How Much Business Debt Is Too Much? Three Tests Before You Borrow Again

On this page
Business debt is too much when a realistic bad year would break any one of these three tests:
- Coverage. Cash available for debt service, in a downside year, is less than 1.25 times your total principal and interest payments.
- Leverage. Total funded debt is more than 2.0 times downside EBITDA.
- Liquidity. Cash above your operating reserve covers fewer than three months of debt service.
Those thresholds are the decision rule this article uses for an established service business. They're not a lender standard, and your bank's covenants may be looser or tighter. The point is to set the limit on your own downside numbers before a lender sets it on their base case.
The question owners usually ask is "will the bank approve this?" That's the wrong test. A bank can approve a loan that still leaves the business fragile. SBA's own criteria in 13 CFR 120.150 say loans must be "so sound as to reasonably assure repayment" and let the lender weigh the applicant's earnings or cash flow. That's a lender's view of repayment, not an owner's view of whether the business can still invest, hire, and survive a bad year while paying.
Why the downside year is the only year that matters
In a good year almost any reasonable debt load works. The damage shows up when revenue dips 15%–25% and the payments don't. Principal and interest payments are fixed. Payroll is nearly fixed in the short run. The cushion between them is your operating margin, and that's what shrinks first.
So every test below uses a downside case. Build it by cutting revenue by an amount you've actually experienced or could plausibly experience, such as losing your largest client, then flowing that loss through at your gross margin, not your net margin. A service firm with 45% gross margin that loses $1M of revenue loses about $450,000 of gross profit, and most of that hits EBITDA directly unless you cut people.
Test 1: downside debt service coverage
Cash available for debt service (CADS) =
EBITDA − Capital spending − Cash taxes or tax distributions
DSCR = CADS ÷ Total annual principal and interest payments
Use every required payment: term loans, equipment financing, seller notes, and any scheduled line paydowns. Leave out none of them because they sit in different places on the balance sheet.
The detailed mechanics, including how lenders adjust the numerator, are in the DSCR guide. For this decision you need just two numbers: the base-case DSCR and the downside DSCR. If the downside is below 1.25x, the debt is too much for the business's risk profile, regardless of what the base case says.
Test 2: leverage against downside earnings
Leverage = Total funded debt ÷ Downside EBITDA
This test asks a different question: how many years of downside earnings would it take to repay everything? DSCR depends on loan terms, so a long amortization can make heavy debt look affordable. Leverage ignores the schedule and looks at the total.
For a service business with few hard assets, 2.0x downside EBITDA is a sensible ceiling. Lenders may lend more against strong recurring revenue, and an acquisition model might justify more with a clear payback plan, as in the acquisition debt-capacity model. But for routine operating borrowing, crossing 2.0x on downside earnings means one bad year leaves the company owing more than it can realistically pay down in two.
What about debt-to-equity?
Debt-to-equity is the ratio owners find most often online. For most service businesses it's a weak guide. S corporations and partnerships often distribute most of their profit, so book equity stays small even in a healthy company, and the ratio looks alarming. A company that has retained everything looks conservative even if its earnings are falling. Earnings-based tests describe repayment ability better than balance-sheet equity does in a business whose main asset is its people.
Test 3: cash above reserve
Liquidity months = (Unrestricted cash − Operating reserve) ÷ Monthly debt service
This is the test owners skip and the one that catches problems first. A company can pass coverage and leverage on an annual basis and still miss a payment in a slow quarter if there's no cash cushion above the reserve it needs for payroll. Set the reserve with the risk-based reserve formula, then require at least three months of debt service on top.
An undrawn line of credit doesn't count here. Lines get reduced after weak quarters, which is exactly when you'd want to draw.
These three tests take an afternoon if your monthly numbers are clean, and much longer if they aren't. If yours aren't, Bennett's free Leak Check is the faster first step: it flags whether margin, payroll, or tax timing is squeezing the cash you'd be borrowing against.
A hypothetical $8M firm asking for $600,000
Assume a hypothetical $8M IT services firm with $960,000 of EBITDA, $60,000 of annual capital spending, and $220,000 of tax distributions. It already has a $1.2M term loan with $300,000 of annual principal and interest. It wants a new $600,000 five-year loan at an illustrative 9%, about $149,500 a year. Cash is $500,000; its operating reserve is $450,000.
The downside case cuts EBITDA by 25% to $720,000, with tax distributions falling to $150,000.
| Test | Threshold | Base case | Downside case | Result |
|---|---|---|---|---|
| DSCR | ≥ 1.25x | $680,000 ÷ $449,500 = 1.51x | $510,000 ÷ $449,500 = 1.13x | Fails downside |
| Leverage | ≤ 2.0x | $1.8M ÷ $960,000 = 1.9x | $1.8M ÷ $720,000 = 2.5x | Fails downside |
| Liquidity | ≥ 3 months | $50,000 ÷ $37,460 = 1.3 months | Same or worse | Fails |
The bank might well approve this loan on the base case: 1.51x coverage and 1.9x leverage look fine. But in one bad year, the business couldn't cover its payments with a cushion, would owe two and a half years of earnings, and has barely a month of payments in spare cash.
Now try $300,000 instead, about $74,700 a year:
| Test | Downside result at $300,000 | Result |
|---|---|---|
| DSCR | $510,000 ÷ $374,700 = 1.36x | Passes |
| Leverage | $1.5M ÷ $720,000 = 2.08x | Just over |
| Liquidity | $50,000 ÷ $31,225 = 1.6 months | Fails |
The smaller loan passes coverage and nearly passes leverage. Liquidity still fails, so the right move is to build about $45,000 more cash above the reserve first, or fund part of the project from the next two quarters of cash flow. That's a slower plan, and a much safer one.
What to do when the tests fail
If the tests fail, there are four levers, in order of preference:
- Borrow less and fund the rest from cash flow over a longer timeline.
- Lengthen the term on new or existing debt. It improves DSCR but not leverage, so it's only half a fix.
- Raise EBITDA before borrowing. Pricing and gross margin usually move faster than overhead. A two-point margin gain on $8M is $160,000 a year.
- Replace debt with equity or a seller-friendly structure if the borrowing funds an acquisition.
Avoid the fifth option owners reach for under pressure: expensive short-term money. A merchant cash advance fixes a liquidity test for a few weeks and makes all three tests worse afterward. If the debt is mostly a revolving line that never clears, separate the permanent balance from the seasonal one using the line of credit vs term loan matching rule.
Also remember who carries the downside. If you've personally guaranteed the debt, failing these tests is a personal balance-sheet problem as well as a company one. The guarantee exposure formula shows how much.
Run the three tests on your next borrowing request
Take last year's EBITDA, cut it by the revenue loss you'd least like to experience, and flow that through at gross margin. Add every required debt payment, including the new loan. Then calculate downside DSCR, debt to downside EBITDA, and months of debt service above your reserve. If any test fails, change the size, term, or timing before you sign, not after. Bennett's fractional CFO services include building this downside model alongside the lender package, and if you'd like a fast read on which line of the P&L is limiting your borrowing capacity, start with a Profit & Tax Leak Check.
Frequently asked questions
What DSCR should a service business hold in a bad year?
This article's decision rule is at least 1.25x on downside cash available for debt service. Lender covenants vary, so treat 1.25x as the owner's own floor rather than a bank requirement.
Why use downside EBITDA instead of last year's EBITDA for leverage?
Because debt payments don't fall when earnings do. In the hypothetical, $1.8M of debt is 1.9x base EBITDA but 2.5x after a 25% earnings drop, which is the year the leverage actually matters.
Is debt-to-equity a good way to judge a service company's debt?
Usually not. S corporations and partnerships often distribute most of their profit, so book equity stays small and the ratio looks alarming even in healthy firms. Earnings-based tests describe repayment ability better.
Does an undrawn line of credit count toward my cash cushion?
Not for the liquidity test. Lines can be reduced after a weak quarter, which is when you'd need them. Count only unrestricted cash above your operating reserve.
How do I build a downside case for debt planning?
Cut revenue by a loss you could plausibly face, such as your largest client, and flow it through at gross margin. A firm at 45% gross margin losing $1M of revenue loses about $450,000 of gross profit.
Does a longer loan term solve a debt problem?
Only partly. Stretching the term lowers annual payments and improves DSCR, but total debt relative to earnings stays the same, so the leverage test doesn't improve.
What does SBA require to show a loan can be repaid?
13 CFR 120.150 says SBA loans must be so sound as to reasonably assure repayment and lets the lender consider the applicant's earnings or cash flow, among other factors.
What should I do if my borrowing request fails the debt tests?
In order: borrow less and fund the rest from cash flow, lengthen the term, raise EBITDA through pricing or gross margin before borrowing, or use equity or seller-friendly structures for acquisitions. Avoid expensive short-term money.