Doing the right thing and being able to show it are two different things.

An investor says: "nobody explained the risk to me."

Three years later. The deal went badly. The firm remembers the conversation clearly; it always explains the risk. Now it must show it.

This is where firms discover something uncomfortable about how these disputes actually run: the file does more work than the memory.

• Under German law, once a disclosure duty has been breached, fault is presumed: the firm has to establish it was not at fault, not the other way round (§ 280(1) BGB).
• Germany's Federal Court of Justice went further: where a disclosure duty was breached, it is presumed the investor would have acted differently had they been told, a rebuttable presumption that reverses the burden on causation (BGH XI ZR 262/10, 2012).
• In an insurance-intermediation case, a required warning was not documented at all. The court held the intermediary had to prove it had given it (BGH III ZR 544/13, 2014).
• And where only one side knows what happened in the room, a bare denial is not enough; that side carries a secondary burden to give a concrete account (BGH VI ZR 343/13, 2015).

The offer stayed private. Now that must be proven.

An exemption is not something a company asserts. It is something the company carries the burden of proving, and the proof is a record of exactly who was approached, how many, and what they were told.

• Under the EU Prospectus Regulation, it is the offeror who must show the offer stayed within the exemption: fewer than 150 non-qualified people per member state, or qualified investors only (Reg. (EU) 2017/1129, Art. 1(4)).
• In Germany, an issuer facing a defective-prospectus claim escapes only if it proves it did not know of the defect and was not grossly negligent, and the investor's reliance is presumed in their favour (§ 12 WpPG). Commentary notes the standard is strict precisely because the issuer holds all the underlying information.
• The classic statement is American and seventy years old: a company claiming the private-offering exemption bears the burden of proving its buyers had access to the information registration would have disclosed. The purchasers "were not shown" to have had it. The company lost on a failure of proof, not on a finding of bad conduct (SEC v. Ralston Purina Co., 346 U.S. 119 (1953)).

No client harmed. No deal mis-sold. Sanctioned anyway.

• In one two-week period, a single European regulator fined two international banks a combined sum in the tens of millions, not for harming anyone, but for a decade of transaction reports that were merely inaccurate. The misconduct was the record.
• A small brokerage was fined and its directors personally fined and banned for failing to report thousands of transactions, and for not monitoring its own calls at all. This is not a big-institution problem.
• One European bank was fined twice in three years for the same failure: filing its suspicion reports late. Not wrong. Late. Timeliness is itself a compliance artefact, and a timestamp is the only evidence of it.

Anonymised deliberately. Every one is public, findable, and permanent for the people in it.

And the regulator publishes before anything is final.

A different pattern, and not a record failure: a steady cadence of orders to cease and unwind unauthorised business, naming companies and named individuals, each requiring repayment to every lender, and each published while still under appeal, months or years before any final determination. Publication is not a finding of wrongdoing. It is still the first thing a search engine returns.

A suspicion is enough to freeze the account. The explanation comes after.

In the UK alone, tens of thousands of transactions a year are placed on a statutory hold while a suspicion is worked out: 57,081 consent requests in one year, £190.3m refused, and a refusal triggers a 31-day moratorium during which the money does not move (year to March 2024).

This describes the mechanism, which is documented. It does not claim the mechanism routinely lands on legitimate businesses. That claim is often made, and it could not be evidenced here.

Everything about this gets tighter from here.

• Germany decided in 2010 that undocumented advice should cost the firm something. Before that reform, it cost nothing. That direction has never reversed.
• The EU's new anti-money-laundering authority, AMLA, begins direct supervision in 2028, of 40 selected cross-border groups. But the part that reaches everyone else is the other half of its mandate: common supervisory standards, joint supervisory teams with national authorities, and regular peer review of those authorities. The floor is being raised for everyone, not a fence built around forty firms (AMLA).
• The harmonised EU rulebook applies in every member state from 10 July 2027.
• Switzerland's regulator, which today cannot fine institutions at all, is publicly asking for more powers over individuals.

This page describes published law and enforcement practice, with sources, as at the dates given. It is general information, not legal advice, and not a description of any obligation applying to any particular reader, nor any prediction of any outcome. Cases are anonymised; nothing here is a statement about any identified company or person. Exedra Gate is not a law firm and never holds or moves investor funds.