Balance sheet
Every screener leads with debt-to-equity, and on a company that has bought back enough stock, that ratio returns a negative number and tells you nothing. The answer is not another ratio. It is a calendar.
9 min read· Corva Research
The short version
A company has too much debt when an obligation comes due before the cash to repay it arrives. That is the whole definition, and notice that it contains a date. Every popular leverage ratio drops the date.
Which is why the ratio-only approach produces two predictable errors in opposite directions. It calls a business with a wall of debt maturing next spring safe, because the ratio looks moderate. And it calls a business with nothing due for six years dangerous, because the ratio looks high.
Below: the four ratios, what each one is actually asking, the case where the most popular one returns a meaningless answer, and the table in the debt footnote that replaces it.
There is no threshold that holds across industries. A regulated utility with contracted revenue can carry leverage that would end a semiconductor company inside one cycle, and both facts are ordinary.
What does travel is a question in two halves. Can the company cover its interest bill from operating profit, and can it repay or refinance each maturity as it lands? The first is a rate question, the second is a timing question, and they fail independently. A company can cover interest twelve times over and still default because it could not roll a bond in a shut market.
So run the ratios first, because they are quick. Then find out when the money is owed.
These four appear on every stock screen and in nearly every article on the subject. Each asks something narrower than its reputation suggests.
Three of those degrade gracefully. They get less informative in edge cases, but they keep returning a number you can compare. The fourth does something worse than degrade.
Book equity is not the value of the business. It is a residue: what the company raised, plus what it has earned and kept, minus what it has paid back out. Buy back enough of your own stock for enough years and the residue goes below zero while the business underneath is unchanged.
McDonald's is the clean case. At 31 December 2025, per its Form 10-K filed 24 February 2026, the equity section reads:
The treasury line alone exceeds every dollar of profit the company has ever retained, by $9,034m. That is what pushes the total below zero. Nothing about it is a loss.
Now run the ratio. Long-term debt on that balance sheet is $39,973m, so debt-to-equity is 39,973 divided by -1,791, or -22.3x. A screen sorting descending on debt-to-equity will not show McDonald's at the top of the risk list. It will show the company somewhere below zero, or blank, or filtered out entirely.
The ratio did not report a risky company. It reported a broken denominator, and those look nothing alike.
The same year, McDonald's earned $8,563m of net income on $26,885m of revenue and generated $10,551m of cash from operations. Interest coverage, operating income of $12,393m over interest expense of $1,582m, is 7.8x, the Aa2/AA band on Damodaran's table. The headline leverage ratio is the only thing in the file suggesting difficulty.
It is not an isolated case. Starbucks reported a total shareholders' deficit of $(8,096.6)m at 28 September 2025 in its FY2025 Form 10-K filed 14 November 2025, while earning $1,856.4m of net income. AutoZone reported a deficit of $(3,414.3)m at 30 August 2025 in its FY2025 Form 10-K filed 27 October 2025. Starbucks retires the shares it repurchases rather than holding them in treasury, so the reduction lands in retained earnings instead, as an accumulated deficit of $(8,272.5)m. Same act, different line, same broken ratio.
Substituting market capitalisation for book equity does not repair it. That turns a balance-sheet measure into a sentiment measure, which falls fastest exactly when leverage matters most. Drop the ratio and ask the question it was standing in for.
Companies do not fail because a ratio crossed a line. They fail on a date, when a specific instrument matures and the refinancing market is closed to them.
So two companies at the same leverage can be in completely different positions. A business at 3.0x net debt to EBITDA with nothing due for six years is safer than one at 1.5x with its revolver expiring in eleven months. The second has to persuade a bank inside a year.
The information is already published. US filers state the maturity of each debt issue in the balance sheet or a note, per Regulation S-X Rule 5-02.22, which asks for "the date of maturity, or, if maturing serially, a brief indication of the serial maturities". In practice it arrives as a contractual maturities table in the debt footnote: five named years, then a single "thereafter" line.
It takes about a minute to find. Search the 10-K for "thereafter" and read the table it sits in.
Here is McDonald's own table, from note 8 of the same filing. The figures are stated before fair value adjustments and deferred debt costs, which is why the total is $40,145m rather than the $39,973m carried on the balance sheet.
| Due in | $ millions | % of total |
|---|---|---|
| 2026 | 0 | 0.0% |
| 2027 | 3,201 | 8.0% |
| 2028 | 5,166 | 12.9% |
| 2029 | 3,637 | 9.1% |
| 2030 | 3,011 | 7.5% |
| Thereafter | 25,130 | 62.6% |
| Total | 40,145 | 100.0% |
Nothing matures in the next twelve months, and 62.6% of the debt falls after 2030. Source: McDonald's Corporation Form 10-K, FY2025, Debt Financing note, filed 24 February 2026. Percentages computed from the figures shown.
Read that against the ratio. Net debt is $39,199m after $774m of cash, and EBITDA computed as operating income plus $2,199m of depreciation and amortisation is $14,592m, so net debt to EBITDA is 2.7x. On a screen, 2.7x with negative equity is a flag. On the calendar, the company owes nothing this year and has five years before the bulk of it lands.
The shape matters as much as the total. A flat, spread schedule is a company that has managed its issuance. One year carrying four times its neighbours is a maturity wall, and that year is the date the business is really being underwritten against.
The framework has a hole, and it is worth opening before anyone else does. The schedule tells you when principal is due. It is silent about two things that can arrive sooner.
Floating-rate debt reprices without maturing. A bond that matures in 2034 but pays a floating coupon costs more every time the reference rate rises, and no line in the maturity table moves. McDonald's is lightly exposed here, with $1,298m of floating-rate US dollar debt at 5.1% against a fixed US dollar book of $23,233m at 4.4%. A company where the floating share is half the total has a real interest-cost problem that a six-year runway does nothing to fix.
Covenants can accelerate a maturity regardless of the schedule. A covenant breach can make debt due immediately, which turns a 2032 obligation into a 2026 one. McDonald's addresses this directly and favourably in the same note: there are "no provisions in the Company's debt obligations that would accelerate repayment of debt as a result of a change in credit ratings or a material adverse change in the Company's business", though "certain of the Company's debt obligations contain cross-acceleration provisions". Most companies are not that clean, and the covenant terms are in the credit agreement exhibits rather than the footnote.
A third qualification sits in the footnote to the footnote. That $0 for 2026 includes $1.5B of short-term obligations, $798m of commercial paper and $725m of current maturities, reclassified as long-term because a $4.0B credit line expiring in June 2028 supports them. The zero is genuine under the accounting. It also rests on a facility with an expiry date of its own.
So read the schedule, then read what can move it.
Six steps, in this sequence, because each changes what the next one means.
Steps one and two take two minutes and tell you the temperature. Steps four and five take ten and tell you the date. Most published answers to this question stop after step two.
The calendar is better than the ratio. It is not a verdict, and it has limits worth stating plainly.
None of that argues for going back to a single ratio. It argues for reading the debt note and the lease note together, roughly fifteen minutes on any large filer. Our guide to reading a 10-K in 30 minutes covers where those notes sit.
There is no cross-industry number, and the more useful answer is that the ratio is the wrong tool for several large, healthy companies. Where book equity is negative, as at McDonald's, Starbucks and AutoZone, the ratio returns a negative or blank value that cannot be compared to anything. Use interest coverage and the maturity schedule instead.
Because book equity records cash raised and profit retained, less capital returned. A company that repurchases more stock over its life than it has retained in earnings ends up with a deficit. McDonald's treasury stock stood at $79,316m against $70,282m of retained earnings at the end of 2025. The gap is the deficit, and it is a distribution history rather than a loss.
In the notes to the consolidated financial statements, in the debt or borrowings note. It is usually the last item in that note, or a footnote to the debt summary table, as it is at McDonald's. Searching the document for "thereafter" finds it faster than scrolling.
No. Contractual maturities are principal only. The interest cost is on the income statement, and if you want the combined figure you have to build it yourself from the coupon and rate disclosures in the same note.
No, and treating it as one is why the ratio-first approach misfires. Debt at a fixed rate, spread across many years, against repeatable revenue, is a financing choice. The same amount concentrated into one near maturity against cyclical revenue is a different situation, at the identical ratio.
A year in the schedule carrying far more than the years around it, usually because several issues were sold at once. It concentrates refinancing risk on one date, which becomes the year the company has to persuade the credit market whatever its ratios say by then.
Corva pulls the maturity schedule, the fixed and floating split, the lease obligations and the coverage ratios out of the filings, and flags where book equity makes a leverage ratio unusable. Figures are computed from the filed statements and cross-checked against SEC EDGAR. Where a number cannot be found, it says so rather than inventing one.
Research a company free →Live financials on any listed company are free. No card.
Related: How to read a 10-K in 30 minutes · Gross margin decline: 4 causes, and the one nobody checks · all posts
Corva is a research tool, not a broker or investment adviser, and nothing here is a recommendation to buy, sell or hold any security. McDonald's Corporation, Starbucks Corporation and AutoZone, Inc. are used here as documented examples of negative book equity and of published maturity disclosure, and for no other reason. Figures are taken from each company's most recent Form 10-K as cited above; ratios and percentages are computed from those figures and EBITDA is a computed non-GAAP figure, operating income plus depreciation and amortisation. Verify anything you intend to act on against the primary filing. See terms and disclaimer.