FIBO — Legal Entities, Ownership & Control

Model the BE domain: legal persons, control relations, and reasoner-inferred ownership chains.

0/4 done

Theory

Drilling into FIBO Business Entities (BE)

Lesson 1 framed FIBO as finance's standard library. Now we go deep on the module a data professional touches most for KYC, risk and regulatory reporting: BE — Business Entities, and specifically ownership and control.

The class backbone

FIBO distinguishes ideas a home-grown model usually conflates:

  • LegalPerson — an entity the law treats as a person (a company, a fund). Distinct from a LegalEntity, which is the broader notion of something with legal standing.
  • AutonomousAgent / Party — FND-level abstractions that let a person and an organisation both play a role in a contract without pretending they're the same kind of thing.
  • Roles vs players — FIBO models a party in role (e.g. the borrower role an organisation plays in a loan) separately from the organisation itself. The same legal person can be a borrower in one contract and a guarantor in another.

Ownership vs control — not the same relation

Regulators care about both, and they differ:

  • Ownership is about equity — who holds shares, and how much.
  • Control is about power — who can direct the entity, which can exist with little or no equity (via voting agreements, board seats, golden shares).

FIBO gives you separate properties for each (illustratively hasOwnership / isControlledBy) so 'A owns 30% of B' and 'A controls B' are independent, queryable facts.

Why this is a reasoning problem

Control is often transitive: if A controls B and B controls C, A indirectly controls C. Declare the control property transitive in OWL and a reasoner walks the whole chain for you — no recursive SQL, no hand-maintained closure table. That inferred chain is exactly what beneficial-ownership and concentration-risk reports need.

Use Case Example: A bank must find the ultimate beneficial owner of a corporate client buried under three holding companies. Modelling each tier with FIBO control relations and marking the control property transitive lets a reasoner surface the top of the chain automatically — the same dataset then answers 'how much total exposure do we have to anything that entity controls?'

Analogy

Ownership vs control is the shareholder vs the steering wheel. You can own a sliver of a company's shares yet sit in the driver's seat through a voting agreement and three board seats; you can own a huge passive stake and steer nothing. FIBO refuses to glue the seat to the shares — it keeps 'who holds equity' and 'who can turn the wheel' as separate dials, because regulators audit each one on its own.

Transitive control closure

Click a node to focus its neighbourhood · drag to pan · scroll to zoom

An indirect control chain a reasoner can close

Mark the control relation transitive and the engine infers the dashed top-to-bottom link no one stated explicitly.

Reflect

Beneficial ownership looks like a data-entry problem and is really a reasoning problem. FIBO's separation of ownership from control, plus one transitive property, turns a brittle pile of join queries into a fact the engine derives.

  • Where does your current data conflate 'owns' with 'controls', and what risk does that hide?
  • Could a transitive property replace a hand-maintained closure table in your stack?

Reading in progress · 0 of 4 activities done