Making the Best of it: Best Execution in Europe
Citi’s Salvador Rodriguez and Daniel Mathews explain how best execution has evolved alongside MiFID and how the latest proposals are likely to affect buy-side and sell-side trading desks.
How is best execution under MiFID II different from MiFID I and from pre-MiFID? Where have we come?
Daniel Mathews, Citi: Although we are referring to MiFID II, we are at the early stages of the MiFID II legislative drafting process. The European Commission (EC) published its proposals last October, and there will be a number of amendments proposed by the European Parliament over the coming months. The Council of the European Union (EU) will also table amendments and ensuing agreement between the EC, the European Parliament and the Council of the EU (27 member states) will then be required. It is not yet clear what MiFID II and MiFIR will look like in final form as there are a number of key areas which are acknowledged by both sides to need addressing. What is clear is that there will be significant changes to the drafting that is on the table at present.
Salvador Rodriguez, Citi: With the MiFID II process we have seen closer alignment between buy-side and sell-side interests. There is more cooperation between the buy-side and sell-side with a view to what may or may not come out of MiFID II; and this in itself, is a clear improvement from the earlier iterations, which is encouraging for the business at large.
DM: Certainly, the meetings we have had with the buy-side indicate that they are taking far more interest in what MiFID II will mean for them; they want to participate in the debate and are keen to understand what we are doing from a sell-side perspective.
Tools and strategies that have become accepted best execution for many brokers are now under review (e.g. broker crossing and the Systemic Internaliser (SI) regime). How difficult will it be to provide comparable offerings within the SI framework?
DM: One of the current challenges is understanding how the SI regime will operate under Ferber’s proposals and understanding the unintended (or intended) consequences of the amendments, such as ‘all OTC trades must be conducted under the SI regime’. The scope of the SI only extends to liquid stocks, so what happens to non-liquid stocks? Can we trade them outside an SI? Will all risk trades need to be executed within SI and therefore within published firm quotes, even though a risk trades to client may warrant a price outside firm quotes? At the moment, there are many unanswered questions raised from Ferber’s proposals concerning what the SI regime actually means for us and our clients.
SR: As Dan has alluded to, many of the requirements have fallen into the lap of the sell-side. From a trader’s point of view, the tools, strategies and decisions around how to execute a trade will probably not change significantly. Traders will continue to send VWAP or participate with volume-type orders. Naturally, there are pending questions around how risk is employed and whether firm capital can be used within an OTF environment. As a broader business, how clients and orders interact with risk, and how we internalize house flow, are wide-ranging questions. There is no one clear answer; it is a multi-layered problem and there are still grey areas to be resolved.
From which of the MiFID updates will institutional investors notice the biggest change in execution quality and/or strategy?
SR: The MiFID proposals will clearly affect everyone. Under the Commission’s proposals, a BCN would be an OTF and an SI is not a venue. I think this needs to be unpacked. Legally the classifications are fairly clear – you can trade on either a trading venue (RM, MTF, OTF) or OTC and if the latter then to the extent the trading is systematic and frequent then the firm must be an SI. As Dan says though, Ferber’s amendments have muddied the waters so that it’s not clear how an SI is intended to function. Our job is to figure this out through dialogue with our buy-side clients, and we have been seeing many of them recently on market structure road shows, explaining where the current process is at and the areas still to be resolved.
DM: Our clients value choice and one reason they participate in broker crossing networks is because they trust and value the liquidity in them. There can be a level of discretion in how they interact with different types of liquidity within the pool. Under MiFID II, it is questionable whether clients will be have the same choices; being able to participate in broker pools and have that same level of ‘discretion’ as to what kind of flows they interact with.
SR: One of the things I touched on at the EMEA Trading Conference in March was ensuring that customers have a menu of choice, particularly in dark pools. There is no ‘one size fits all’ solution. For example, some traders are aggressive, others are more passive, so you are never going to have a solution that suits everyone. As brokers we need to tailor solutions to particular client needs. It falls back to the ‘know your client’ adage, so discussion with them and understanding how they want to execute their business helps us to identify and understand consensus in the industry. While we want consensus on how to deliver solutions to a client, it is equally important to understand the bespoke execution requests to ensure that we align their expectations with what we can deliver.
DM: Within our broker crossing system, there are different types of flow – for example, client flow and principal flow from unwinding of risk trades. Both generate important sources of liquidity. Under the EC’s proposals, clients can interact with an OTF but we could not deploy our own capital (i.e., principal flow from unwinding risk trades) within in the OTF. This means that our clients would be unable to access the liquidity of our risk trade unwinds. It’s back to investor choice – clients value the choice of interacting with different types of flow within broker crossing networks and under MiFID II they would not be able to do so. In addition, under the Commission’s proposal BCNs would be OTFs; under Ferber, OTFs are limited to the non-equities space.
How is greater post-trade transparency, both in terms of a consolidated tape and reduced trade reporting delays, going to affect best execution?
DM: A consolidated tape will help post-execution analysis, but it is not necessarily going to change trading behaviour, per se. Establishing the consolidated tape from various venue feeds will increase latency compared to direct feeds from trading venues. Brokers will still need to subscribe market data directly from trading venues, whether it is a MTF, an OTF or primary exchange, to ensure they have the lowest latency of market data against which to execute their trading strategies.
SR: The best execution requirement states that we must ensure that we are accessing the correct venues, at the correct times, for the correct liquidity and at the correct price. This is an important point because historically brokers have always had access to the post-trade element. We have recently seen the rise of third party providers offering this service to the buy-side, and more sophisticated clients who now have their own tools and personnel, with full-time staff to analyze the executions that we, the broker, are feeding back to that client. Not only are we being measured against our own internal TCA, but also against third party TCAs and clients with their own dedicated staff that measure each execution.
Alternative venues gave European brokers greater choice in execution, but also added complexity and cost in terms of IT and connectivity. What will happen to the balance of cost and choice under the new OTF and MTF regime?
SR: There are undoubtedly going to be costs attached to this, in the sense that brokers have to ensure that they fully comply with MiFID II. There will be ongoing costs in terms of personnel, and if we need to make changes to areas such as the MTF, OTF, or SI regime, then there may be adjustments in quant and IT resources, which might not necessarily increase costs but will divert attention.
DM: There will be more trading venues – We have already seen this as firms begin to pre-empt the outcome of MiFID II. There is an overhead to joining any trading venue and a substantial uplift in the amount of market data that needs to be consumed and processed.
Will it be harder for a broker to get a new algo to market under MiFID II? Why, or why not?
DM: This depends on the definition of algo trading, whether it is the EC’s definition of an algo or Ferber’s definition. Under the EC’s definition, any “computer algorithm” was considered an algo and therefore fell under the obligation to provide a continuous two-way quote, which is not going to work for clientstyle VWAP algos. Whereas, if you look at Ferber’s amendments, he splits out algorithms used for client orders from algorithms used by high frequency trading (HFT) strategies. It is unclear which approach will prevail although we know that neither will stand as it is drafted currently. In addition, the recently published European Securities and Markets Authority (ESMA) Guidelines on Automated Trading would suggest that it will be no more difficult for a investment firm to introduce a new algo to the market than it is today.
SR: Client directed flow – the traditional VWAP, Participate with Volume business – will not be very different. There might be one or two new rules but as we mentioned earlier, we have dedicated staff within our legal and compliance departments who will monitor, oversee, and approve any new strategies, amendments or customizations. Gone are the days where you simply delivered an off-the-shelf VWAP algorithm to a client. Clients are more demanding regarding customizations and adjustments to meet their particular requirements. Once we have identified the client’s specific needs, we then have an implementation process. We discuss it internally, involving our quant and product teams and agree what can be done from a technical or quantitative point of view. Throughout this process, we also liaise very closely with our compliance colleagues before delivering an execution solution to our clients.
DM: There is already a high degree of oversight and governance around algos as it is part of the regulatory framework in which we operate. There will be small tweaks that we have to make, for example, having the traders’ IDs reported to the regulator. Non-regulated entities, like hedge funds, will need to ensure that they have the correct oversight in place, but from our perspective and that of most large brokerage houses, it will be business as usual. Sometimes we forget that many of the firms operating these algos are already thoroughly regulated by the Financial Services Authority (FSA).