What the ratio asks
Times interest earned asks whether the business is earning its interest bill, and by how wide a margin. Coverage of 4.0× means operating profit is four times the interest charge, so three-quarters of EBIT could disappear before interest stopped being covered.
EBIT is the right numerator because interest is deducted before tax. Using net income would be circular — net income is already after the interest you are trying to test — and would understate coverage by the amount of the tax charge. This is also why the ratio is sometimes written as (net income + interest + tax) ÷ interest, which is the same thing rebuilt from the bottom of the income statement.
The measure's blind spot is that EBIT is an accrual number, not cash. Depreciation has already been deducted from it, and depreciation is not a payment. That is why the EBITDA-basis ratio exists: adding back non-cash charges gives a closer approximation to the cash actually available to meet interest in the short run. Both figures are worth having, and the gap between them tells you how asset-heavy the business is. The EBITDA version is more generous, and it is generous in a way that is only honest over a short horizon — the assets being depreciated will eventually need replacing, in cash.
The arithmetic of headroom
Because interest is fixed within a period, coverage moves in exact proportion to EBIT. Halve EBIT and coverage halves. That linearity makes the headroom arithmetic simple and worth doing explicitly.
The interest you can carry at a covenant of k is EBIT ÷ k. Subtract what you already pay and you have the additional interest capacity — the extra annual charge the current earnings could absorb, whether from new borrowing or from a rate rise on existing floating-rate debt.
The EBIT you need at a covenant of k is k × interest. The percentage fall from today's EBIT to that level is your real cushion: 1 − (k × I) ÷ EBIT. This is the number to plan with, because it converts an abstract multiple into an operating question — could this business lose that much of its operating profit in a bad year?
Note what a coverage covenant does not see. It responds to interest rates, so a borrower can refinance identical debt at a higher rate and breach a coverage test without changing a single thing about its balance sheet. Conversely a leverage covenant responds to the amount of debt and is blind to the rate. That is exactly why credit agreements usually include one of each: neither catches what the other misses.
One definitional trap: use gross interest unless your agreement says otherwise. Netting interest income against interest expense raises coverage, and the income may be earned on cash that is not available or not repeatable. Check also whether capitalised interest — interest rolled into the cost of an asset under construction — is included; it is a real cash payment even though it never appears in the income statement.
Worked example: $900M of EBIT against a $220M interest bill
Take the defaults, in millions: EBIT $900, depreciation and amortisation $350, interest expense $220, and a covenant minimum of 2.50× on the EBIT basis.
- EBITDA. 900 + 350 = $1,250M.
- Times interest earned. 900 ÷ 220 = 4.0909×.
- EBITDA coverage. 1,250 ÷ 220 = 5.6818×. The 1.59-turn gap between the two is exactly 350 ÷ 220, the D&A expressed in turns of interest.
- Headroom. 4.0909 − 2.50 = 1.5909 turns.
- Interest permitted at the covenant. 900 ÷ 2.50 = $360M, so there is 360 − 220 = $140M of additional annual interest capacity.
- EBIT required at the covenant. 2.50 × 220 = $550M.
- Permitted EBIT decline. (900 − 550) ÷ 900 = 38.89%.
Steps 5 and 7 are two views of the same cushion, and each is useful for a different question. The $140M figure sizes how much more debt service the company could take on: at a 7% coupon that is roughly $2bn of additional borrowing, though a leverage covenant would almost certainly bind first. The 38.89% figure sizes the downturn the covenant can tolerate. A treasurer facing floating-rate debt should watch step 5, because a 300 basis point rise would consume the whole $140M on 140 ÷ 0.03 = $4.67bn of floating-rate debt, without any change in trading.
How to read the number
Below 1.0× the company is not earning its interest at all. Whatever the balance sheet looks like, the income statement is not paying the lenders, and the shortfall must be coming from cash reserves, asset sales or new borrowing.
1.0× to 2.0× is thin. Interest is covered, but almost nothing is left for tax, reinvestment or principal repayment. Most credit agreements set their minimum somewhere in the 2.0× to 3.0× range precisely to keep borrowers out of this band.
3.0× to 6.0× is a normal range for a leveraged but sound corporate. Above 6.0×, interest is not the constraint on the business and the ratio stops being informative — a company at 25× coverage and one at 40× are equally unconstrained.
Interpret the level against the volatility of EBIT. Coverage of 3.0× on contracted, inflation-linked revenue is far safer than 5.0× on a cyclical order book. The permitted-EBIT-decline figure is the right way to make that comparison, because it states the cushion in the units the business actually varies in.
Read the EBIT and EBITDA versions together. A wide gap means large non-cash charges, which flatters short-term coverage and understates the long-term capital burden. For a business whose assets need replacing on a predictable cycle, the EBIT-basis figure is closer to the truth, and a measure that includes capital expenditure, such as debt service coverage, is closer still.
Coverage by the share of EBIT consumed by interest
| Interest as % of EBIT | Times interest earned | EBIT left after interest | Permitted EBIT fall at a 2.50× covenant |
|---|---|---|---|
| 10% | 10.000× | 90% | 75.0% |
| 15% | 6.667× | 85% | 62.5% |
| 20% | 5.000× | 80% | 50.0% |
| 25% | 4.000× | 75% | 37.5% |
| 30% | 3.333× | 70% | 25.0% |
| 33⅓% | 3.000× | 66.7% | 16.7% |
| 40% | 2.500× | 60% | 0.0% |
| 50% | 2.000× | 50% | already breached |
The 40% row is where coverage equals the covenant exactly, so the permitted decline is zero. The default case sits between the 20% and 25% rows: interest of $220M is 24.4% of $900M of EBIT, giving 4.09× coverage and a 38.9% cushion.
Mistakes that distort interest coverage
- Using net income instead of EBIT. Net income is already after interest, so the ratio becomes circular and understates coverage by the tax charge as well.
- Netting interest income against interest expense. Unless the covenant defines it that way, use gross interest. Interest income can be earned on cash the business cannot spare.
- Omitting capitalised interest. Interest rolled into the cost of an asset under construction is a cash payment that never touches the income statement, and lenders normally include it.
- Reading EBITDA coverage as though it were cash coverage. EBITDA is before capital expenditure, cash tax and working capital. An asset-heavy business with 8× EBITDA coverage may have very little genuinely spare.
- Testing a single period on a seasonal or cyclical business. Use twelve months, and test the trough as well as the average.
- Forgetting floating-rate exposure. Coverage falls when rates rise even if the business is unchanged. Model the covenant at the top of your rate scenario, not at today's rate.
Coverage next to leverage and debt service
Credit analysis rests on two questions that a single ratio cannot answer together. Can the borrower afford the debt? is a coverage question, and times interest earned is its simplest form. Is there too much debt? is a leverage question, answered by net debt to EBITDA or by balance-sheet ratios such as debt to equity. The two can diverge sharply: cheap debt gives comfortable coverage at high leverage, and expensive debt gives poor coverage at modest leverage. Most credit agreements therefore test both.
Debt service coverage is the stricter member of the coverage family, because it includes scheduled principal repayment as well as interest. For amortising debt, that is the test that actually predicts a missed payment; a company can cover interest three times over and still fail to make a bullet repayment.
On the equity side, the same fixed charges that this ratio tests are what amplify earnings, and the degree of financial leverage measures that amplification directly. There is a clean relationship worth knowing: DFL for a company with no preferred stock is TIE ÷ (TIE − 1). Coverage of 4× gives a DFL of 1.333; coverage of 2× gives 2.000. Where operating leverage is also high — see the degree of operating leverage — the two multiply, and a coverage ratio that looks adequate can sit on top of very volatile EBIT.
Key terms
- Interest coverage ratio
- Another name for times interest earned. Some analysts reserve the term for the EBITDA-basis version, so check which definition is in use.
- Capitalised interest
- Interest incurred during construction of an asset and added to its cost rather than expensed. It is paid in cash but does not appear in interest expense.
- Maintenance covenant
- A covenant tested every period regardless of any action by the borrower, as opposed to an incurrence covenant tested only when the borrower does something specific such as raising new debt.