Schema: placement of the 'beneficialOwnershipOrControl' flag at Interest level

[kd-ods]

What's the rationale for having the 'beneficialOwnershipOrControl' flag/boolean at the Interest level as opposed to the Ownership-or-control Statement level?

(From a publisher's point of view, a BO declaration may be made on the basis that a person has a collection of interests in a company which cumulatively afford them status as a beneficial owner. To put it another way: none of a person's multiple interests in a company might in itself meet the BO threshold, yet taken together they meet a threshold. In this case a flag at the statement level would make more sense.)

[ScatteredInk]

I think there are two different use cases here:

1. Have a flag at statement level that allows publishers to definitively state if a natural person is a BO or not. We could call this declaresInterestedPartyAsBeneficalOwner or similar.
2. Have a flag at interest level that allows publishers to distinguish between legal title and beneficial ownership (and unknown ownership type) of interests held by natural persons. This would be useful in the case of integrating shareholder registers, where the flag would be false (or if just nominees had to declare their shareholdings).

2 is needed for BODS to work. 1 sounds useful and would be particularly interesting for cross-jurisdictional data use - but would we have much data where it was set to false?

[kd-ods]

1 sounds useful and would be particularly interesting for cross-jurisdictional data use - but would we have much data where it was set to false?

Well, in the case of integrating shareholder registers and the like it wouldn't necessarily be set to true.

Can you say a bit more why '2 is needed for BODS to work'? (2 is effectively what we have now.)

[ScatteredInk]

Can you say a bit more why '2 is needed for BODS to work'? (2 is effectively what we have now.)

BODS is about beneficial ownership but inevitably represents information about legal ownership to get to beneficial ownership. Without that flag at interest level, we can't distinguish (in a single statement) between a natural person who just holds legal title and a natural person who has a beneficial interest through legal title. Both statements could be either legal or beneficial ownership of shares. So the flag at interest level says "this interest is definitely about beneficial ownership".

But that makes me agree with you that a flag at statement level would be useful because, without it, you couldn't distinguish between a person who is a beneficial owner, according to some disclosure regime, and a person who has a beneficial ownership interest. Given the difficulties of aggregating beneficial ownership data, this seems useful to push onto publishers rather than users.

The only difficulty I see for implementers is when, e.g., thresholds or other regulations change over time. We would therefore need to say that systems maintained the ability to output BODS data in accordance with all current and past disclosure requirements?

[kd-ods]

I'm not sure about this:

Without that flag at interest level, we can't distinguish (in a single statement) between a natural person who just holds legal title and a natural person who has a beneficial interest through legal title.

If the 'beneficialOwnershipOrControl' flag were at the OwnershipOrControl statement level and its scope took in all the interests declared in that OwnershipOrControl statement, then if it were set to false then you would know that all share ownership interests were not BO interests.

Also:

a flag at statement level would be useful because, without it, you couldn't distinguish between a person who is a beneficial owner, according to some disclosure regime, and a person who has a beneficial ownership interest.

Surely 'a person who has a beneficial ownership interest' only has that interest according to some disclosure regime. So I'm not sure what we're distinguishing between.

Probably just my confusion!

[timgdavies]

Going back to the definition of the beneficialOwnershipOrControl flag, this is:

Does this statement assert this as a beneficial ownership or control interest? A beneficial ownership or control interest is always between a natural person and some entity, and exists where the person ultimately benefits from, or has a degree of control over, the entity. There may be cases where a person has an interest in an entity, but where there are arrangements or other conditions that mean this interest does not constitute beneficial ownership or control.

Alas it looks like I didn't reference an issue when creating this field (note to self: always commit against an issue!), but as I read it, the operative part is does this statement assert this as a beneficial ownership or control interest?.

I.e. Is there an explicit assertion that the interest disclosed qualifies, according to the source and the publisher, as a direct beneficial ownership relationship between interestedParty and subject.

IIRC, the concern that led this to be introduced was more about publication of intermediary parts of an ownership chain.

Using one of Kadie's great diagrams, we can see the issue:

The BO has 10% shares in Company 4. If disclosing this as part of full chain disclosure of ownership in the Primary Declaring Company, then, under a 25% disclosure regime, this would not qualify as an assertion of beneficial ownership.

However, if (and I don't think in this diagram they actually do) the indirect ownership between 'Beneficial Owner' to 'Primary Declaring Company' added up to 25% ownership, that would be disclosed as an assertion of beneficial ownership and control.

I believe this explains also why this exists at the interest level, as it may be possible (particularly merging data from multiple sources) to have some interests that qualify as ownership and control, and some that do not.

On the question of whether it's possible to have N interests, where no one interest qualifies as beneficial ownership or control, but cumulatively they do, I think:

(a) In part, the focus of the beneficialOwnershipOrControl on assertions of ownership or control mitigates against this. If there is OoC, then this should be asserted in the disclosure.

(b) If there were still an issue, I would think this is best handled by adding some sort of explicit interest alongside the other interests, which is then flagged as beneficialOwnershipOrControl and which captures this concept of a cumulative interest.

Before worrying about that in any depth, it would be useful to check if there are such examples in law where N non BO interests = 1 cumulative BO interest at the level of a single point in the ownership chain.

[kd-ods]

I.e. Is there an explicit assertion that the interest disclosed qualifies, according to the source and the publisher, as a direct beneficial ownership relationship between interestedParty and subject.

I think this should be "direct or indirect".

IIRC, the concern that led this to be introduced was more about publication of intermediary parts of an ownership chain.

In that case, we need to review this beneficialOwnershipOrControl property alongside the new isComponent and componentStatementIDs properties. (Since the latter exist at the statement level, their use to describe the detail of an indirect beneficialOwnershipOrControl interest will often require breaking up collections of interests into individual o-o-c statement packages.) So... yes...

Before worrying about that in any depth, it would be useful to check if there are such examples in law where N non BO interests = 1 cumulative BO interest at the level of a single point in the ownership chain

... I agree we need to find some examples and look at modelling them.

[kd-ods]

In light of the suggestion in #446 and the new documentation linked to from it, I'm happy for this issue to be closed.

[kd-ods]

I think we need to revisit this issue of a 'beneficialOwnershipOrControl' flag being at Interest level. We will come to review of interest modelling in due course. But I think this issue can can dealt with independently of that.

Before worrying about that in any depth, it would be useful to check if there are such examples in law where N non BO interests = 1 cumulative BO interest

Not only are there examples in law of this, but Open Ownership now recommend as best practice that:

5. Legislators should aim to create a broad catch-all definition of what constitutes BO, and couple this with a non-exhaustive list of example ways in which BO can be held. This is because those seeking to use legal entities for illicit gain are constantly devising new ways to derive economic benefits from and exercise control over companies, meaning that a definition based solely on a purportedly comprehensive list of typologies would need constant revision.

(From OO's Definitions Briefling )

I think this leads us to a point where the standard needs to represent:

1) That a person is asserted to be a beneficial owner of an entity, and 2) These are the bundle of interests (direct and indirect) which mean that they meet that definition.

Whereas, at the moment, the current use of the flag represents that:

1) This individual interest (which is between a person and an entity, directly or directly) is part of a statement which asserts that this interest is one of beneficial ownership or control.

The clunkiness of that wording is due to the definition of the beneficialOwnershipOrControl property.

I'd be in favour of simplifying our modelling by:

• removing the beneficialOwnershipOrControl property from interests
• adding an isBeneficialOwner boolean property to the InterestedParty object in relationship records.

I suspect this is going to help as we grapple with how we model interest types and how beneficial ownership information overlaps with and intersects with legal ownership information.

@tymonk @oalannao - when you have some mulling time, I'd love your input on this.

[tymonk]

Do we need BODS to say they are a BO? The conditions for this will differ per country (e.g. lower thresholds for PEPs, or majority voting rights carrying through to the lowest entity, even if the threshold then goes below 25%) and when we are joining datasets from multiple countries, this may also get confusing. Is the main point not to show what type of ownership/control they have over a corporate vehicle, rather than whether that meets a specific definition or not? Apologies if I have misunderstood.

[kd-ods]

Do we need BODS to say they are a BO?

This boolean flag indicates that they have been disclosed as a BO to the publishing registry. Without a flag like this, it wouldn't be possible to distinguish in the data between - eg - managing officials who are being disclosed in lieu of BOs, and BOs themselves.