Debt has a public relations problem in both directions. One camp treats all borrowing as failure, and quietly starves good businesses of growth they could easily have afforded. The other treats capital as oxygen, and borrows its way into payments that eat the company alive. Both camps are running on feelings. Debt is neither virtue nor vice; it is a tool with a price tag, and the only interesting question is whether the tool earns more than the tag.
This piece is the framework for answering that question with arithmetic: the one-question test, worked examples in both directions, and the patterns on each side of the line, written by people who arrange financing and still tell a fair number of applicants that the honest answer is no.
The one-question test
Strip away the product names and every borrowing decision reduces to this: what does this specific money earn, and does it earn more than it costs? Good debt buys something that generates return, a machine that produces billable work, inventory that sells through at margin, a contract the business could not otherwise accept. Bad debt fills a hole that will still be a hole when the money is gone, and adds a payment on top.
Notice the test says nothing about the interest rate or the factor rate. Cheap money spent on nothing is bad debt at any price; expensive money that captures a larger return can be good debt despite its cost. The rate decides how high the bar sits, and the use of funds decides whether you clear it. Owners who fixate on the price of money while staying vague about its purpose have the analysis backwards.
The math when debt is good
Run the test on a concrete case with invented round numbers. A cabinet shop can take a $60,000 commercial job but needs a $24,000 CNC upgrade to hit the spec. It borrows $24,000 and repays $28,800 over the term: the money costs $4,800. The job's margin after materials and labor is $21,000. The debt bought $21,000 of profit for $4,800, the shop nets roughly $16,000 it otherwise had to decline, and the machine remains for every future job. That is good debt with the receipts to prove it.
Two features make the example work, and they generalize. The return was specific and near: a signed job, not a hope of jobs. And the return exceeded the cost with room to spare, enough that delays or overruns would not flip the sign. When either feature is missing, the same borrowing becomes a coin flip wearing a business plan. The same arithmetic drives the equipment decision guide and financing inventory at a discount, which are this test applied to two specific purchases.
The math when debt is bad
Now the mirror image. A shop is losing $3,000 a month because two long-standing customers left and nothing replaced them. The owner borrows $30,000 to, in the phrase underwriters hear daily, get through a rough patch. Ten months later the money is gone, spent on the same rent and payroll that were unaffordable before, the customers are still unreplaced, and the business now carries a payment besides. The hole is deeper by exactly the cost of the money, and the rough patch turned out to be the business model.
That is the signature of bad debt: the money maintains a loss instead of buying a change. Nothing about the borrowing altered what the business earns; it only moved the reckoning and raised its price. The same logic flags borrowing to cover a chronic leak that operations could fix for free, and stacking new advances to service old ones, which is the same trade at compounding speed. A bridge needs a far bank: defined money arriving on a date. Bridging to a hope is not bridging, it is sinking slowly.
The gray zone, where most real decisions live
Plenty of borrowing is neither a signed contract nor a chronic loss, and the test still works if you are honest with its inputs.
- A payroll bridge is good debt when the gap has a date and a source, a slow-paying invoice, a season that turns, and bad debt when the gap is simply what the business does now.
- Marketing spend is good debt only as far as your own history prices it: if past campaigns returned a known multiple, borrowing scales a machine; if you have never measured one, you are borrowing to run an experiment, and experiments deserve the smallest budget that answers the question.
- A hire funded by debt is a wager the person produces more than payroll plus the cost of money, which is why it deserves the hire-or-wait framework rather than optimism.
- Refinancing existing debt is good precisely when it lowers the total cost or fits the payment to real cash flow, and bad when it merely resets a clock while the balance grows.
The second gate: can the cash flow carry it?
Passing the earn-versus-cost test is necessary and not sufficient, because returns arrive on their schedule and payments arrive on theirs. A profitable use of funds can still wreck a business if the remittance lands weekly while the revenue lands in ninety days. So every borrowing decision clears two gates: does it earn more than it costs, and can the account carry the payment through the worst realistic stretch before the return shows up?
The second gate is checkable tonight. Put the proposed payment against your actual deposits with the payment affordability checker, using your weakest recent month rather than your best. If the numbers only work in the optimistic version, the structure is wrong even if the idea is right: a longer term, a smaller amount, or repayment that flexes with revenue may fit the same purpose. And when offers exist, price them side by side with the offer comparison tool before judging any of them; the framework for that call is in is this offer too expensive.
A broker's honest summary
We are in the business of arranging capital, so weigh the source, and then weigh this: the clients who thrive with debt all sound the same on a first call. They can say what the money buys, what that purchase earns, when the earnings arrive, and what payment their slowest month can carry. The clients who struggle borrowed against discomfort instead of a plan, and no product on any funder's menu fixes that.
If your use of funds passes both gates, borrowing is not a compromise; it is how businesses that lack rich uncles grow, and timing it well is its own decision: take the money now, or wait. If it fails either gate, the strongest move is the one we put in writing: sometimes we tell clients not to take funding, and the reasons in that piece are this framework, applied.
Frequently asked questions
Is expensive short-term funding always bad debt?
No, and it is never automatically good either. Cost raises the bar the return must clear: expensive money that captures a specific, larger, near-term return can be a sound trade, while cheap money spent maintaining a loss is still a loss with a payment attached. Judge the use first, then check whether this particular price still leaves a worthwhile margin.
How do I calculate whether borrowing is worth it?
Put three numbers on paper: the total cost of the money, which is everything you will repay minus what you receive; the dollar return of the specific use, priced from your own history rather than hope; and the payment against your weakest recent month's cash flow. Worthwhile borrowing shows a return comfortably above the cost and a payment the slow month still carries.
Is it good debt to borrow to pay off other business debt?
It can be, when consolidating genuinely lowers the total repaid or replaces a payment structure that fights your cash flow with one that fits it. It turns bad when a new advance mostly services old ones while balances grow, a pattern called stacking that compounds cost quickly. The test is one honest number: total owed and total cost, before versus after.
What does debt service coverage mean for this decision?
It is the second gate in ratio form: what your business earns in a period divided by what it would owe in debt payments over the same period. Lenders compute a version of it before offering terms, and running it on yourself, with a slow month's earnings rather than an average, tells you whether a proposed payment fits before anyone else's underwriting weighs in.