How extra-territorial is MiFID?

How extra-territorial is MiFID? This is a question that we have been asking ourselves, and been being asked by clients, for some time now. We sometimes hear the view that MiFID is terribly extra-territorial and, for a global group, every corner of the empire will be affected. Others seem to take the view that nothing much changes in this regard between MiFID I and MiFID II, and MiFID I does not seem to be too extra-territorial at the moment.

In this article, we analyse what MiFID actually says and does, and what regulators have said about the way they intend it to apply. We are able to draw some clear conclusions for clients that we hope will help them as they enter the delivery phase of MiFID II implementation projects.

Background

MiFID II does not appear to set out to be deliberately extra-territorial. Its scope is different from, for example, the market abuse regime. MAR applies to activities undertaken anywhere in the world provided that the relevant instrument is listed or tradable in the EU. This made a good deal of sense under MAD, which largely restricted itself to anti-insider dealing and market manipulation measures. It became decidedly awkward when MAR was expanded to include conduct matters, such as disclosures in investment recommendations. But it has always been clear that MiFID takes a different path. MiFID applies to investment firms – a definition which attaches to EU entities, and activities undertaken in the EU. So, as a starting point, under MiFID it generally does not matter either where the instrument is listed/traded, nor where the client is based. If the firm is in the EU, MiFID's rules apply. And at the moment, that seems to be backed up by the reality of the situation "on the ground". Let us take best execution as an early example, and it is one to which we shall return. If a transaction is executed in (say) Hong Kong, then no matter where the client is based (EU or not), nor where the instrument is listed, nor where the trade is eventually booked, does it make any difference to the extent to which MiFID's best execution rules apply – they do not. There is some logic to this position. Clients who choose to deal with Hong Kong based brokers presumably expect to receive the protections afforded by the Hong Kong system, no matter where in the world the client may be based. If they insisted on receiving EU protections, they would presumably deal with or through an EU-based firm. So far so good.

The FCA's second Consultation Paper on MiFID II (CP16/19) sets out the UK's approach to this topic. Admittedly, there are other regulators out there, but FCA's view is likely to remain influential. FCA says that it does not want to alter the MiFID I position in relation to the application of conduct rules to non-EEA branches of UK investment firms. In brief, FCA's stance is that the rules do not apply. FCA has done this by retaining the SYSC provision at 2.15 that states "the common platform requirements, except the common platform record keeping requirements, apply to a firm in relation to activities carried on by it from an establishment in the United Kingdom". We have previously termed this the "what happens in Vegas, stays in Vegas" approach to non-EU branches. In other words, local standards of behaviour will apply. In FCA's view, unless there is a specific requirement to the contrary, MiFID II is not extra-territorial.

So, where are there specific requirements to the contrary?

Transaction reporting by branches

The clearest example of extra-territoriality in MiFID II is the obligation that branches submit transaction reports to the Home State regulator, even where the execution occurs in a non-EEA branch. Article 26 of MiFIR invites ESMA to draft regulatory technical standards on "the application of transaction reporting obligations to branches of investment firms", and this they have done in RTS 22. Therefore, transaction reporting is explicitly extra-territorial. Isn't this bad news across the board? Is it the thin end of the wedge, showing us that MiFID II generally is extra-territorial?

In our view, no – quite the opposite. Because MiFID II makes clear that business undertaken in a branch must, nonetheless, be transaction reported within the EU, we think that the alternative view is strengthened – the rest of MiFID II (or at least, any part that is not made so explicitly extra-territorial) does not apply. Otherwise, it would not have been necessary to have made this point.

Position limits

MiFID II introduces a significant new regime applying to commodity derivatives traded on EU trading venues, and economically equivalent OTC contracts. The regime permits the competent authority of the trading venue with the largest volume to set position limits to particular contracts. Article 57 of MiFID II sets out the regime, and RTS 21 provides the technical standards on the calculation of the limits. A reporting regime also applies under Article 58 of MiFID II. And Article 1(6) of MiFID II applies the position limit and reporting regime to unregulated entities. 

So it can be seen immediately that this regime is significantly different from the general MiFID II approach, which applies only to firms. First, the position limits regime has a similar trigger to MAR – it applies to those trading in commodity derivatives listed or tradable in the EU. And second, it is expressed to apply to unregulated firms. So the scope of this aspect of MiFID II appears to be global (in the same way as it could be said that MAR is global), and is therefore significantly extra-territorial. But we take similar comfort as a result that this is another isolated example, not part of a general trend.

Product governance

MiFID II introduces, on a pan-European basis, a product governance regime which owes much of its DNA to the existing UK regime applicable to the manufacturers and distributors of retail structured products. The regime is worthy of detailed treatment, but, for now, suffice to say that the regime requires manufacturers to take responsibility for product design, and requires manufacturers and distributors to co-operate to ensure that products are sold into designated target markets.

Whilst the MiFID product governance regime is not explicitly extra-territorial (the rules clearly only apply to Firms, when conducting their business in the EU) it does give rise to the question – what is the obligation on an EU manufacturer if it is using a non-EU distributor who is not caught by the MiFID II product governance rules? Or an EU distributor who wants to sell products not manufactured under the MiFID II product governance regime because the manufacturer is outside the EU? ESMA has given detailed guidance on the obligations applying to the EU firm in these circumstances, and they may well lead to a measure of extra-territorial effect, because the EU firm may need to use distribution agreements to ensure that the non-EU firm is meeting certain EU standards. It would probably be wrong to call this extra-territorial, because it is not the MiFID II rules that apply to the non-EU firm, but they may have an indirect effect of a similar nature because the contract between the EU and the non-EU firm is likely to include some MiFID obligations. But ultimately, the non-EU firm would not be at risk of enforcement action – it would be at risk of being sued by the EU firm if it did not live up to its contractual obligations. And it remains to be seen how this regime works in practice – will non-EU firms be prepared to engage on MiFID II terms? Or will the EU firm end up taking on more of the responsibilities because of the reluctance of the non-EU firm to do so?

Inducements and research

MiFID II contains rules relating to inducements (which have much in common with those contained in MiFID I) and new rules relating to the research industry (which cannot be found in MiFID I – some wags may say the rules cannot be found in MiFID II either!). One way or another, EU firms are prevented from either offering or receiving prohibited inducements, and research content is caught by this regime. This means, in practice, that an EU firm might not be able to offer, or receive, a payment which would be permitted under the rules of the non-EU jurisdiction where a third party is based. And, to take research as an example, it will be difficult for EU brokers to offer research content for free into third countries where it can be consumed for free (and the US is a good example here, because it is potentially unlawful in the US for firms to offer research to US-based managers, and charge for it, without being registered as investment advisors). So again, it would probably be wrong to classify this as extra-territoriality, because the rules do not apply to non-EU firms. However, the way that the rules apply to EU firms exports some of the EU concepts into arrangements between EU and non-EU firms.

Mandatory trading obligation

The Mandatory Trading Obligation ("MTO") is contained in MiFIR Articles 23 and 28 – 34. The obligation is further developed in RTS 1. In summary, the MTO requires EU firms (subject to some limited exceptions) to trade on venues, within the EU, or on non-EU venues that are found to be of equivalent status. Even if a share has a non-EU market as its most liquid venue, provided it is capable of being traded in the EU, the MTO will therefore have an effect outside the EU (take Apple Inc. as an example of such security). Again, this is not classic extra-territoriality – the rule does not apply to firms outside the EU. But it may impact upon the way that EU firms go about conducting their trading business, in particular if a sensible approach is not taken to finding that the most used non-EU venues are, indeed, equivalent.

DMA

MiFID II sets out detailed rules for trading venues relating to electronic, algorithmic, and direct market access trading. In summary, an EU trading venue must (to maintain its EU authorisation) ensure that members are only permitted to provide direct electronic access if they are authorised under MiFID II or CRD IV (or deemed equivalent under the third country rules). This restricts the ability of non-EU firms to trade directly on EU regulated markets. Again, this is not classic extra-territorial legislation – the rule only applies to the EU firm (the market). But it does have an extra-territorial effect in that it limits the ability of non-EU firms to do what they would otherwise want to do. It is something more like a "fortress Europe" measure (whereby barriers are erected) rather than an extra-territorial measure (which is an attempt to apply an EU rule to a non-EU firm).

Summary of MiFID's extra-territorial effect

So, as can be seen, and having scoured MiFID for examples, we think from the above that there are only two areas where MiFID has explicit extra-territorial effect – transaction reporting in branches, and the position limits regime. There are other areas where MiFID has an effect on non-EU markets (product governance, inducements, the Mandatory Trading Obligation, and DMA access). There are probably more – each time one delves into MiFID deeply, the implications can seem more and more widespread. But we think it is clear from the above that the extra-territorial effect of MiFID is limited. In general, MiFID applies to EU firms, and not others. 

We now need to test that assumption against a number of different scenarios.

Subsidiaries V branches

How does our analysis hold up in relation to the activities undertaken by subsidiaries as compared to branches? We think, in fact, it holds up very well. We have already identified one measure where there is a difference – transaction reporting by branches (non-EU subsidiaries do not need to transaction report trades that they do on their own book – non-EU branches of EU firms do).

There is, of course, a line of reasoning that goes in the following way. MiFID applies to firms. The non-EU branch of a MiFID firm is still a MiFID firm, because the branch is legally indivisible from its head office. Therefore, isn't everything the branch does caught by MiFID II?

We do not think that this is the case – indeed, it is quite clear, in MiFID I, that no one treats the rule in that way. And the FCA has (in the UK) already made clear that it does not take that view (see above). MiFID II contains nothing that makes clear a break from the past in this regard either – so the assumption must be that the activities of branches are still not caught by MiFID II, except where an explicit reference is made to the contrary, as with transaction reporting. It is difficult to see how anything at Level 3 could alter this position – such a key policy change would, surely, need to be enshrined somewhere in the legislation. (We look below at a best execution example which we hope illustrates this point.)

It is also worth considering MiFID's new third country regime. This is an attempt, for the first time, to harmonise the approach across the EU to incoming business from third countries who are deemed equivalent. If a third country passes an equivalence test, and permits equivalent access by EU firms, then this opens the door to the firm in that third country to do business into the EU without complying with MiFID's rules (on the basis that its national rules are equivalent). This measure has been subject to considerable scrutiny in the light of Brexit. It is also at least questionable whether it will ever be invoked given how radical and liberal a measure it would prove to be. But we think that this is further evidence that the extra-territorial effects of MiFID II are limited – surely the third country regime would have been the perfect place to cover the impact of MiFID outside the EU for non-equivalent jurisdictions if MiFID II had been intended to change the position compared to MiFID I.

Examples

Let us take an example in order to try to illustrate the point we are making about the limited extra-territorial application of MiFID.

A client in Singapore telephones a trading desk in Hong Kong in order to execute a transaction on the Hong Kong market. Let us assume that the Hong Kong trading desk is either a branch, or subsidiary, of an EU entity (we think it makes no difference, with the exception of the transaction reporting obligation for branches referred to above). Let us then assume that the transaction is booked back to the relevant EU entity (the head office, or the affiliate). To what extent do the MiFID II obligations apply to that transaction?

In our view, all of the activities undertaken up to, and including, the execution of the transaction take place outside the EU. Therefore, MiFID II does not apply to those elements. So local rules on the provision of any advice that might have been given to the client, on client order priority, and on best execution should apply.

Once the transaction has been executed, though, it is booked back with the EU. Therefore, at this point, the post-trade MiFID II obligations (trade reporting, transaction reporting etc.) would apply. MiFID kicks in as soon as the handling of the order in some way crosses back into the EU, in this case through the remote booking.

Now let us further expand on this example, and go beyond a single transaction to look at a series of transactions. Let us say that the activities of the trading desk in Hong Kong are such that, overall, they lead the EU firm to trigger the Systematic Internaliser limits. This does not, in our view, mean that the entity in Hong Kong has to meet the systematic internaliser obligations. So Hong Kong does not, to put it simply, need to act as a market maker. However, if the activities undertaken in Hong Kong are of a sufficient volume to lead the EU entity to trigger the Systematic Internaliser limits, then the EU entity would need to act as a systematic internaliser, and hold itself out with firm prices. It would be wrong to classify this as an extra-territorial effect – it is in fact the EU entity which is trading above the Systematic Internaliser limits. But it is an example where activities undertaken outside the EU could have a MiFID consequence.

Conclusions

We believe that the extra-territorial effect of MiFID has, on occasions, been both overstated and misunderstood. MiFID II is not, at heart, an extra-territorial piece of legislation. Where MiFID II is extra-territorial, it specifically says so (and we identify several instances above). On other occasions, the activity being undertaken must be tracked and the moment when responsibility is assumed by the EU MiFID firm must be identified. Requests for detailed guidance on these points may well disappoint, as they are unlikely to reduce the complexity of this analysis. We think there is sufficient clarity for firms to press on with MiFID II implementation in the meantime.