Picture a plumbing contractor who’s had a good couple of years. He’s outgrown his shop, he’s turning down work, and he finally books a meeting with his bank to talk about borrowing. The banker runs the numbers and comes back with a figure: six hundred thousand.
He walks out feeling pretty good. And from that moment on, six hundred thousand is the number in his head. Everything after that is a conversation about whether he wants all of it, or most of it, or maybe just a bit less to be safe.
See the problem? He let someone else set the anchor. The bank told him what it was willing to risk, which is a completely different question from what his business can comfortably carry.
The better order is the other way round. Work out what you can afford to pay back. Work out what you actually need. Then borrow whichever of those two numbers is smaller.
That sounds obvious enough that you’d think everyone does it. Almost nobody does, because the first question takes a phone call and the second takes an afternoon with your accounts open. Here’s the afternoon version.
Start with what’s genuinely left over
Pull your EBITDA for the last twelve months. Not last fiscal year, the last twelve months, because that’s where the business actually is right now.
Then take three things out of it, and be honest about all three.
Pay yourself properly first. If you’ve been drawing below market rate to make the numbers look better, add the difference back as a cost. You’ll have to pay it eventually, or pay someone else to do your job.
Take out tax. If you’re a pass-through, the money you pull out to cover your personal tax bill on business income was never available to a lender.
And take out maintenance capex. Not the exciting growth spending, just the routine replacement of the things that wear out. The van that needs doing next year. The compressor that’s on borrowed time.
Whatever survives all that is the cash you have to service debt with. That’s the only figure in this whole exercise that matters, and it’s usually a good bit smaller than people expect. If it comes out lower than you hoped, the number is doing its job.
Then leave yourself room
Most banks want to see that your cash covers your loan payments at least one and a quarter times over. Use the same test on yourself, and treat it as a floor rather than something to aim at.
So divide that cash figure by 1.25. That’s the absolute most you should be spending on debt payments in a year, all in. Now subtract what you’re already paying, and don’t forget the bits that are easy to overlook: the truck notes, the equipment finance, the interest you carry on a line that never quite gets back to zero.
What’s left over is your room to move. Divide it by twelve if you think in monthly terms.
Turning a payment into a loan amount
Here’s where a lot of owners trip up, because a monthly payment and a loan size aren’t the same thing at all, and the length of the loan matters more than most people realise.
Round numbers, at roughly eight percent: every thousand pounds or dollars a month of capacity will support somewhere near forty-nine thousand over five years, sixty-four thousand over seven, and eighty-two thousand over ten. Real estate stretched over twenty-five years supports about a hundred and thirty.
Look at what that means. Moving from a five-year to a ten-year term gets you roughly two-thirds more borrowing power on exactly the same monthly payment. That’s usually worth far more to you than grinding your banker down a quarter of a point on the rate, and it’s a much easier thing to ask for. Rates move around, so run the arithmetic at whatever you’re actually quoted rather than my eight percent.
Now the other half of the sum
Put capacity aside for a minute and work out what the project genuinely costs, from the ground up. Two things get underestimated nearly every time.
The first is the all-in cost of whatever you’re buying. An equipment quote is not the cost of the equipment. There’s freight, rigging, the electrical work, permits, sales tax, commissioning, training your people to use the thing, maybe the first year of a service contract. That lot can add fifteen to thirty percent on top of the sticker price. Borrow the quote and you end up funding the rest out of working capital, which is the most expensive way to do it.
The second is the cash that growth quietly swallows, and this is the one that catches good businesses out.
There’s a quick way to size it. Take how long your customers take to pay you, add how long stock sits around, then subtract how long you take to pay your own suppliers. That’s your cash conversion cycle in days. Multiply your expected extra annual revenue by that number of days, divide by 365, and you’ve got a rough figure for the cash the extra work will absorb before any of it comes back.
A business on a 55-day cycle taking on another nine hundred thousand of revenue will tie up something like a hundred and thirty-five thousand in cash, just in the mechanics of doing the work. That’s not profit you’re waiting on. It’s money out the door in advance. Companies go under this way while showing a profit on paper, and they never see it coming because the P&L looks fine.
Then add a contingency. Ten or fifteen percent on anything involving a build, an install, or a timeline. Going back to your lender halfway through for a smaller top-up facility is slow, faintly embarrassing, and almost always priced worse than getting it right the first time.
Match the money to the job
Most small business loans fall into a few basic shapes, and putting the wrong money against the wrong need is the most common mistake in the whole process.
A line of credit is for working capital and seasonal swings. It’s supposed to breathe: drawn when your receivables and stock build up, repaid when they turn back into cash. If yours hasn’t touched zero in two years, you’ve funded something permanent with temporary money, and your renewal meeting is going to be awkward.
A term loan is for equipment and fixed assets, and it should run roughly as long as the asset lasts. A machine with seven good years in it on a three-year note creates a cash squeeze you didn’t need. A van on a seven-year note means you’re still paying for it long after it’s gone.
Property goes on a mortgage or an SBA 504 over twenty to twenty-five years. And SBA 7(a) is often worth a look purely because the longer amortisation is what makes a deal serviceable in the first place.
One rule with no exceptions: never fund payroll or a trading shortfall with long-term amortising debt. That isn’t financing, it’s postponement, and it shows up later as a permanent drag on everything you do.
What this looks like with real numbers
Back to our contractor. Say he’s doing $4.2 million in revenue with trailing EBITDA of $310,000.
Knock off $55,000 for maintenance capex and $55,000 for tax distributions, and he’s got $200,000 to service debt with. Divide by 1.25 and his ceiling for total annual debt payments is $160,000. He’s already paying $84,000 a year on equipment notes, so he’s got about $76,000 of room. Call it $6,300 a month.
What does he need? The machine is $340,000, plus $45,000 to freight it, install it and train the crew, so $385,000. The new contract adds $900,000 of revenue on that 55-day cycle, so another $135,000 of working capital. Add a bit of contingency and he’s looking at around $560,000.
Run $560,000 over seven years at eight percent and the payment is about $8,700 a month. That’s $105,000 a year against $76,000 of room. It doesn’t fit.
Which is the useful part. He wasn’t turned down by anybody, he found out himself, and now he can fix the structure instead of the answer. Put the $385,000 of equipment on a ten-year amortisation and it costs about $4,700 a month. Put the $175,000 of working capital on a revolving line where he only pays interest on what he draws, so maybe $8,000 across the year. Total new debt service lands near $64,000. Now it fits, with something left over.
Notice he didn’t need less money. He needed the money shaped differently. That’s a conversation worth having with a lender before you’re under pressure, not after.
Now try to break it
One more pass, and this is the one owners skip. Run the same sums on a bad year. Revenue down twenty percent. What happens to the coverage?
For our contractor, it goes under 1.0. The debt works in the plan and stops working in a downturn.
That doesn’t automatically kill the deal. Plenty of good businesses carry debt that would strain in a bad year. But it changes what he does next: keep that line of credit genuinely undrawn as a reserve rather than treating it as spending money, sit on a few months of payments in cash, or spec a cheaper machine and keep the difference.
Owners rarely get into trouble borrowing for a bad reason. They get into trouble borrowing an amount that only works if nothing goes wrong.
Two last things
Don’t borrow to your ceiling just because it’s there. The customer who pays sixty days late, the roof that goes in month four, the opportunity you couldn’t have predicted — all of that is far easier to absorb with unused capacity than without it. Borrowing seventy percent of what you could service is often the smarter call.
And when you do sit down with a lender, bring a number rather than a question. A specific amount, what it’s for, your coverage calculation, and an honest downside case. It changes the whole tone of the meeting. You’re not asking what’s available, you’re proposing a structure — and that tends to get you a faster answer, better terms, and a banker who actually understands your business the next time you need something.
