Rules
Part of Building a portfolio case study that follows the entity, not the person
Portfolio case studies valuation: why no figure is stated
Portfolio case studies valuation: what a published figure is made of, why the arithmetic cannot reach a value, and what a study should say about worth instead.
A case study is not the place to put a number on anything. Not because numbers are unimportant, but because the inputs sit in private accounts and private agreements, and everything a researcher can reach is either an estimate somebody else made or an arithmetic result derived from terms nobody has read.
Portfolio case studies valuation comes down to a writing decision as much as a research one: what a published valuation is actually made of, how to attribute one, and what a study can say instead of a figure. The method applies to any venture whose numbers sit in private accounts, and it matters most for celebrity-fronted brands, where the name is often most of what is claimed.
What to take away
- Every valuation you encounter in coverage was produced by somebody, on assumptions, for a purpose. Name all three or leave it out.
- A study can describe how a number would be built without ever supplying one, and that is usually the more useful page.
- A figure repeated across outlets is one figure with several bylines, not corroboration.
What a published valuation is made of
| Type of figure | Who produced it | What it rests on | How to treat it |
|---|---|---|---|
| A price paid | The parties | A signed agreement | A fact about a transaction, if disclosed |
| An implied valuation from a round | Arithmetic | The amount raised and the share issued | Attribute, and say implied |
| A third-party estimate | An analyst or publication | Assumptions, usually undisclosed | Attribute to the estimator |
| A figure supplied by a participant | The subject or an adviser | An interest in the number | Attribute, and note the interest |
| A figure with no attribution at all | Unknown | Unknown | Do not use it |
The bottom row is where most numbers in this category actually belong, and it is the row publications least often use.
What a Published Valuation Rests On
Figure type
- Price paid
- Parties
- Implied from round
- Arithmetic
- Third-party estimate
- Analyst or publication
- Supplied by participant
- Subject or adviser
- No attribution
- Unknown
Producer
- Price paid
- Signed agreement
- Implied from round
- Amount and shares
- Third-party estimate
- Undisclosed assumptions
- Supplied by participant
- Interest in the number
- No attribution
- Unknown
Rests on
- Price paid
- Fact if disclosed
- Implied from round
- Attribute, say implied
- Third-party estimate
- Attribute to estimator
- Supplied by participant
- Note the interest
- No attribution
- Do not use it
How to treat
- Price paid
- Implied from round
- Third-party estimate
- Supplied by participant
- No attribution
Three ventures, and the record each leaves
The rule is easier to test against names readers know.
The Honest Company. Jessica Alba's consumer brand is listed in the United States, so its registration statement and later annual reports sit on the SEC's EDGAR system. It filed for its initial public offering in April 2021 and began trading on the NASDAQ under the ticker HNST on May 5, 2021. The figure most often repeated about it is older: the valuation reported for its 2014 venture financing, which belongs in the table's second row. It is an implied number from one private round, produced by arithmetic on the amount raised, not a price anyone paid for the company. A study can tie anything it says about size to a dated filing, and it still has no figure for the brand itself.
Mint Mobile. Ryan Reynolds' wireless venture is now a mobile virtual network operator owned by T-Mobile US, which acquired it in May 2024 for cash and stock. The transaction is described in the buyer's own filings, from the buyer's side. The headline price is a price for the whole company, paid partly in shares, so it is not a figure for Reynolds' stake, and it says nothing about what his name was worth apart from the business.
Fenty Beauty. Rihanna's cosmetics line was launched on September 8, 2017, after she signed a deal with LVMH in 2016 to produce it through LVMH's Kendo division. LVMH's reporting runs at division level rather than by brand. The numbers that circulate are the third row of the table: a trade-press report of what LVMH may have paid for the agreement, and sales totals attributed to an LVMH executive. Each is attributed and none is a valuation. A study can name the owner and those statements, and stop there; it cannot convert them into a value for the line.
Three ventures, three different records, and not one of them hands a case study a number it may state as a value. Nearly every misreported brand figure starts the same way: a number attached to a company, a division or a whole transaction, repeated as if it were a price for the name alone.
Why the arithmetic does not reach a value
Three things break it, and they break it in every sector.
Three Things That Break the Arithmetic
- Classes rank differentlypreferred stock ahead of ordinary
- Residual unmeasurablegoodwill cannot be counted outside
- Right in play may be narrowmark owned or licensed
- A price for one class cannot multiply across all shares
- A study must say which right is held
Classes rank differently. Investors typically hold preferred stock, which stands ahead of ordinary shares when a company is sold or wound up. Multiply that price across every share and you have valued a senior claim as though it were the whole company; the technology investments valuation page works through what that does to a headline number.
The residual is unmeasurable. What remains after the identifiable assets are counted is goodwill, and by construction nobody outside the transaction can measure it independently. In a name-attached venture it is often most of the claimed value.
The right in play may be narrow. A venture may own a mark, or only hold a license to use one for a term. The categories of protection differ and so do their durations, as the U.S. Patent and Trademark Office sets out in Basic Facts About Trademarks. A study has to say what the record shows: whether the application was filed through the agency's TEAS system, whether the mark reached the principal register, which classes of goods and services it covers, and whether the maintenance filings are current. Saying a brand is worth something without saying which right is held has not described the asset.
What a study can say instead
- What the venture would have to earn for a given kind of buyer to be interested, described as a mechanism rather than a figure.
- Which asset a buyer in that sector would actually be acquiring.
- Which method was used on any figure that does appeara discounted cash flow, a comparable-company multiple, a comparable-transaction multiple, or a Section 409A valuation of private stock — since each answers a different question.
- What is publicly known about scaleoutlets, categories registered, territories, years trading.
- Which inputs are missing, and who holds them.
- What has been claimed publicly, by whom, and on what stated basis.
That list is not a consolation prize. A page that explains why nobody outside can price a venture, and shows what would have to be known, beats one that guesses or repeats an estimate.
A real buyer adjusts the accounts before pricing anything: restating earnings, testing what the assets would fetch on their own, then setting debt and other obligations against them. The acquisitions and exits valuation page sets out that process in order. The distance between it and a published figure is the whole problem.
Attribution, done properly
Where a figure has to appear, write it with its origin attached in the same sentence. Say who produced it, on what date, and the period the figure covers. Then say what it does not include: debt, contingent payments, other classes of share, and the rest of the group.
An estimate handled that way informs a reader. The same estimate written flat becomes a fact the piece cannot support, and the portfolio case studies failures page treats that transformation as one of the standard ways a study goes wrong.
The comparison trap
The strongest pull in this category is toward ranking: putting two ventures side by side with a number under each. It fails for a reason that has nothing to do with research effort. The two figures were produced by different methods, on different bases, at different dates, by people with different interests.
A real comparison would need the same basis, the same date and the same method on both sides, which is what coverage almost never supplies. The best celebrity companies 2027 page sets out those requirements; the celebrity brands valuation page explains why three parties can hold three defensible values for the same name on the same day.
The portfolio case studies page places all of this on the evidence ladder, where valuations sit lower than almost anything else a study handles.
Common questions
Can I ever state a price as a fact?
Where a party has disclosed it in a filing — a Form D notice, a registration statement or an annual report on the SEC's EDGAR system — or in an official statement, yes, cited to that document and described as what it covers. Otherwise, no.
What about widely reported figures?
Wide reporting is not verification. Trace the figure to its earliest appearance and you will usually find one source, often a participant.
Is refusing to give a number unhelpful to readers?
Not if you replace it with the mechanism. Readers who understand how a price would be built can assess the figures they meet elsewhere, which is a better outcome than being handed one more.
How should I handle a subject who offers me their own valuation?
Take it, attribute it to them, and say it is theirs. It is real information about what they claim, and it is not a measurement.
