# The name on the certificate is not the owner.

A company name is where the record starts. It is not where accountability ends. The Beneficial Ownership Data Standard (BODS) does not reveal the truth about who owns a company; it makes the ownership claim inspectable: linkable, dated, and testable.

By Michael Cengkuru · June 14, 2026 · Interactive version: https://cengkuru.com/essays/the-name-on-the-certificate/

## The story

A public registry gives you a clean record with a company's name on it. It does not tell you who is behind it. A company name is not an owner. It is a door. Beneficial ownership transparency begins when the company name stops being treated as the answer.

One disclosure before the story starts: the company network in this essay is fictional; the data model and the tests are real. The companies and the person in the chain that follows, Mirembe Infrastructure Ltd, Lakefront Holdings Ltd, Blue Gate Nominees Ltd, Riverbank Works Ltd, and Amina K., were invented to teach the Beneficial Ownership Data Standard (BODS v0.4), published by Open Ownership. The standard facts come from the published BODS documentation.

Mirembe Infrastructure Ltd has won a public contract. Public money: a road, a clinic, a classroom block. Here is why the name on the certificate is not enough. One member of the committee that awarded the contract is named Amina K. If the person who benefits from Mirembe is also Amina K., that is a conflict of interest worth knowing. Maybe it is a coincidence of names. You cannot tell from a certificate. You have to open the door.

Pull the line and it leaves the certificate. The first answer is another company: Lakefront Holdings Ltd, a 60 percent shareholder. Useful, but not enough, because a company cannot ultimately enjoy private benefit or hold a conflict of interest. Ownership is not a list. It is a route. Lakefront leads to Blue Gate Nominees, a nominee arrangement, which leads toward a person. Direct ownership is a straight line; indirect ownership is a route that can pass through companies, nominees, and legal arrangements before it reaches whoever benefits or controls.

The line reaches a human being, and it is the same name. The person who benefits from Mirembe is Amina K., the same Amina K. who sat on the committee that awarded its contract. This does not mean guilt. It means the question can finally be asked properly: of a person, not a name on a door. Naming the conflict is not the same as proving it.

BODS turns that chase into three kinds of statement. An entity statement records the legal thing: a company, trust, or other arrangement. A person statement records the natural human being who may own, control, or benefit. A relationship statement is the hinge: it names a subject and an interested party, then lists the interests between them, each with its type, its share, and whether it is direct or indirect. When ownership is indirect, the primary relationship still names the beneficial owner, and the intermediary hops are kept as component records. Every statement carries a source and a date. A claim is not a fact; a dated, sourced claim is something you can test. That is the essay's refrain: Name it. Link it. Date it. Test it.

Statements are never edited in place. A correction is a new statement that amends or confirms the old one. Two identifiers do different jobs: a statement identifier names this claim, and a record identifier names the thing the claim is about. That persistence matters when a flagged company dissolves and the same person appears behind a new one, Riverbank Works Ltd in the essay's telling: the company can be replaced overnight, but the person record cannot. A regulator can see the pattern and ask the next question. A pattern worth checking is not a verdict on the person.

The essay closes with a query console that runs four questions against the fictional record. Who is the beneficial owner? The line runs Mirembe to Lakefront to Blue Gate to Amina K.: a person reached through relationship statements, a sourced and dated claim, not proof. Is the interest direct or indirect? Indirect, with two intermediary hops held in component records. What changed over time? One record, two statements: a 60 percent share in 2023 became 20 percent in 2026, and nothing was overwritten. Known, unknown, exempt, or not disclosed? Blank is not the same as none: where no owner is disclosed, the relationship records a reason instead of a person, and the line is drawn broken, not hidden.

The limits, stated plainly: BODS can show who was declared as owner or controller, by which source, on which date, through which interest. It cannot prove by itself that the declaration is true, that the person is guilty of anything, that the structure shown is complete, or that the register is current. The certificate did not lie. It was incomplete. A company name is where the record starts. The public-interest question is who the line reaches.

The essay's own editorial note, repeated here in full: "The company network is entirely fictional: Mirembe Infrastructure Ltd, Lakefront Holdings Ltd, Blue Gate Nominees Ltd, Riverbank Works Ltd and Amina K. Any resemblance to real companies or people is coincidental. The award-committee conflict and the dissolve-and-re-form (phoenix) sequence are illustrative stakes to show why following the line matters; they are not a real case, and a repeated pattern behind one person is a flag to investigate, never a finding of wrongdoing."

## The data

What BODS statements record, as the essay presents them:

| Statement or field | What it records | In this essay |
| --- | --- | --- |
| recordType: entity | The legal thing: a company, trust, or other arrangement | Mirembe Infrastructure Ltd, Lakefront Holdings Ltd, Blue Gate Nominees Ltd, Riverbank Works Ltd (fictional worked example) |
| recordType: person | The natural human being who may own, control, or benefit | Amina K. (fictional worked example) |
| recordType: relationship | The link: a subject, an interestedParty, and a list of interests | Mirembe Infrastructure Ltd (subject) and Amina K. (interestedParty) (fictional worked example) |
| interests: type, directOrIndirect, beneficialOwnershipOrControl | Each interest carries its own type, direct-or-indirect setting, and ownership-or-control flag | shareholding, indirect, true (fictional worked example) |
| share (exact, minimum, maximum) | The size of the interest, modelled as an object | exact: 60, later 20 (fictional worked example) |
| statementDate and source | When the claim was made and by whom, under a disclosure rule | 2026-06-14, source.type ["selfDeclaration"] (fictional worked example) |
| statementId and recordId | statementId identifies this claim; recordId identifies the thing the claim is about | Statements s-001 (2023) and s-002 (2026) share recordId rel-118 (fictional worked example) |
| recordStatus | new, updated, or closed; statements are append-only, never edited in place | new in 2023, updated in 2026 (fictional worked example) |
| componentRecords and isComponent | The intermediary hops of an indirect route, referenced from the primary relationship | Amina K. to Blue Gate, Blue Gate to Lakefront, Lakefront to Mirembe (fictional worked example) |
| interestedParty.reason | Absence recorded rather than hidden, for example noBeneficialOwners | The broken line in the query console (fictional worked example) |

## Sources

- Beneficial Ownership Data Standard (BODS v0.4), published by Open Ownership. It is the schema the essay teaches: https://standard.openownership.org/
- Companion essay, the use case: The Black Light for Public Money, joining OC4IDS procurement data to ownership data via a shared identifier: https://cengkuru.com/essays/black-light/
- Method note: the company network is fictional; standard facts verified against the published Beneficial Ownership Data Standard (v0.4) documentation, June 2026. The data line is a drawing device, not a feature of the standard. Built as a self-contained page: no framework, no tracking, no third-party scripts.