Costs
Part of Reading richest rankings with a sceptical eye
5 things worth knowing about richest rankings estimate
Richest rankings estimate inputs audited: what a list compiler actually holds in the file for each entry, and how thin that folder turns out to be.
Imagine the folder a compiler keeps for one entry on a rich list. Open it and see what is inside. That exercise is more informative than any argument about accuracy, because the folder is usually thinner than the confident number on the page suggests, and its contents are almost never what a reader assumes.
What to take away
- Most entries rest on prior articles rather than on any document.
- The documents that do exist describe companies, not people.
- Nothing in any folder addresses debt, which is half the definition.
What is in the folder
| Item | How often it is present | What it can support |
|---|---|---|
| A figure from a previous list | Nearly always | Only that somebody published that number before |
| News coverage citing that figure | Nearly always | The same, repeated |
| A company filing | Sometimes, for company officers | Disclosed pay and holdings at one company, at a date |
| A property record | Sometimes | A price on a date, with no view of the borrowing behind it |
| A court filing | Rarely | Whatever that specific dispute established |
| A registration record | Rarely | A claim in a mark or a work, and recorded transfers |
| A statement from the subject or their representatives | Occasionally | An interested party's assertion |
| Anything about liabilities | Almost never | Nothing, because nothing is there |
The first two rows are present for essentially every entry, and neither is evidence. Rows three to six are real documents and they appear for a minority of entries. The last row is the one that matters most and it is empty for everyone.
What the folder holds
- Previous list figurenearly always
- News coveragenearly always
- Company filingsometimes
- Property recordsometimes
- Court filingrarely
- Registration recordrarely
- Subject statementoccasionally
- Liabilitiesalmost never
Why the strong documents do not close the question
The documents that do exist are strong within their scope and their scope is narrow.
What each document can prove
Company filing
- Shows
- Pay at one firm
- Assets elsewhere
- Nothing
- Debt
- Nothing
- Coverage
- Minority of entries
Property record
- Shows
- A price on a date
- Assets elsewhere
- Nothing
- Debt
- Nothing
- Coverage
- Minority of entries
Court filing
- Shows
- One dispute's findings
- Assets elsewhere
- Nothing
- Debt
- Nothing
- Coverage
- Rare
A company filing describes a relationship between a person and one company, produced under a disclosure obligation, retrievable through the SEC's EDGAR full text search. It says nothing about assets held elsewhere and nothing at all about debt.
A property record shows a transaction. It does not show what was borrowed to fund it, whether the buyer was a company, or what the property is worth now.
A court filing is the most reliable document in the whole set, because it survived an adversarial process, and the judiciary describes what is available in its guide to court records. It also exists only where something went wrong, so it is a selected sample.
None of them is a balance sheet, and no combination of them becomes one.
What is never in the folder
Personal debt of any kind. Mortgages, margin borrowing, pledged shares, private loans, guarantees given to other people's businesses. No disclosure regime anywhere requires an individual to publish what they owe.
That single absence is decisive. Net worth is defined as assets minus liabilities, so a folder with no liability information cannot produce a net worth at any level of care. It can produce an estimate of some assets, which is a different quantity that should carry a different name.
The aggregate alternative
There is one place where household wealth is measured properly, and it is worth knowing about because it shows what the real thing looks like. National statistical work surveys households, publishes its sampling design, and states its own limitations. The Federal Reserve's Survey of Consumer Finances is the standard example for the United States.
Compare that with a rich list. One has a documented method, a stated population, published uncertainty, and no claims about named individuals.
The other gives a confident number for each named person, with no method anyone can re-run. The difference is not effort. It is whether the object being measured is accessible at all, as argued on estimate methodology.
How to audit a folder you cannot see
You can infer its contents from the article. Count the documents it names, with issuers and dates. Most articles will yield zero. Then look at whether any liability is mentioned anywhere. It will not be. Then check whether a valuation date is given for anything priced in a market.
Three checks, and they tell you roughly what is in the folder. The way those thin folders are then ordered into a table is on richest rankings.
How that same problem plays out for company officers is on executive net worth, and how the lifetime version fails is on career earnings.
Common questions
Do compilers ever contact the people they list?
Some do, and a response is a data point from an interested party rather than a document. It also introduces a selection effect, because people who engage can shape their entry and people who ignore the request cannot.
Is a property record a solid asset input?
It is solid about a transaction and silent about the financing. A recorded purchase adds an asset to the estimate while leaving the matching liability invisible, which biases every folder in the same direction.
Why not simply publish the folder?
Because published folders would show readers how little is in them, and the format depends on the number looking derived. That is a commercial observation rather than an accusation of dishonesty.
