Starbucks carries an $8.1 billion shareholders' deficit and $3.4 billion of goodwill. It trips Layer 2 of the Goodwill Mask harder than anything this series has run — and it is clean. The instrument was built for roll-ups. This is not one.
The negative equity is buybacks and dividends, disclosed. A choice, not a concealment.
From the outside, Starbucks in fiscal 2025 has the shape of something this series exists to find. Shareholders' equity is negative $8.1 billion. Goodwill sits at $3.4 billion against tangible equity that is not merely thin but roughly negative $11.5 billion. Operating margin fell 710 basis points in a single year. Transactions declined for seven consecutive quarters. The company closed 627 coffeehouses in one quarter, took an $892 million restructuring charge, and is twelve months into a new chief executive and a turnaround with a slogan.
Run the arithmetic and the second layer of the scorecard fires as hard as it is possible to fire. Goodwill is not merely larger than tangible equity — tangible equity does not exist. On the face of the balance sheet, intangible assets are all that stands between this company and a hole.
And the signal is wrong. Not marginal, not arguable. It is measuring something the layer was never built to see, and the filing says so plainly enough that a hostile reader would find it before we did. That is the investigation.
Those three figures answer the question between them, and the third is the one that matters. Additional paid-in capital is $634.1 million. The deficit is not an accumulation of losses against contributed capital, and it is not the residue of a write-down. It is retained earnings, gone.
Starbucks was profitable in every year of the period. Net earnings were $4,124.7 million in fiscal 2023, $3,762.3 million in fiscal 2024, and $1,856.7 million in fiscal 2025 — the worst of the three, and still nearly two billion dollars. Over the same window the company returned $3.8 billion to shareholders in fiscal 2024 and $2.8 billion in fiscal 2025 through dividends and repurchases.
Return more to shareholders than you earn, for long enough, and retained earnings go negative regardless of how healthy the business underneath is. Nothing has been hidden and nothing has been impaired. The money left as dividends and buybacks, on purpose, with the arithmetic disclosed every quarter.
Negative book equity produced this way is a capital allocation choice. Negative book equity produced by unwritten acquisition goodwill is a concealment. They look identical in a screen.
The Goodwill Mask Scorecard was calibrated on Roper, Wayfair, B. Riley, and B&G Foods, and has since been run on Embracer, Unity, and Take-Two. Every one of them is a roll-up or something adjacent to one. Layer 2 asks whether goodwill exceeds tangible equity, and in a company that grew by acquisition the answer carries real information: the balance sheet is being held above zero by an asset that represents prices paid for other companies.
Starbucks did not grow by acquisition. It grew by opening stores. The question Layer 2 asks still has an answer here, and the answer still comes back yes, and it means nothing at all.
| Route to negative tangible equity | What produced it | What the layer should conclude |
|---|---|---|
| Unwritten acquisition goodwill | Debt-funded deals, stale recognition | Investigate. This is the mask. |
| Accumulated operating losses | The business does not earn | Investigate. Different problem, real problem. |
| Capital returned above earnings | Buybacks and dividends, disclosed | Exclude. A choice, not a concealment. |
Reading the statement of shareholders' equity separates the three in about ten minutes, and the scorecard did not previously require anyone to do it. That is the amendment this investigation produces, and it is stated at the end.
Operating income of $2,936.6 million against interest expense of $542.6 million. Interest cover of 5.41 times in the worst operating year of the decade, against an OECD threshold of 1 and a healthy-acquirer threshold of 2. Fiscal 2024 covered roughly ten times.
Goodwill and intangibles ran from $4,584 million at fiscal 2018, immediately after the East China buyout, to $3,535.7 million at fiscal 2025. Stale recognition requires goodwill nobody has been willing to touch. This has been coming down for seven years.
Nine months of fiscal 2026 generated $3,604.1 million of operating cash against $887.8 million of capital expenditure and $2,118.0 million of dividends. No burn, no dependence on outside capital.
And then the fact that settles it. Long-term debt fell from $14,575.9 million at the end of fiscal 2025 to $11,780.2 million at June 28, 2026 — roughly $2.8 billion, about a fifth, in three quarters. Repurchases of common stock stopped. Approximately $1.3 billion of the proceeds from the China transaction went to retiring senior notes through cash tender offers.
A company masking a dead core behind an unwritten balance sheet does not retire a fifth of its long-term debt in nine months. The mechanism that built the deficit has not merely paused. It has reversed direction, and the reversal is visible in a single line of the balance sheet.
The instrument introduced in Investigation #4 asks whether a company is spending its brand rather than building it. The silhouette is price rising while volume falls. Starbucks wore it, and then stopped.
| Quarter | Global comps | Transactions | Ticket |
|---|---|---|---|
| Q1 FY25 · Dec 2024 | −4% | −6% | +3% |
| Q2 FY25 · Mar 2025 | −1% | −2% | +1% |
| Q3 FY25 · Jun 2025 | −2% | −2% | +1% |
| Q4 FY25 · Sep 2025 | +1% | +1% | — |
| Q1 FY26 · Dec 2025 | +4% | +3% | +1% |
| Q3 FY26 · Jun 2026 | +7.9% | +4.2% | +3.5% |
Four consecutive quarters of comparable sales growth, led by transactions. In the United States, comparable sales rose 7.9% on transactions up 4.2% and ticket up 3.6%. Price up and volume up is the Gate 2 exclusion — the same one Unilever received, on the same grounds.
Which means the instrument returns opposite verdicts on this company depending on when it is run. Run it in March 2025 and Gate 1 fires. Run it today and Gate 2 excludes. Both are correct readings of the data available at the time, and that is worth saying out loud rather than discovering later: a flag is a statement about a period, not about a company.
A 710 basis point contraction in one year is the kind of number that looks like an emergency. The filing decomposes it, and the decomposition is the most useful thing in the document.
The remainder is investment the company chose and announced. Within store operating expenses, which rose 410 basis points as a share of company-operated store revenue, the filing attributes roughly 160 basis points to additional labor and roughly 90 basis points to increased marketing.
So of a 710 basis point fall, about 80 is a cost the company absorbed and the large majority is spending it elected to do, disclosed in advance, under a named strategy. That is not a company whose margin is being taken from it. It is a company spending margin on purpose and saying so.
It is tempting to call this a third state the instrument cannot name — neither cost shock nor brand harvest but deliberate reinvestment — and to write a new exclusion for it. We considered exactly that and decided against it, for a reason worth stating: the Brand Mask already routes this correctly. Price up, volume down, margin down declines to flag. The verdict was right. Only the label on the row was imprecise, calling the cause a cost problem when it can equally be a chosen investment.
An exclusion that never fires is not a safeguard. It is decoration, and this instrument already carries one such rule. We are not adding a second.
The strongest case against this clearance is structural, and it is the same shape as the one Investigation #4 faced.
On April 2, 2026, funds managed by Boyu Capital acquired 60% of Starbucks retail operations in China at a cash-free, debt-free enterprise value of approximately $4 billion. Starbucks retained 40% and continues to own and license the brand. Roughly 8,000 company-operated coffeehouses convert to a licensed model. The business deconsolidated in the third quarter of fiscal 2026 — which is why consolidated revenue fell 1% to $9.3 billion in a quarter when comparable sales rose 7.9%. The company received $3.1 billion of consideration and recorded a $536.3 million pre-tax gain.
The obvious challenge: if China was the weak limb, this clearance is flattered by removing it. Store mix moved from 53% company-operated a year ago to 33% today. A company can always look healthier with its hardest market taken out of the picture.
The shed changes the composition of the company. It does not manufacture four quarters of transaction growth in the United States, and it did not create the balance sheet position that made Starbucks look like a target in the first place.
Second clearance in five investigations, and worth naming plainly rather than leaving a reader to notice it. An instrument that only ever flags is not discriminating between companies; it is describing its author's expectations. The clearances are the evidence that the flags mean something.
One amendment, and it requires no recalibration because it cannot move any company across a threshold — it only requires that a question already being asked be answered correctly.
Where tangible equity is negative, read the statement of shareholders' equity and identify the cause: acquisition goodwill, accumulated losses, or capital returned above earnings. The third excludes. The first two proceed. Without this step the layer cannot distinguish a concealment from a dividend policy.
Price up, volume down, margin down currently reads as input-cost pass-through. It should read as cost shock or chosen investment — same routing, honest label. Where margin falls, read the company's own attribution before assuming which.
Neither change alters an output on any company in the calibration set. Relabeling cannot flip a verdict, and a decomposition step that informs an exclusion already available cannot either. Calibration is required when a change can move a company across a threshold. These cannot.
No terminal, no subscription. The three steps that produced this clearance are all public and all quick.
Not the balance sheet line — the statement. It shows what moved the number: earnings, dividends, repurchases, impairments. A negative equity figure means nothing until you know which.
Three years. Below 1 for three consecutive years is the OECD zombie definition. Above 5 and the balance sheet debate is usually academic.
Most companies attribute a margin move to named causes in basis points. Read it before deciding whether a contraction was suffered or chosen. It is frequently the most candid paragraph in the filing.
What could not be verified. A clearance carries the same duty of care as a flag, so here is the honest ledger. Advertising expense as a separate line was obtained from the filings for fiscal 2022 and 2023 only; the fiscal 2024 and 2025 figures were not read from the 10-K. The marketing conclusion above therefore rests on the company's own basis-point attribution within store operating expenses, which is a filing disclosure, and not on a spend series. The precise decomposition of the goodwill decline into impairment, disposal, and currency was not performed — the direction is established across seven years, the split is not. Fiscal 2024 interest expense is derived rather than read from a single annual line, though the exclusion holds at double the figure.
Read this carefully. This is research, not investment advice, and not a recommendation to buy, sell, or hold anything. The Goodwill Mask and Brand Mask produce flags for investigation — questions worth asking — not accusations, and not a verdict on any company's value or prospects. This investigation records a confirmed negative: Starbucks cleared every instrument applied to it. That is a finding about a method, not an endorsement of a security. Figures come from named filings — Starbucks Form 10-K for fiscal 2025 and Form 10-Q for the third quarter of fiscal 2026, together with quarterly earnings releases. Every verdict here is provisional and time-stamped, and should be re-run against the next annual filing. Do your own work, and consult a licensed professional before making any financial decision.