Which brokers are the most popular among European buy-side traders? To what extent are brokers using primary exchanges versus alternative venues? Which brokers are the biggest internalisers of order flow? These questions and many more can be answered from the latest MiFID II RTS 28 ‘best execution’ reports. Analyst Tim Cave highlights TABB Group’s top findings from analyzing the professional equities portion of the reports for 20 major sell-side firms and the top 100 European buy-side firms by assets.
Which brokers are the most popular among European buy-side traders? To what extent are brokers using primary exchanges versus alternative venues? Which brokers are the biggest internalisers of order flow, and how has that changed in the first year of MiFID II?
These questions and many more can be answered from new “best execution” disclosures under MiFID II, Europe’s new trading rulebook, which went live in 2018. Such information on order routing previously was highly confidential, but European regulators believe there is value in greater transparency over the execution process to help inform and protect end investors.
TABB Group’s latest report, “Opening Pandora’s Box: Exploring MiFID II’s Equities RTS 28 Reports,” analyzes the second set of annual “RTS 28” best execution reports, with the help of Cappitech, a provider of regulatory reporting and intelligence solutions for the financial services industry. As these reports refer to trades executed throughout 2018 – the first year that MiFID II was in force – they have been hotly anticipated by Europe’s trading community.
As a recap, MiFID II’s RTS 28 requires firms to disclose their top five execution counterparties in all asset classes in which they are active, for both retail and professional orders, and summarize execution quality achieved. Both buy-side and sell-side firms are required to disclose the top five venues on which they execute directly (RTS 28), as well as the top five brokers where orders are executed on their behalf (DA 65.2). The reports require disclosure on the proportion of total volume and orders executed with each top-five counterparty, and the split between passive and aggressive flow, as well as the proportion of directed orders (where the execution venue is specified by the client). Separate reports have to be produced for equities in three liquidity bands, broadly defined as large-cap, mid-cap and small-cap stocks.
Here is a selection of our findings from analyzing the professional equities portion of the reports for 20 major sell-side firms and the top 100 European buy-side firms by assets.
Market Share of Top Five Brokers Increasing
One long-held concern of MiFID II was that it might reinforce the dominance of larger organizations, based on the assumption that bigger firms of all types can simply devote more resources to compliance and implementation. The buy-side DA 65.2 reports can be used to assess whether this is the case in execution, by estimating the average proportion of buy-side flow that goes toward their top five brokers (see Exhibit 1, below). During 2017, this figure was 71.8% among the top 100 European buy-side firms in our sample group. In 2018, activity appeared to concentrate significantly, with the buy side’s top five brokers accounting for 82.7%, 75.2% and 79.8% of activity in small-cap, mid-cap and large-cap instruments, respectively.
Exhibit 1: Average Proportion of Buy-Side Equity Order Flow Sent to Top Five Brokers
Source: Cappitech and TABB Group analysis of RTS 28 Reports
In short, the biggest brokers appear to be getting bigger. This may be a positive development if the buy side experiences an improvement in execution quality with a more focused group of counterparties. The concern is the potentially damaging influence wielded by a smaller group of large brokers – both in terms of choice and for competitors attempting to get a foothold in the market.
Agency Brokers on the Rise
As the market share of the top five brokers is increasing, the firms within that top five are beginning to change. While an inexact measure, we have measured Europe’s most popular top five brokers by the number of mentions in DA 65.2 reports among our sample group of 100 buy-side firms. Exhibit 2, below, shows the top 20 most-mentioned firms in 2017 and 2018, ranked as a percentage of total mentions (total mentions were much higher in 2018 because most buy-side firms split activity by liquidity band).
Exhibit 2: Europe’s Most Popular Equity Brokers by Mention in RTS 28 Reports
Source: Cappitech and TABB Group analysis of RTS 28 Reports
This tells us several things. UBS was the most-mentioned broker in both 2017 and 2018. JP Morgan was a significant riser in 2018, moving from fifth to second-most mentioned. Significant fallers included Bank of America Merrill Lynch, which fell from second to seventh, and Deutsche Bank, which fell out of the top 10.
But perhaps most interesting is that all three major agency brokers – ITG, Liquidnet and Instinet – are among the top 10 most-mentioned firms during 2018, ranked eighth, ninth and tenth, respectively. Only ITG was in the top 10 in 2017. This is most likely explained by MiFID II’s unbundling provisions, which force the buy side to choose execution counterparties on the basis of execution quality alone, rather than as a way to pay for other services, such as research.
The shift toward agency brokers under MiFID II becomes more apparent when viewing total mentions by firm type (see Exhibit 3, below). As a group, agency brokers accounted for 12.8% of total mentions in 2018, compared with 7.3% in 2017. This was the biggest year-over-year increase of any firm type, though there were also increased mentions for tier two/regional banks and electronic liquidity providers. By contrast, relative mentions fell for both European and US bulge-bracket banks. The key caveats here are that more mentions do not necessarily translate into higher market share, and some OTC activity undertaken by bulge-brackets might not be caught within RTS 28 reporting. Year-over-year comparisons are also distorted by the fact that total mentions were higher in 2018 because most buy-side firms split activity by liquidity band.
Exhibit 3: Europe’s Top Five Equity Brokers by Type, by Mentions in RTS 28 Reports
Source: Cappitech and TABB Group analysis of RTS 28 Reports
Internalisation Falls and Focuses on Small-Cap Names
A closely watched figure in the sell-side reports is the level of internalisation. A like-for-like comparison is difficult to achieve in this area given the 2017 reports covered a period in which broker crossing networks (BCNs) existed. BCNs were operated by the large investment banks effectively as mini off-exchange trading venues, used to execute client orders against one another on an agency basis, using the bank’s own capital and liquidity provided by third parties. Most major banks, but not all, included internal entities (with the MIC code XOFF) in their 2017 reports that likely represented BCN activity to some extent. Under MiFID II, BCNs were banned and replaced with an enhanced SI regime, the regulatory designation for capital commitment. Orders that are internalised by banks under MiFID II must be undertaken in their capacity as an SI and involve the use of bank capital. The intention of the rules was to encourage more volume back onto public markets and prevent banks internalising agency crosses.
It is logical to assume, therefore, that some internal activity (XOFF) will have transitioned into SI entities in the 2018 top five reports. This shift will have depended on multiple factors, including: the bank’s routing strategy; investments in risk provision and a bank’s ability to commit capital; restructurings heading into MiFID II; and the specific nature of MiFID I BCN activities.
The results are intriguing. In 2017, internal (XOFF) entities were mentioned in top five reports by 10 brokers in our sample sell-side group of 20 firms and were used to execute 26.1% of order flow on average among those firms (see Exhibit 4, below).
Exhibit 4: Proportion of Equity Flow Sent to Internal SI
Source: TABB Group analysis of RTS 28 Reports
In the 2018 reports, SIs were mentioned by our sample sell-side group most often as a top five venue for small-cap stocks (11), followed by mid-cap (9) and then large-cap (8). However, the percentage of volume handled by SIs was lower, on average; SIs executed 20.4%, 13.9% and 18% of order flow for small-cap, mid-cap and large-cap stocks, respectively. The higher utilization of SIs in small-cap names probably reflects the difficulty faced by investors in sourcing liquidity in these stocks on-exchange.
There some wide variations in SI levels, reflecting different business models, client type and overall success at risk provision. Goldman Sachs, JP Morgan and Credit Suisse were the heaviest users of their own SIs across all three liquidity bands. Bank of America Merrill Lynch and Morgan Stanley sent a relatively lower amount to their own SIs, and they also saw the biggest falls in XOFF versus SI activity. At BAML, 31% of its activity was internalised in 2017, but it used its SI for 4.1% of activity on average during 2018 across all three liquidity bands. Morgan Stanley internalised around 27% of activity during 2017, but its own SI was used to execute 9.7% of activity during 2018 on average across all liquidity bands.
There could be several explanations for this. These banks operated some of the biggest BCNs under MiFID I and may have been the most impacted by the ban on these venues. This might imply they deployed minimal amounts of their own capital through their BCNs, with the majority of activity made up from matching client orders on an agency basis and executions from third-party liquidity providers.
Directed Orders and DMA Providers
While BAML and Morgan Stanley appear to have sent less flow to their own SIs, this figure may have been diluted by the fact that these brokers routed larger volumes to exchanges. It is possible to gain some insight into Europe’s biggest equity brokers by volume from the sell-side RTS 28 reports. Brokers are required to disclose, for each top five counterparty, the proportion of order flow which was directed – meaning the client specified the venue on which it wanted to execute. This can generally be viewed as a proxy of direct market access (DMA) activity, a service allowing trading firms to access markets directly using the technology/connections of a broker. It is used for various reasons, including by proprietary trading firms wishing to access exchanges anonymously, and by high-volume buy-side firms without exchange memberships.
Those brokers with the biggest proportion of directed order flow are likely to have the highest share of on-exchange activity. Among our sample group of 18 sell-side firms, nine had meaningful levels of directed order flow on their top five venues (see Exhibit 5, below).
Exhibit 5: Average Proportion of Directed Orders sent to Top Five Venues
Source: TABB Group
Those brokers with the highest proportion of directed orders were Instinet, Bank of America Merrill Lynch and Morgan Stanley. This fits with what we know about these brokers – they are all big providers of DMA. At Instinet, 44% of large-cap order flow, 29% of mid-cap flow and 1% of small-cap activity came from directed orders. There were some interesting highlights within these numbers. At BAML, 82% of the small-cap executions it sent to Nasdaq Stockholm were directed orders – amounting to 9.3% of its small-cap volume overall. DMA services are popular in the Nasdaq Nordic markets because, until recently, it operated a model whereby counterparties are identified on a post-trade basis. Using DMA services helps provide prop trading firms with anonymity.
Agency brokers ITG and Kepler Cheuvreux both had a relatively high proportion of directed orders (43% and 18%, respectively), but these were not broken down by liquidity band. ITG was something of an anomaly in that the majority of directed order flow was to its own block trading venue, POSIT, which its buy-side clients can access directly.
As with many of MiFID II’s new transparency regimes, RTS 28 reporting is still in its infancy and inconsistencies exist in the way information is disclosed. There are also some natural limitations to the reports: For some firms, their top five brokers/venues comprise a limited dataset, and the figures are all relative, as opposed to absolute. Comparing the 2017 and 2018 reports is particularly challenging given the shift from MiFID I to MiFID II and the fact that the 2017 reports were produced on a “best efforts” basis and were incomplete in many cases. The real value in the reports – to regulators and others – is likely to come over a longer period of time. Even so, it is still possible to draw a number of intriguing observations about execution services provided by the sell-side and the shifts brought about by MiFID II.
For a deeper dive into the second set of annual RTS 28 best execution reports, please contact TABB Group for information on our latest research, "Opening Pandora's Box: Exploring MiFID II's Equities RTS 28 Reports."