on it is one sentence on what the money is for.
A customer that has just given notice. Concentration is visible in the deposits; the email cancelling the contract is not, and the statements go on looking excellent for another sixty days.
The accrual position. A business with $180,000 billed and not yet collected shows none of it, and a business holding $180,000 of deposits it may have to refund shows all of it as ordinary credits. A statement records cash movements, not obligations, and nothing in it distinguishes a deposit that is earned from one that is not.
Anything outside the window. A trough five months back, a lawsuit settled last year, a lease expiring next quarter. Four months is short enough that a bad patch two years ago is genuinely gone, and short enough that a seasonal cycle does not fit.
Whether a deposit is revenue or the owner's own money. The description usually resolves it. A wire from a related entity may not, and an underwriter will ask rather than guess.
Whether the business should borrow at all. This is the real limit. The statements answer whether the account can carry a remittance. They cannot answer whether adding one is a good idea, and the two have different answers more often than the industry admits — there are situations where borrowing is simply the wrong answer, and a clean statement window is not evidence against any of them.
That is why the application asks in writing for what the statements cannot say, and why an underwriter calls. Four months of PDFs is fast, cheap, hard to dress up and narrow; the trade is deliberate, because a bank's file answers a wider question and takes weeks to assemble. How these figures turn into an amount, a term and a price is written out separately.
Everything above describes a first file, where the evidence about how a business handles a remittance is indirect. At renewal the strongest line is your own remittance history, in your own statements: four months of a fixed debit clearing on every banking day, no returned items, a balance that held while it ran. No first-time file can produce that, which is why a business that survives one cycle cleanly usually sees a different offer rather than the same one — how renewal terms are built is worth reading before the call comes.
We fund $5,000 to $10 million in working capital against monthly revenue of $10,000 or more, from four months in business, with a decision within 24 business hours and, on deals up to $2 million, funding within 24 hours of approval. A merchant cash advance and our flex line of credit are both sized from the same four months of statements this article describes. Checking your options is a soft credit pull only, four months of recent statements is the whole document ask, and the application takes about ten minutes. If you would rather see the sizing arithmetic first, start with how much your business can actually borrow.
The signal table lists what the firm's underwriters derive from four months of statements and how each moves an offer. It is the firm's own practice, stated from the decision side, not a survey of the market. No decline threshold is stated: the remittance share and days of cover are computed for the worked file and read qualitatively, because each is judged against the whole file rather than against a published cut-off, as the article on why applications are declined explains. The worked file is arithmetic on stated inputs, not a client: $62,000 of credits netting to $54,000 of verifiable revenue once an $8,000 transfer from the business's own savings is removed; a $2,863 average daily balance from a stated month in which the account closes at $14,500 on three days, $4,300 on five and $950 on twenty-two; and $365.71 a day on a $54,000 advance at a 1.28 factor over nine months of 21 banking days, which is 14.2% of daily deposits and 7.8 days of cover. The remittance shares for other terms are the factor divided by the number of months, at the factor-by-term ladder the site's worked examples are priced from, and the draw-date effect is a $9,000 draw moved twenty-five days later across a thirty-day mean. The approval rate — 90% of complete applications that meet the firm's requirements, on deals up to $2 million — is the firm's own figure.
True monthly revenue with transfers and funding proceeds netted out; how many deposits there are and how they are spread; average daily balance; negative days and overdraft patterns; existing funder debits; the revenue trend across the window; and other recurring obligations such as tax payment plans or chargebacks. Revenue gets you in the door. The rest sets the size and price of the offer.
Four months of recent business statements is the whole document ask for revenue-based funding. That window is long enough to show a trend, a balance pattern and any existing positions, and short enough that a rough patch two years ago is simply not in it. Older history matters far less than most applicants expect.
It is the average amount actually sitting in the account across the month, rather than the total that passed through it. It matters because a daily remittance is paid out of the balance, not out of the monthly total: a business depositing $40,000 a month while riding near zero between deposits will overdraft in the first slow week, while the same $40,000 with a $12,000 average balance absorbs the same debit without noticing. It is usually the reason two businesses with identical revenue get very different offers.
Not by themselves — the pattern matters more than the count. Three isolated negative days across four months, each explained by a timing collision, is normal and survivable. Overdrafts in the same week of every month say something structural, and no responsible funder should add a daily debit on top of that. If you have any choice in the timing, apply from a window where the statements are clean.
Yes, and it is one of the most common self-inflicted problems. Revenue landing in a personal account is generally not counted, so the business looks smaller than it is and the offer is sized against the understatement. Routing everything through one dedicated business account is one of the highest-return preparation steps, and it costs nothing; it is worth doing a full month before applying rather than the week of.
Because they inflate apparent deposits without adding any revenue, so an underwriter nets them out. A business that moves money between operating and savings looks larger to itself than it does in the file, and the correction can be substantial. Loan and advance proceeds are removed for the same reason — and they also disclose the position that funded them.
Clear any recurring overdraft pattern; route all business revenue through one dedicated business account; stop discretionary transfers between your own accounts during the window; time large discretionary purchases after the statement window rather than before it; leave a float in the account so the average daily balance reflects the business; have payoff letters ready for positions you plan to retire; and write down the explanation for anything unusual so it goes in the application rather than being extracted later.
It is pulled — softly, so applying leaves no mark — and it is one input among many rather than the decision. Revenue-based underwriting exists precisely because a personal score is a weak predictor for a business with real cash flow, which is why, on deals up to $2 million, 90% of complete applications that meet our requirements are approved. Profitability as an accounting concept is not assessed at all here: the account shows cash, not accrual.
Because the summary box adds up every credit, and underwriting subtracts the ones that are not revenue: transfers in from your own accounts, and loan or advance proceeds. On a month showing $62,000 of total credits, made up of $54,000 in customer payments and an $8,000 transfer from savings, the figure the offer is built on is $54,000, which is $8,000 lower. A one-off receipt such as an insurance settlement is read as a one-off rather than as a month of trade, so explain it in the application. Quote the net number when you apply and the offer will start from the figure you expect.
By a combination of four properties in the debit column, no one of which is enough alone (a position funded before the window shows the first three): a daily or same-weekday cadence, banking days only and never a weekend or a federal banking holiday, the same amount to the cent, and a start date inside the window. It usually comes from an originator whose name recurs across thousands of files. Rent and insurance run monthly on a date; payroll moves with hours; a remittance does neither. The funding deposit that started it is usually visible too, a few banking days earlier on the credit side. Finding one takes seconds, and the cost is not the position itself — it is that the file stops being a question about cash flow and becomes a question about accuracy.
It can move it more than the size of the draw does, because average daily balance is a mean of every day's closing balance. A $9,000 draw taken on the 3rd of a 30-day month is out of the account for twenty-eight days; taken on the 28th it is out for three. That twenty-five-day difference adds $9,000 times 25 divided by 30, or $7,500, to the average daily balance on identical revenue. This is measurement rather than presentation — the cash really was there — but it only helps a business that can genuinely afford to wait. If the draw has to leave early, the thinner file is the accurate one. A draw that empties the account on the last day of the month is read as a sweep and sized against as no cushion, so the gain holds only while the account keeps a floor after the draw.
Quite a lot, and it is worth knowing which questions still have to be asked. They cannot show a customer who has just given notice, only the concentration that customer represents. They cannot separate cash from obligations: money billed but not yet collected does not appear at all, and money collected but refundable looks exactly like money earned. They cannot see outside the window, which means a full seasonal cycle does not fit. And they cannot say whether borrowing is a good idea — only whether the account could carry a remittance. That gap is why the application asks for things in writing and why an underwriter calls.
Yes, and the second read has evidence the first one could not. Alongside deposits and balance, a renewal file carries your own remittance history: four months of a fixed debit clearing on every banking day, no returned items, and a balance that held up while it ran. No first-time application can produce that, which is why renewal terms are usually better than first-advance terms rather than the same. The reverse is also true — returned debits and a falling balance during the term are in the same statements, and they are read.
Travis Yule worked in business funding before starting Full Send Funding in 2021, and has more than five years in business funding in all. He leads the company from Middle Grove, New York, and writes about working capital from the underwriting side of the table — what the numbers actually have to say before a business gets funded.