Some of MiFID II’s consequences were widely anticipated, such as the rise of block trading and the migration of order flow towards systematic internalisers (SIs) or other venues, such as periodic auctions. But the rise of request for quote (RFQ) in equities trading was one of the trends admittedly unforeseen by the majority of market participants grappling with MiFID II compliance.
Typically used for trading in more illiquid markets, such as fixed income, exchange-traded funds (ETFs) or derivatives, RFQ allows traders to ‘request a quote’ from counterparties based on the security and quantity. However, various senior figures on the buy-side have expressed their scepticism of deploying RFQ in the equities world, questioning the logic behind launching such a tool in a liquid market. Despite this, two major trading venues confirmed plans to launch RFQ functionality for equities trading six months into the MiFID II lifespan, with plans to go-live by the end of 2018.
As one of the early pioneers of the electronic RFQ protocol in fixed income, derivatives and ETFs it’s perhaps little surprise that Plato Partnership knocked on Tradeweb’s door with the idea of building an RFQ platform for cash equities, known as eBlock, which aims to provide buy-side traders with the ability to source and view in aggregate broker principal risk and execute orders on a regulated venue via Tradeweb’s multilateral trading facility (MTF) using RFQ.
The London Stock Exchange Group (LSEG) is the latest to launch an RFQ system for equities, following in the footsteps of Tradeweb, and agency broker Instinet which was first off-the-mark in January. The exchange operator said that it had taken the traditional and manual RFQ functionality it has for ETF trading and upgraded it for equities to provide a more automated solution. The tool looks to provide traders with a fully automated, on-exchange and centrally-cleared RFQ service, removing the need for various bilateral relationships, and alleviating the technological costs and burdens often associated with sell-side access.
Tradeweb and LSEG are both in the early days of the launch of their respective RFQ systems for cash equities, and both systems are similar in terms of what they offer, although there are several key differentiators between the two. For example, LSEG’s RFQ functionality offers automated execution, central clearing and the opportunity to request quotes anonymously if the trader wishes to do so. On the other hand, Tradeweb and Plato Partnership’s tool functions with no anonymity and full disclosure of the counterparties involved in the transaction, but with the ability to hide if they are looking to buy or sell from the RFQ. Such differences are key to the concerns outlined by the asset managers on RFQ for equities.
Last chance saloon
With the context of market structure changes caused by MiFID II, the logic behind launching RFQ for equities trading comes into focus. MiFID II encourages risk trading and on-exchange activity, having banned broker crossing networks (BCNs) and placed restrictions on dark pool trading via the double volume caps (DVCs), which trigger bans on certain types of trading when a transaction accounts for 4% of the total activity on a single dark venue, or 8% of total trading market-wide. Under the regulation, RFQ is listed as one of five pre-trade transparent trading methods, meaning it is exempt from the DVCs and it is considered to be lit trading activity.
Regardless of this, genuine demand for the product on the buy-side has been difficult to find. Senior heads of desks and traders have instead expressed their bewilderment about why a trader would willingly and voluntarily send out signals to a market usually considered to be liquid and risk creating a lot of unnecessary noise. Many buy-siders didn’t foresee the rise of RFQ in equities and remain uncertain about where the execution protocol will sit in a market which is still adapting to MiFID II and the introduction of new block trading venues, periodic auction systems and SIs.
“RFQ for equities is just a potential alternative way of accessing different sources of liquidity, and a lot of buy-siders are sceptical about the idea,” says Dan Nicholls, head of trading at Hermes Investment Management. “Since MiFID II has tried to limit access to traditional liquidity sources with the DVCs to drive liquidity to lit venues, the sell-side is forced to develop alternative routes to match the best price. To misquote Jurassic Park: ‘Liquidity finds a way’. The RFQ works well for fixed income and ETF trading, but these are very different instruments to equities.”
Nicholls adds that the RFQ is often perceived on the buy-side to be the ‘last chance saloon’ when sourcing liquidity, explaining that information leakage tops the list of concerns around RFQ for equities, and remains curious as to why a trader would RFQ a sell order in a rising market and, conversely, a buy order in a falling market.
“The signal to the market is that the order is in the same direction as the market move and that you can’t find any natural liquidity, which is powerful information,” he says. “I would be happy to use RFQ for equities to finish an order on risk, because the liability for the unwind then lies with the risk provider. However, I would not use RFQ at the start of a multi-day order, as there is no way of knowing the origin of the price. If it is natural or an unwind, then great, but otherwise the order starts to compete with the unwinding of the risk price.”
For Neil Joseph, head of equity trading at JP Morgan Asset Management, and Ed Wicks, global head of trading at Legal & General Investment Management (LGIM), the RFQ for equities trading is a compelling trend. But similar to Hermes’ Nicholls, overarching concerns with information leakage means that it is unlikely the system will be widely used across their respective equities trading desks.
Joseph adds that the reasons behind the development of each RFQ for equities platform are unique, but scenarios where the RFQ would be used tangibly, and as the platform providers hope, are in some ways restricted: “There are conceptually some valid use-cases for RFQ in equities, although these are currently limited,” he says. “For our specific order flow, it will likely be utilised for a lower proportion of orders than RFQ for ETFs and in a different manner.”
LGIM’s Wicks recognises that RFQ for equities has its relevance for some market participants, but for smaller players that don’t have the ability to transfer risk as seamlessly as Legal & General, alongside integrated actionable indications of interest (IOIs) and direct-to-capital type arrangements with brokers, could experience problems.
“The equity RFQ model as it stands is an interesting concept and it will have relevance for some people,” says Wicks. “If I am very honest, it is not number one on my to-do list, I don’t see the relevance for our book of business currently. There are other ways to transact your business that result in potentially less information leakage. I can see that procedurally you may be able to tick some boxes from a best execution perspective, but you have to be cognisant that you are surrendering information when you are going out on an RFQ.
“As long as you are aware of the risks and you mitigate them to the best of your ability, then it can have relevance for certain asset managers, and maybe it will for us in the future, but as the model stands, I don’t think it is something we are going to be driving particularly aggressively.”
Among the buy-side, information leakage is considered to be more of an issue with the Tradeweb-Plato Partnership RFQ model, as eBlock is fully disclosed, meaning that both counterparties are identifiable throughout the process. In contrast, the LSEG RFQ platform provides the user with the option to trade anonymously.
Adriano Pace, managing director for equity derivatives at Tradeweb, and Richard Bateson, product manager for equities, who was hired by Tradeweb from Macquarie in May 2018 to lead the RFQ for equities project, argue that the tool should be used cautiously and with information leakage in mind. Pace stresses that eBlock is not an uncontrolled RFQ, but it was designed to provide the buy-side trader with authority on the negotiation and control of information, particularly when it comes to interaction with market makers, or liquidity providers (LPs).
“The buy-side user has full control over who they go to and the type of IOIs they interact with – all of that is at their complete discretion,” Pace adds. “So the control is completely with the buy-side, and in fact, they can decide which type of orders they want to send to RFQ – knowing the name, and size of the trade, the LPs they want to interact with, which should give them a lot of comfort in that regard.
“When the buy-side trades on the lit exchange they don’t know who they are trading with, maybe an ELP or through an ELP with an SI. At least with our protocol if they want to trade with an ELP, they do so with their eyes wide open and that’s under their control. For them, that is potentially a better option than trading anonymously with a smart order router with an ELP and getting picked off.”
Discussing the LSEG’s RFQ for equities model and the topic of concerns over information leakage, Scott Bradley, who currently heads up sales and marketing for LSE cash markets and Turquoise, explains that there are certain situations where a trader would not necessarily want to RFQ anonymously, but providing users with the option to request on a named and unnamed basis is key to ensuring the buy-side is truly in control – which is particularly important for interactions with market makers.
“If you choose to be named you are making that informed decision that you wish for activity to be highlighted to a selected number of LPs because you have chosen to be named,” Bradley says. “But that’s where the power of bilateral relationships come into play because you will choose to go named when you have a good relationship with that market maker. So the requester in both cases, named and unnamed, is really taking control over that information sharing. So the expected order size, the fact you can control who you request a quote from, and whether you decide to be named or unnamed are all facilities that enable the requester to take full control of the information.”
For the buy-side, information leakage seems to be at the top of the list of concerns with using an RFQ system in equities, but it isn’t the only apprehension. Fabien Oreve, global head of trading at Candriam Investors Group, highlights that with actionable IOIs now firmly placed in the market, it’s difficult to see where RFQ will add value to equities trading.
IOIs have suffered a bad reputation in the past, in terms of defining which indications were representing true liquidity, but were reformed a few years ago following the introduction of an industry Code of Conduct which was put together by the Investment Association and the Association for Financial Markets in Europe (AFME).
“We do not see any interest in using electronic RFQs for cash equity, because we currently have easy access to IOIs from our brokers who advertise reliable, tradable sizes and prices,” Oreve says. “IOIs have gained much in quality since the European industry agreed on an IOI reform and a new code of conduct some years ago. The IOI reform introduced a classification model which has improved transparency and helped investors like us find liquidity. This model is helpful, as it makes a clear difference between IOIs that reflect natural liquidity and those that reflect non-natural, principal liquidity.”
Oreve adds that by using IOIs, decent size orders can be filled in their entirety with predictable market impact: “From my point of view, the RFQ protocol in cash equity does not add value to the current IOI model in place. RFQ is best suited to fixed-income where many instruments do not trade very frequently.
“Electronic RFQ has also become popular in some derivatives like options and in the ETF space where the role of market-makers is predominant. RFQs help evidence best execution but, if they are overused, they can make a lot of noise and risk moving the market. Investors need to be very careful how they use them.”
In terms of where the IOIs fit in this particular space it’s important to note that the LSEG RFQ for equities model does not interact with IOIs, but the eBlock model allows risk providers to stream actionable IOIs into the system. In response to the notion that RFQ is redundant with IOIs firmly placed in the market, Tradeweb’s Pace highlights that not all asset managers have the technological power to manage such streams of information, but the eBlock platform opens up that avenue for those buy-side firms that are perhaps struggling to consume IOIs.
“The reality is the wider part of the buy-side community doesn’t have that easy access to IOIs or don’t find them particularly helpful,” he says. “It’s not every buy-side that has the luxury of having access to the tier one bank quotes, and the ELPs might not be within that construct. With eBlock, we are essentially widening out and getting a much broader universe with the ability to interact with the actionable IOIs.”
Alongside questions around information leakage and the importance of IOIs, there is also concern among the buy-side when it comes to central clearing. The LSEG’s RFQ for equities tool provides automatic central clearing upon completion of a trade, similar to trading in the exchange’s central limit order book. Centrally cleared RFQ in equities means that the buy-side do not need to be papered with the market makers or risk providers on the platform. This benefits both counterparties by freeing up balance sheet and providing liquidity without necessarily entering into multiple bilateral relationships.
In contrast to this, the Tradeweb-Plato Partnership RFQ model does not provide central clearing for users, so the buy-side has to settle trades directly against the counterparty, and in some cases, that will be with an ELP. Neil Bond, head of trading and partner at Ardervora, says he finds it “a little odd” that the industry is excited about the rise of RFQ in equities as the buy-side weren’t necessarily “screaming out for it”. Noting its validity in less liquid markets, he adds that he is unsure how the trend will play out in the future, but eBlock’s lack of central clearing could prove to be problematic for Tradeweb and Plato.
“Without central clearing you have to clear against the counterparty, but there are very few buy-side’s that have direct relationships with the ELPs,” Bond explains. “I think the RFQ venues are expecting the buy-side to onboard with those ELPs directly and settle against them, but I don’t think that will happen in a significant way. When connecting directly to an ELP, what happens is you have someone firing in thousands of bids and offers a second, so you must have the infrastructure to handle that type of information flow. The buy-side simply doesn’t have that and that’s why I don’t think you’ll have users going direct. But that could all change.”
Tradeweb’s Pace counters that on the risk side, when clients trade bilaterally either through direct capital or an actionable IOI with a broker, they effectively net down against that broker at the end of the day, unless it’s on-exchange. He adds that this will be no different with the RFQ for equities protocol. Looking back on the success of Tradeweb’s RFQ functionality for other asset classes, Pace is confident that as the universe of dealers on the equities platform expands, so will the buy-side’s comfort in interacting with them.
“You have to offset the fact that it’s bilateral, so you settle directly with the view that in being bilateral and disclosed, we believe it gives clients more pricing power by not being anonymous,” Pace explains. “I get that being anonymous and centrally cleared RFQ might save 10% in terms of compression, but we believe that being disclosed far more outweighs the pricing power clients get compared to participating as an anonymous counterparty.”
Regardless of the distinctions between the systems, advantageous or not, it appears as though the RFQ for equities model has been developed and established without any serious demand from the market. But the LSEG, Tradeweb and Plato Partnership are not the pioneers of this particular trend. Agency broker Instinet quietly developed the function at the same time MiFID II came into force across Europe, with the firm highlighting that the demand for the product came from its own business.
Ben Stephens, head of business development for Instinet Europe, explains that the RFQ for equities has been central to Instinet’s MiFID II strategy and it solved the problems the regulation presented through restrictions in over the counter (OTC) trading. He says Instinet required a venue that would allow it to trade in a regulated fashion by moving the matching on-venue whilst at the same time aggregating and recombining the liquidity out there to help clients navigate the new landscape.
“Instinet operates a trading venue and we are a broker,” Stephens says. “We felt we had to build the best venue out there to solve the problems we were facing as a broker as a result of MiFID II. There was no way we could have preferenced our dark pool over another, unless there was something special about it. This is the same for our RFQ venue – if we can’t get a better performance out of the RFQ venue then we can’t really use it.
“For our own broker to use our own venue we have a much higher bar to cross, hence we did a lot of thinking around how we could solve our own problem as we couldn’t see a solution in the market. If you’re a bank or you trade as principal, then you have SI options that can solve a lot of problems, but we don’t and MiFID II prevented us from trading OTC and doing client crosses on the desk, therefore RFQ is very much central to our MiFID II strategy.”
Instinet’s BlockMatch RFQ is fully automated, centrally cleared and allows participants to stream IOIs straight into the system. Regulatory changes in Europe and Instinet’s position in the market meant that it prioritised conditional trading opportunities and the RFQ was key to this. Stephens adds that order size is another important specification here, with both LSEG and Tradeweb targeting larger-sized orders with their respective RFQ models. Tradeweb has said that its RFQ is aimed at orders in the value spectrum of around $500,000 to $3,000,000, whereas LSEG has said the RFQ is for trades from a minimum of 25% of LIS or £50,000, the larger of the two.
Instinet’s BlockMatch RFQ, on the other hand, accepts trades at any size, which Stephens adds is significant for the conditional trading opportunities the broker wanted to centralise for its clients. “We have gone the whole way and said we can do RFQ trades at any size, and that’s a key differentiator between our RFQ venue and others,” Stephens explains. “What happens is, if the quote is able to be seen on the platform then it has to be public. This is in contrast to other models where quotes are all above LIS and nothing is made public, then after a certain amount of time, there’s an auto-execution. This is significant because it means that someone has to commit ahead of time to trade, whereas ours is inherently conditional.”
Ardevora’s Bond echoes Stephen’s sentiment on targeted trade size and says that the RFQ has worked successfully in the retail industry for a long time, but has failed to translate into the institutional world because those investors require larger sizes. However, the RFQ for equities models targeting large sized trades in equities, as both LSEG and Tradeweb have, could face another barrier to adoption from the buy-side.
“The two new services are looking at doing larger size, and that works well for fixed income and ETFs when you’ve got the broad-based underlying securities or large balance sheets behind it, but with individual equities, I’m not sure there will be the appetite from market makers to provide large in scale quotes inside at the touch or better for single equities.”
Since going live with BlockMatch RFQ at the beginning of the year, Salvador Rodriguez, Instinet’s head of electronic trading for Europe, the Middle East and Africa (EMEA), claims that of all the asset managers currently onboarded and using BlockMatch RFQ, Instinet is yet to see a buy-side firm opt out of the service.
“We have effectively reengineered a new process that allows all of our other systems, including our algo trading, our EMS blotter scraping technology, smart order router and FIX connectivity, to access conditional opportunities,” Rodriguez adds. “The power of conditional trading in a MiFID II world has been something we have focused on and that’s why the RFQ matters. Ultimately, we’ll have one router if you like, and if you show interest we will find you the other side. From a client-facing perspective, it’s about how do we solve the liquidity conundrum and increased fragmentation that we’re seeing.”
Listening to buy-side sentiments on the idea, RFQ in equities may have quite a way to go before it achieves market-wide adoption. eBlock, with no central clearing, no automation and no anonymity has a seemingly more difficult journey, but with Plato Partnership behind it, the importance of industry relationships may play a key role in getting the new platform off the ground.
LSEG’s RFQ for equities service doesn’t necessarily come up against the same concerns in terms of system specifications as Tradeweb and Plato, but the question remains as to whether the industry really needs the protocol. Finally, Instinet has come up with a rather innovative solution to the problems it found when preparing for MiFID II. Regardless of industry adoption, the agency broker has used RFQ to solve those problems in order to carry on serving clients in much the same way as it did before MiFID II came into force.
One scenario, laid out by Ardevora’s Bond, could see wide-spread adoption of RFQ in equities trading. As MiFID II ushered in the age of periodic auction systems, systematic internalisers and block trading venues, regulatory review and consequential reform in the near future could prove to be a game changer for Tradeweb, Plato Partnership and the LSEG. But those lingering concerns laid bare by the buy-side will have to be addressed if those venues want to stay in the game.
“MiFID II closed down broker crossing networks, which the buy- and sell-side found very useful, and some of that business has moved into periodic auction systems,” Bond says. “Those systems are now under scrutiny, and if action is taken, I think that flow could move towards the RFQ model.
“The big winners will probably be those that offer automated trading with central clearing rather than the large negotiated RFQ venues. Large negotiated RFQs that don’t have central clearing and are not automated won’t be able to pick up that flow which may move from periodic auctions in the near future.”