Skip to content
// FOR THE OTHER SIDE OF THE TABLE
FREE

How banks
read you.

Everything on this site is written for the person on the bank's side of the table. This page is written for the person on yours — the founder or CFO whose facility is up for review, whose bank has started asking different questions, or who is about to ask for more.

It is not a guide to getting approved. It is a description of the order a credit-trained reader goes through your business in, written by someone who spent sixteen years doing it. Knowing the order is worth more than any phrasing, because the order is what decides which of your numbers gets looked at first.

// 01 — THE THIRTY-SECOND VERSION

  1. 1 · Does the amount match the arithmetic? Your cash conversion cycle applied to your revenue gives a number. Yours is compared to it before anyone reads a word about your industry.
  2. 2 · Is the cycle the same shape as last year? Growth that carries working capital is one thing. A cycle quietly lengthening is a different thing wearing the same clothes.
  3. 3 · What repays this if trading does not? Every loan has two ways out. The second one is priced whether or not anybody mentions it in the meeting.
  4. 4 · Who else can decide your outcome? One large customer, one landlord, one supplier, one contract clause. Concentration is read as a transfer of control, not as a sales achievement.
  5. 5 · Did you say it first? Everything above is discoverable. The only variable you control is whether the bank hears it from you or finds it in a schedule.

// 02 — THREE FILES, READ FROM YOUR SIDE

These three requests are published on this site as case drills for bankers — the full credit reading of each is a click away. Here is the same file from the borrower's chair: what is on your desk, what is being computed about it, and what you could do before the meeting rather than during it.

// THE GROWTH REQUEST

You are up 36% and want a bigger revolver

What is on your desk
Revenue has gone from NZ$28m to NZ$38m in two years on two three-year supply agreements. You want the working capital revolver lifted from NZ$2.5m to NZ$4.0m. It feels obvious.
What is being computed
Not your growth — your cycle. Debtor days 44, inventory around 60, payables around 38: a cycle near 66 days, which on NZ$38m of revenue is roughly NZ$6.9m of working capital, of which the incremental NZ$10m of revenue accounts for about NZ$1.8m. You asked for NZ$1.5m. You asked for less than the arithmetic supports, and that is established before the file is opened.
Three things to do first
Do that calculation yourself and put it in the application — a borrower who can say “the cycle supports 1.8 and I am asking for 1.5” is a different kind of counterparty from one who asks for a round number. Volunteer that debtor days moved two days in two years; it is your strongest fact and it is buried in a note. And make clear the capacity expansion is already funded by a separate amortising term loan, or the reader will wonder whether the revolver is quietly covering capex.
THE BANK-SIDE READING OF THIS FILE →
// THE SEASON THAT DID NOT CLEAR

Last Christmas did not sell, and you are buying more

What is on your desk
NZ$1.1m of last season is still in the distribution centre. You believe the problem was buying the wrong categories rather than absent demand, so this year's buy is up 22% and you need the seasonal line lifted from NZ$2.0m to NZ$3.5m.
What is being computed
One question: did last year's loop close? A seasonal facility funds a cycle that returns to zero — that clean-down is the whole covenant. It did not close, so part of the NZ$1.5m you are asking for is not seasonal money at all. It is the unsold past, asking to be refinanced under the season's name. The reader can also see operating cash flow down 60%: you enter this bet with less cushion than last year, not more.
Three things to do first
Split the two amounts yourself — this season's purchase, and the carry-over from last. If you do not, the reader will, and their version is less generous. Buy the increment with evidence rather than conviction: September pre-orders convert part of the 22% from belief into demand, and that part is fundable. And do not rest the case on “we bought the wrong categories” — it may well be true, and it cannot be verified from the file, which makes it worth nothing at this table.
THE BANK-SIDE READING OF THIS FILE →
// ONE CUSTOMER

Your biggest customer is now 58% of revenue

What is on your desk
You won the renewed national contract. That customer has gone from 31% to 58% of revenue, and you need NZ$6.0m for eighteen linehaul units and a dedicated distribution centre to serve it. The contract runs five years.
What is being computed
Not that customer's credit — they are investment grade and they will pay. What is being priced is the asymmetry of commitment: you are taking five-year assets against a relationship the other side can leave in ninety days. And your assets split in two. Tractor units are generic and haul for someone else at a discount. Racking and fit-out configured to one customer's network is close to scrap the day the notice arrives. The 90 basis points you gave up on renewal is read as the customer already exercising the power the concentration handed them — politely, on schedule, and again at every negotiation from here.
Three things to do first
Split the capex yourself by what survives the contract ending, and propose the shorter amortisation on the dedicated half before it is imposed on you — offering it and being handed it are different conversations. Be open about fleet age: if part of this “growth capex” is deferred replacement, the reader will see it, and it costs less coming from you. And work out what actually makes you hard to replace, because under a ninety-day clause that is what the bank is underwriting, not the contract.
THE BANK-SIDE READING OF THIS FILE →

Every borrower named on this page is a composite. The figures are internally consistent and the patterns are drawn from sixteen years of commercial banking practice, but no company here is real and none represents any actual borrower, transaction, or institution.

// 03 — WHAT YOU ARE ALREADY BEING READ ON

WHAT YOU THINK IS BEING JUDGED
WHAT IS ACTUALLY BEING READ
Profit
Whether cash behaves the way profit says it does
Your growth story
Whether the cycle kept its shape while you grew
The security you are offering
What that security is worth to somebody who is not you
How well the meeting goes
What the schedules said before the meeting started
Whether the bank trusts you
Whether anything in the file arrived later than it should have

// 04 — QUESTIONS BORROWERS ASK

What does a bank actually look at first in a business loan application?
Not the profit. The first thing a credit-trained reader does is size the request against the cash conversion cycle — how long money is tied up between paying for something and being paid for it — and ask whether the amount requested matches that arithmetic. A request that matches the cycle reads as the mechanical consequence of growth. A request that exceeds it, or that is a round number reached for in a meeting, invites the question of what else the money is for. This happens before anyone forms a view about the business.
Why does the bank keep asking for aged debtors and stock reports?
Because those two schedules are where a lengthening cycle shows up first, and a lengthening cycle is the difference between working capital that growth carries and working capital that funds a hole. Debtor days moving from 44 to 79 says something the profit line will not say for another year. The request is not administrative box-ticking; it is the bank looking at the one thing that moves before everything else does.
Is it better to ask for more than I need, in case they cut it?
No, and it is one of the few genuinely self-defeating moves available. A credit reader sizes the requirement independently before reading your number. If your number is under what the arithmetic supports, that is evidence of discipline and it is noticed. If it is over, the gap becomes the subject of the meeting, and you spend your credibility explaining a buffer instead of your business. Ask for what the cycle supports, and show the working.
What makes a bank nervous about a customer being a large share of my revenue?
Concentration is not a linear risk. Below about a third of revenue, losing the customer is a bad year. Above half, losing them is a solvency event — and both sides' procurement teams know it, which is why margin usually starts giving way before anything else does. What the bank is pricing is not that customer's credit; it is the asymmetry between assets with a five-year life and a contract the other side can leave in ninety days.
Should I tell my bank bad news before they find it?
Yes, and the reason is mechanical rather than moral. A bank that learns of a problem from you weighs your explanation. A bank that finds it in a schedule weighs your explanation at a discount, because it now has a second question — what else has not been mentioned. The cost of volunteering a bad number is smaller than the cost of the discount that follows discovery, and the discount attaches to everything you say afterwards, including the true and favourable parts.

// THE SAME DISCIPLINE, AS THE DRILLS ARE READY

New case drills go out by email as they are ready. They are written for the banker's chair, which is exactly why they are worth reading from yours — it is the reasoning you are being subjected to, in full, before it is applied to you.

// IF YOU WANT A WHOLE WORKED CASE NOW — one is free, in full