In 2017, the Court of Justice of the European Union handed down a ruling that should have made the trade press — but didn’t. In Khorassani v Pflanz (Case C-678/15), the Court was asked whether the act of putting a retail client in touch with an asset manager and helping them sign a portfolio-management mandate is an investment service within the meaning of MiFID. If the answer were yes, every firm acting as a go-between in that flow would need a MiFID authorisation. The CJEU said no.
What the Court actually ruled
The reference came from Germany’s Federal Court of Justice (BGH). At issue was whether brokering a portfolio-management contract fell within the “reception and transmission of orders in relation to one or more financial instruments” in Annex I, Section A, point 1 of MiFID I (now MiFID II). The Court’s answer was structural: the activity of brokering is intermediation toward the conclusion of a contract, not the transmission of an order for a financial instrument. The mandate itself isn’t a financial instrument; what the manager subsequently does inside the mandate is the investment service, and that activity has its own MiFID perimeter.
Translated out of the legal grammar: the introduction is outside MiFID. The portfolio management on the other end of the introduction is inside MiFID. Two different perimeters; one commonly-confused single transaction.
Why this matters in practice
For most of the post-MiFID-II era, European retail-investment operators have defaulted to one of two postures: hold an investment-firm authorisation themselves (expensive, slow, capital-bound), or hide inside a regulated parent as a tied agent. Khorassani v Pflanzopened a third, narrow path: act exclusively as an introducer to a separately authorised manager, without ever taking custody, without ever executing, without ever exercising discretion. Done strictly, that activity sits outside MiFID’s authorisation requirement.
It is not, however, outside all regulation. In Germany the activity will typically still require a §34c GewO trade permission, and where consumer credit, advice or marketing rules apply they apply. National-level conduct rules, anti- money-laundering obligations, marketing-communication standards and data-protection law are all unaffected by Khorassani. The ruling is a MiFID-authorisation carve-out, not a regulatory bypass.
The structural consequence for copytrading
The implications are sharpest in copytrading. The dominant copy-trading platforms in retail FX/CFD — Naga, eToro, ZuluTrade — combine the platform, the broker, and the execution venue under one corporate roof. The follower copies a signal-provider; the same firm receives the order, sets the spread, and faces the client on the trade. That is a textbook conflict-of-interest structure. The licence is real, but the incentive geometry of the licence-holder is not aligned with the follower.
The mandate-brokerage model implied by Khorassani separates those functions by construction. The introducer is not a broker. The asset manager is the broker’s client and operates under its own MiFID authorisation. The follower has a contractual relationship with the asset manager, who in turn executes through unaffiliated venues. The introducer’s commercial interest is in matching the follower to a manager whose track record is real; the asset manager’s commercial interest is in compounding the mandate. The broker- as-counterparty problem doesn’t arise because there is no broker-as-counterparty.
What this is not
It is not a free pass. Operators relying on Khorassani must be introducers only. The moment they accept funds, execute, or hold themselves out as offering portfolio management, the MiFID perimeter applies in full. National supervisors — BaFin in Germany — have been clear in administrative practice that the substance of the activity, not the legal label, decides the perimeter.
Nor is it a guarantee about the asset managers on the other end. The introducer’s diligence on the manager — authorisation, track record, conflict policy, custody chain — is the structural-quality question. Khorassani permits the model; it does not validate any particular operator inside it.
Why the ruling has stayed obscure
Three reasons. First, the case was a private-law dispute about contractual nullity, not a regulatory enforcement matter — so the press release didn’t fit the supervisory news cycle. Second, it answered a question incumbents would have preferred not asked: holding a MiFID licence is expensive, and the unlicensed alternative wasn’t something the licensed community wanted to advertise. Third, the operationally-correct model the ruling implies is structurally inconvenient for the platforms that dominate the retail-investment surface today. Public commentary on Khorassani has been thinner than the ruling deserves.
Reading recommendations
The judgment itself is short and clear: Case C-678/15, Mohammad Zadeh Khorassani v Kathrin Pflanz. For the German private-law context, the BGH’s referring order (BGH, 13 April 2017, III ZR 84/16) sets out the dispute cleanly. For the contemporary supervisory practice, BaFin’s published Q&A on §34c GewO versus §32 KWG remains the cleanest summary of where the line sits in Germany.