Best Execution In MiFID II
With Arjun Singh-Muchelle, Senior Advisor, Regulatory Affairs, Institutional and Capital Markets, Investment Association
There are two main requirements which have an impact on the buy-side with regard to MiFID II and best execution. The first is within the MIFID-delegated acts, which look at our ability to ascertain best execution from our brokers. That means looking at the application of best execution to other financial instruments, as previously there were no requirements to demonstrate best execution for non-equity instruments. There is now an expansion of those requirements into cash bonds, derivatives, FX forwards and so on. We are now required to demonstrate best execution on those instruments as well whereas previously, it was not the case.
We are now looking towards third party providers to offer TCA for FX or TCA for cash bonds etc. To date, these service offerings have been somewhat lacking, so we have started from scratch for non-equity instruments. We are also wary of the application of equity TCA to non-equity instruments. One of the reasons for this is that equity TCA is often focused on using VWAP as the primary benchmark whereas (in our view) for equity best execution, VWAP should not be the only benchmark. It may also be difficult to simply copy these concepts across into non-equity instruments, but that is what is being offered by the third party TCA providers. In addition, the way one executes an FX forward contract or executes an illiquid corp or sovereign bond is different to how one would execute an equity instrument.
The regulators have also redefined best execution from all ‘reasonable’ efforts to all ‘sufficient’ efforts. It is not clear what the difference means but the regulators have purposely changed the definition. As a result, additional thought will have to be applied to the consequences of that linguistic shift; do we have to plug into every execution venue in Europe to meet that requirement, no matter how small the exchange might be?
There are many shallow books on exchanges and venues across Europe, and we need to know whether we must start plugging into all the execution venues in order to see whether we can get a better price. It would require new systems development to actually plug into, for example, the Bratislava Stock Exchange. Technically, it would also mean providing additional information to other market participants (which is not normally provided), which would increase the probability of information leakage. There is also the additional impact of the cost of market data in real time from the additional venues, which is a considerable political issue and without a consolidated tape this remains difficult and expensive. There are some provisions for a consolidated tape for equities in MiFID II, but there is no mention of a non-equity consolidated tape offer for bonds, FX or derivatives.
The second issue relates to the reporting requirements for best execution under RTS27 and RTS28. RTS27 is data that goes from the broker or venue to the asset manager. RTS28 is best execution data from the asset manager to the underlying clients – to the funds. The concern with RTS27 is the way it is currently formulated as unexecuted client orders will be made public. In our view, an unexecuted order should never be made public because it gives a false impression of whether a venue or broker is actually within our ‘Top 5’ or not. It also provides additional information to other market participants of holdings that we may have tried to execute previously (that may have been unsuccessful) and/or where we have used an RFQ for a fixed income order, as we would use the RFQ process as a valuation tool.
Another issue with RTS27 is that a Systematic Internaliser Operator would have to disclose all transactions (in aggregate format), even if they were large and had benefited from the large in-scale order protections – all that information will now be made public. Even though there is an in-built delay of at least three months, it is not a sufficient amount of time for brokers to unwind the positions they have taken to facilitate our trade, which then has a downward impact on asset managers. If brokers are going to be exposed to additional or undue market risk, that impacts upon their ability to make markets for us at efficient prices.
The concern we have with RTS28 is more technical. At a high level, the concern relates to who is ever likely to read this information. In the UK there is a ‘pension disclosure code’ – an annual statement given to our providers and everyone with a holding in the pension fund, which in reality is very rarely read by anyone. We will now be required to provide them with an additional 35-40 pages of best execution data. This would contain information about whether the firm acted in an aggressive or passive capacity, what percentage of the trades were with the Top 5 etc. We are also required to disclose any close links with brokers or venues without having any comprehensive definition of what ‘close link’ means. For example, does a shareholding in the London Stock Exchange now need to be disclosed as a close link?
There are other technical issues around delegated executions. If a fund manager is managing a fund domiciled in Luxembourg on behalf of a US client and the order is generated in London but the execution was delegated to the in-house dealing desk in Hong Kong, then is that included in the fund manager’s RTS28 reports or not? We don’t know. If amongst the top 5 brokers, three of them are Morgan Stanley (Morgan Stanley New York, Morgan Stanley London and Morgan Stanley Singapore) can they be aggregated into one as Morgan Stanley? These are the sorts of issues and concerns we have around the best execution reporting requirements.
In our conversations with regulators across the EU, we have referred to the potential impact on global asset managers and other regulators. And whilst talking about US, Asia-Pac brokers and regulators, I don’t think they fully understood the ramifications of the best execution requirements on their zones either.
In addition, RTS27 rules say that the asset manager is legally obliged to receive the information, meaning that a broker must provide it in the first place in order for it to be received by the asset manager. But there is no legal compulsion on a US or an Asia-Pac broker to provide that information. So if a US broker refuses to provide that granular data, then who will be held legally responsible – will it be the US broker or the European-domiciled asset manager?
These issues give the impression that the policymakers have not necessarily appreciated the extent of the technical and financial challenges that will be faced by asset managers in order to achieve best execution, or at least to demonstrate their achievement of best execution.
We’d love to hear your feedback on this article. Please click here