I had a discussion recently with a regulator from a major jurisdiction: a person who has had reason to use Tax Information Exchange Agreements and who knows their pitfalls from the inside.

I asked him a simple question:

Do secrecy jurisdictions really know who are the beneficial owners of the structures that exist in their domains?

He laughed in response.

Look, I deal with the places that are cooperative — who have signed these deals, and where we think we might make some progress. And candidly they just don’t know. They ask the administrator who is meant to have the data and all they know is that the beneficial owner is a structure in another secrecy jurisdiction.

And in a sense it’s not that they’re lying: it may be that discretionary trust has been layered on discretionary trust, upon company and foundation until no one knows.

Was this chance, I asked?

No of course not.

But it does frustrate our purpose.

He said. And does it prove that these places really don’t know who owns the structures that operate within them, I asked?

He just laughed. And then added:

And remember I’m talking about enquiries made of the more willingly cooperative jurisdictions and we get nowhere.

I can’t name names. I’d like to but can’t. But let’s be clear we’re talking best case scenarios here. And candidly Tax Information Exchange Agreements aren’t working. And nor are beneficial ownership rules.

We can tweak them forever, but that’s the wrong direction of travel. Only mandatory identification of the warm bodies behind structures and the automatic exchange of information in accordance with that data can resolve this issue. Whilst the assumption of a right to privacy remains then we’ll get nowhere.