A brand is a promise a company has been keeping for decades. It can also be quietly drawn down to hold a quarter — and for a while, the numbers look better, not worse. This is a new instrument for catching that, the footnote that gives you an early warning, and the company we pointed it at first. Unilever cleared it.
The reservoir was not being drawn down. Volume never went negative.
A brand is a promise a company has been keeping to people, usually for decades, mostly made by employees who have long since left.
That promise has a cash value. It shows up as the price a customer will pay without checking a rival, the demand that arrives without being bought, the customer who comes back, and the quiet steadiness of revenue in a bad quarter. Four fingerprints on a P&L, none of them labeled.
It can also be drawn down. A company under pressure can raise price faster than it improves the product, cut the marketing that earns the price, thin the pack, simplify the recipe, let the meaning go vague — and the numbers will look like discipline. Margin rises. Cost falls. The quarter lands.
The reservoir is finite, and the bill arrives late — long after the people who drew it down have been promoted or paid out.
The people holding the bill are the employees whose jobs go when volume finally cracks, the customers who kept paying a premium for something quietly getting worse, and the shareholders who bought a growth story that was actually a liquidation.
A tree under serious threat can produce an unusually heavy crop of seed. Forestry sources describe heavy seed production being triggered in some species as a reaction to stress — drought, defoliation — and ask exactly the right question: why would a stressed tree run down the reserves that could fund its recovery? Because the crop is the last throw. The tree produces the best-looking harvest in years, funded by the reserves meant to build next season's roots. An orchardist who reads that crop as health plants more of the same tree.
We call it the stress crop. The record year is the tell, not the reassurance.
Here is the strangest fact in corporate accounting.
A brand a company built itself cannot go on its balance sheet. IAS 38 and ASC 350 forbid capitalising internally generated brands. Coca-Cola's brand — arguably the most valuable commercial asset ever created — appears nowhere in Coca-Cola's assets. It was never bought, so it cannot be booked.
But a brand a company bought must be booked, at the price paid. Often with an indefinite life, meaning it is never amortised. It simply sits there at cost, year after year.
The brands a company is destroying by neglect are invisible. The brands it bought are carried at what it paid — forever, until someone admits they are worth less.
This is why "we can't measure brand" survives as an excuse. It is not that brand is unmeasurable. It is that it appears on the P&L constantly — as margin, as acquisition cost, as retention, as resilience — just never with its name on it.
There is one exception, and it is the engine of everything that follows.
Acquired brand intangibles and goodwill must be tested for impairment every year, whether or not anything looks wrong. And — this is the part that gets skipped — companies must disclose how much cushion is left on the ones they did not write down.
Under US GAAP, it arrives as headroom bands, as above. Under IFRS, it arrives differently and says the same thing: IAS 36 requires disclosure of the reasonably possible changes in key assumptions that would reduce headroom to nil. In practice — revenue would have to fall this far, or the discount rate rise this much, before this brand fails its test.
A published, quantified, forward-looking list of which brands are closest to being admitted as worth less than they are carried at.
Two companies, different sectors, same move. VF Corporation attributed its Supreme brand impairments to non-operating factors including higher interest rates and currency movements. Kraft Heinz attributed its 2023 impairments of Maxwell House, Cool Whip and two other brands primarily to an increase in the discount rate.
Rates genuinely do move valuations, and neither statement is false. But it is the explanation that converts a demand event into a macro event. When it appears alongside years of volume and share decline, it is worth asking who benefits from that framing.
In 2015, 3G Capital merged Kraft into Heinz and applied zero-based budgeting — every expense justified from scratch, every year. Operating margins were driven above 25%. Wall Street admired it. Competitors feared it.
It was published more than a year early. The 2017 annual report disclosed that there was not a significant excess of fair values over carrying values. In plain English: these brands are carried at close to what we think they are worth, and there is very little room. The room ran out in Q4 2018. The share price fell 27% the next day, and an SEC subpoena over procurement accounting arrived the same week.
Read the sequence. Margin expanded because brand investment was cut. Revenue held on price. It looked like discipline for three years. Then the balance sheet was forced to tell the truth all at once, and value that had been quietly consumed showed up as a single line. Impairments continued: roughly $1.2bn across 2019, more in 2020, $152m in 2023, $593m in 2024.
That is a stress crop. The harvest was real. It was funded by the roots.
The instinct is to look for price rising while volume falls — more money from fewer customers. That instinct is right, and on its own nearly useless, because at least three completely different situations produce it.
| Price | Volume | Gross margin | What it actually is |
|---|---|---|---|
| ↑ | ↑ | ↑ or flat | A healthy brand. Strength converting to margin. |
| ↑ | ↓ | ↑ or flat | A stress crop. The reservoir is being spent. |
| ↑ | ↓ | ↓ | Input-cost pass-through. Chasing a cost, not harvesting a brand. |
| ↑ | ↓ structural | share ↑ | Category decline. Strong brand, shrinking market. |
A company harvesting brand equity expands margin. A company chasing an input loses margin. Both show price up and volume down. Only one is a mask.
I built this instrument to catch companies spending their brands. Two companies nearly proved it caught the wrong ones.
Textbook flag on the surface. Q3 2025 pricing +8% against volume/mix −4.6%. Q1 2025 chocolate organic revenue +10.1% on volume/mix −5.7%. A cruder screen convicts on the spot.
Three facts stop it. Cocoa rose nearly threefold between late 2023 and early 2025 to a 60-year high. Market share held. And adjusted gross profit fell 11.3% in Q2 2025. Price rising while margin compresses is a company chasing an input and losing ground as it goes — a bad season, not a spent reservoir.
The sharpest divergence in the whole set: net price realization of roughly 8.8% in FY2023 against volume declines of 7.5%, 9.7%, 9.9% and 10.2% across 2021–24. Revenue net of excise crawled from $19.8bn to $20.4bn over five years, on pure price.
And yet Marlboro's share of the premium segment rose 0.7 points to 59.4%. The brand is gaining position inside a shrinking market — a serious business problem, and a completely different one.
An instrument that flags every strong brand in an inflationary year, or every strong brand in a declining category, is not an instrument. It is a mood. These two exclusions are what make the rest of it mean anything.
Does brand meaning drive the price premium? If not — commodity, regulated, pure cost competition — the instrument does not apply.
Is the volume decline explained by regulation, secular category decline, deliberate scarcity, or input-cost pass-through? If so, route out.
Is price held or raised while volume and share also hold or grow? Exclude. That is brand strength working as designed.
Past the gates, twenty points across two modes. Mode A — the Stress Crop reads the drawdown in the P&L: price–volume divergence with margin holding, share loss under rising price, brand-investment starvation, quality withdrawals. Mode B — the Held Valuation reads the concealment on the balance sheet: thin disclosed headroom, impairment deferred or attributed away, cost programs carrying the margin, discretion incentives.
| Score | Verdict | What it means |
|---|---|---|
| 0–6 | Reservoir Intact | Pattern not present. Log the confirmed negative. |
| 7–10 | Drawing Down | Partial. Re-check at the next annual report. |
| 11–15 | Reservoir Thin | Open a full deep dive. |
| 16–20 | Refill Overdue | Map the headroom disclosure and the likely write-down window. |
The instrument does not score brand health. It scores concealment — the gap between what the reported numbers say and what the brand underneath can support. A company can have a badly damaged brand and score near zero, because nothing is being hidden. Two of the ten cases prove exactly that.
Every case below had its expected result written down before the evidence was examined. That is the only way a scorecard means anything; otherwise you find what you set out to find.
| Company | Result | Score |
|---|---|---|
| Kraft Heinz 2015–19 | Refill Overdue | 18 / 20 |
| VF Corporation FY23–25 | Reservoir Thin — pure Mode B | 12 / 20 |
| AB InBev 2023 | Reservoir Thin | 11 / 20 |
| Peloton 2021–22 | Reservoir Intact | 4 / 20 |
| Altria | Screened out — dying category | — |
| Mondelez | Screened out — cost pass-through | — |
| LVMH (2023) | Excluded — healthy brand | — |
| Hermès (2025) | Excluded — healthy brand | — |
| Unilever (2025) | Cleared — all three instruments | — |
Peloton. I predicted a flag. It scored 4/20. Revenue fell 28% year-on-year in Q4 FY22 and churn nearly doubled to 1.41%. The brand was genuinely damaged — and nothing was concealed. The P&L reported the collapse in real time.
AB InBev. I predicted a clear flag. It scored 11/20. Group revenue rose 7.8% in 2023 on price while volume fell, and Bud Light lost the number-one US position it had held since 2001. But the brand-starvation signal scores zero, correctly — the company reinvested behind its brands and disclosed the damage in detail every quarter.
Brand damage is not a brand mask. An instrument that cannot tell a company that got hurt in public from a company hiding the damage is measuring the wrong thing. Where AB InBev does score is on the balance sheet: goodwill of $110.5bn and indefinite-lived intangibles of $36.9bn at end-2024, with no impairment warranted — brands carried at indefinite life on the basis of a 600-year corporate history, through a collapse its own annual report acknowledges.
Unilever should have been obvious. Roughly €22bn of goodwill and €18bn of indefinite-life brand intangibles built over a century of acquisitions. A productivity program affecting around 7,500 roles. A new CEO from March 2025. A major demerger. Slowing markets. From the outside, it has the shape of a company managing decline with financial engineering.
Underlying operating profit of €10.1bn against net finance costs running at roughly 3% of about €26bn of net debt — interest cover in the low teens, against a threshold of 2×. Free cash flow €5.9bn at 100% cash conversion. Net debt/EBITDA 2.1×. ROIC 19.0%. Goodwill is large; it is also irrelevant when the core is this strong.
Unilever generates cash, converts all of it, pays a rising dividend and completed a €1.5bn buyback. No burn, no dependence on outside capital.
| Continuing basis | 2023 | 2024 | 2025 |
|---|---|---|---|
| Underlying sales growth | 7.7% | 4.3% | 3.5% |
| Volume | +1.1% | +3.1% | +1.5% |
| Price | +6.5% | +1.2% | +2.0% |
| Underlying operating margin | 17.6% | 19.4% | 20.0% |
| Gross margin (2025) | — | — | 46.9%, +20bps |
Volume never went negative — not even in 2023, when price ran at +6.5% through the inflation spike. Twelve consecutive quarters of underlying volume growth, all four business groups positive in 2025, and Power Brands at 78% of turnover growing 4.3% with volume up 2.2%. There is no divergence to explain. Gate 2 excludes on the top row of the table: price up, volume up, margin up.
Overheads improved 50 basis points from the productivity program — and the savings went into brands. That is the difference between a cost program that funds a refill and one that funds a harvest. Same activity. The tell is where the money goes.
On 6 December 2025 Unilever demerged its Ice Cream business as The Magnum Ice Cream Company, retaining 19.85%. A stress crop leaves the tree depleted. Here both parts are alive: since the demerger Unilever is up 11.6% and TMICC up 1.3%, together adding over €16bn in shareholder value. That is a healthy shed — a limb removed by a growing organism.
Unilever's own segment data demonstrates why the third number matters. Foods in 2023: price +10.1%, volume −2.2% — the exact Brand Mask silhouette. But its underlying operating margin that year was 18.6%, a three-year low. Margin compressed while price rose: input-cost pass-through, the same pattern as Mondelez.
By 2025, Foods volume had recovered to +0.8% and margin to 22.6%. Home Care ran the same shape in 2023 and recovered too. Two divisions briefly wore the mask, the third number correctly said they were not wearing it, and then they recovered — which is what a bad season does, and what a spent reservoir does not.
Every figure above is on a continuing operations basis, excluding the demerged Ice Cream business, with 2024 and 2023 comparatives re-presented to match.
The obvious challenge: if Ice Cream was the weak limb, this clearance is flattered by the shed. A company can always look healthier by removing its worst division from the picture.
Ice Cream generated €7.7bn of turnover on €0.7bn of operating profit — a margin of roughly 9.1%, against the retained group's 17.9%. So the level of Unilever's margin is genuinely improved by the separation. That should be said plainly rather than buried.
But the clearance does not rest on the level. It rests on three things the re-presentation cannot manufacture:
A skeptic should still want the pre-demerger series, and should ask for it. On the evidence available, the shed changed the level and not the verdict.
The method is not proprietary and the data is public. Run it on an employer, a supplier, a holding — anything you care about.
Most consumer companies split sales growth into price and volume. Find gross margin for the same years. Three years minimum.
Price up with volume up is health. Price up with volume down and margin up is the pattern worth investigating. Price up with volume down and margin down is usually a cost problem, not a brand problem.
Advertising and marketing as a percentage of revenue, over five years. Rising, flat, or falling? This is the clearest single signal of whether the reservoir is being refilled — and the first thing cut, because it is the easiest cut to make and the slowest to hurt.
Search the annual report for "impairment" and read the note on goodwill and intangible assets. Under US GAAP, look for brands in thin headroom bands. Under IFRS, look for the sensitivity disclosure — how far revenue or the discount rate would have to move before headroom hits zero.
If a brand is written down and the explanation is interest rates, ask what volume and share have been doing for three years.
A clearance carries the same duty of care as a flag, so here is the honest ledger.
Unilever's own IAS 36 sensitivity disclosure was not read. I confirmed the mechanism and its location — Note 9 of the Form 20-F — but not the company's specific headroom figures. The balance-sheet half of the analysis is therefore incomplete. Every visible signal points low: brand carrying value has been leaving Unilever's balance sheet by disposal rather than write-down, with no pattern of triggers without impairment. But "points low" is not "verified low."
Also open: interest cover is derived rather than read from a single line, though the exclusion holds at double the finance cost. Third-party market share data was not obtained. Ben & Jerry's was not examined — it almost certainly moved with the demerger, but that was inferred rather than confirmed. Closed 7 August 2026: The Magnum Ice Cream Company's own filings list Ben & Jerry's among its brands. The inference was correct. And the instrument's deliberate-scarcity exclusion has never actually fired across ten cases; it was written on reasoning, not evidence.
This is research, not investment advice, and not a recommendation to buy, sell or hold anything. The Brand Mask produces 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: Unilever cleared every instrument applied to it. That is a finding about a method, not an endorsement of a security.
Figures come from named filings — Unilever's Form 20-F 2025 and results announcements, and the SEC filings of Kraft Heinz, VF, Mondelez, Peloton, Altria and AB InBev. Where a figure comes from trade press or an interested party, the full report says so and reports disagreements as ranges rather than picking the convenient number.
The instrument is version 0.2, calibrated across ten companies in three rounds, with two documented prediction failures. It will change. 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.