Best Execution speaks to Adam Wood, CEO, Turquoise Global, and Head of Equities Trading Commercial Proposition, LSEG, and Dominic Lowres, Head of Execution Strategy for Liberum, about the exponential rise in retail trading, how new platforms are improving best execution, and whether the RSP model is still fit for purpose.
Liberum’s T-REX was originally launched in 2018 as a new trading algorithm to execute retail flow using institutional trading technology, specifically through post-MiFID II periodic auctions allied with dark pool aggregation. T-REX now composes a suite of dark mid, dark lit and block algorithms. The platform offers a faster way of trading than the market making retail service providers (RSP), and always achieves mid-point or better (when using a pure dark version).
In April 2023, Turquoise added to its suite of products with the launch of Turquoise Retail Max™, a new trading service designed to support European retail brokers in meeting their best execution obligations as well as delivering price improvement for end retail investors when trading pan-European securities, through a multilateral auction process. With both services now live, we speak to their respective heads to find out how these new models are being received in the marketplace – and how they can improve the trading experience for the retail user.
What were the drivers for initiating T-REX and Turquoise Retail Max™ and why are they so relevant for proving better best execution?
Lowres: In late 2017 we did a tour of UK retail brokers and we noticed that best execution via the RSP model frequently meant that customers were only interacting with one counterparty, and sometimes that was inside the touch price and sometimes outside. So we decided to develop a product to go live in tandem with MiFID II (April 2018), that was essentially a way of getting institutional quality fills at mid-price, in a simple-to-use form – something that has been available to institutional clients for many years.
Wood: We saw a similar scenario, in the sense that in the UK, executing a retail order outside of the RSP was increasingly difficult. We also looked at how things operated in Europe from a retail perspective and realised that it was very different. But even when Europeans are trading on an exchange, they’re still not being offered a best execution opportunity, it’s typically crossing the spread to be able to execute. What was needed was to offer wholesale services to the retail market. There’s no reason why a retail order should be treated any differently, nor why it shouldn’t have the opportunity to interact with a wider audience. Retail-focused offerings are generally led by just a small number of market participants and what we really wanted to do was open things up, because a retail order should be able to interact with anybody and everybody. And if an institution is prepared to provide a better price than a market maker, then that retail order should benefit from the price improvement the institution is prepared to offer. That was the driver for launching Turquoise Retail Max™ on our side.
What kind of demand are you seeing from the market?
Lowres: We’ve got about 65 active low touch customers signed up. About 25 of those are wealth managers and retail brokers, and the rest are institutions – we unexpectedly found that there’s massive demand from institutions who want to be able to trade at mid-price against flow that originates from the retail world, because they perceive it to be very benign with low information content. If they’re building a position, they can use exchange auction member matching to consummate the retail trade, and everyone wins because everyone’s dealing at mid-price and the institution can build up decent positions, and still have the chance to interact in the block market as well.
Wood: On our side, we launched Turquoise Retail Max™ in April of this year, and it’s been very successful so far. We look at things slightly differently to T-REX, as we don’t typically have retail brokers signing up directly. They come through aggregating brokers to get access to the service, and we’ve started to see a number of those provide access to our service, which entails identifying the retail orders through a specific flag; this in turn allows the retail orders to benefit from both the liquidity providers and the other institutions – although we don’t publish the flag as part of the order message, it remains anonymous.
What kind of trajectory have you seen, and what growth are you expecting?
Lowres: We’ve seen a geometric ‘J’ curve of growth since MiFID II, with volumes up significantly since 2019, so we’ve got really positive momentum. Originally, we set this up for the UK only, but the split is now 30% Europe, 70% UK and we see continental European flows as a very positive catalyst for the product.
Wood: From our side, when we look at the participation of retail in Europe, it ranges from 10-25% of flow on the primary market. If you’ve got €2.1 billion a day addressable for retail, we want to be capturing 20-25% of that. We’ve also seen a huge growth in retail investing into ETPs and ETFs, which is relatively new, and there’s a lot of work being done across different jurisdictions to encourage that, which is a big driver. We think retail could grow upwards quite significantly from around €2.1 billion a day now.
What are the nuts and bolts behind how you offer investors price improvement?
Lowres: To put it very simply, a portion of every T-REX ticket is left resting with exchange auctions that allow member matching (such as Turquoise and Aquis), which are the lit continuous auction venues. That allows us to match between all different types of flow at mid-price, so our T-REX suite is predominantly pegged at mid-price. There’s no retail marker published with a lot of these retail trades, so they’re indistinguishable from the morass of other executions in the market, and the result of mid-price usually means low impact, low pre-trade signalling and low post-trade reversion. Typically most of our investors look out to about one minute post-trade. Dealing at mid-price in a quite benign way gives very good TCA characteristics. We’ve also had two academic studies conducted by Professor Richard Payne of Bayes Business School in 2020 and 2022, the first of which showed that regardless of time, day of the week, buy or sell, or domicile of stock, this product was always mid-price and exhibited very low signalling characteristics, which we believe counts significantly towards better execution. The second study showed that during periods of massive volatility like Covid, T-REX did even better than in a steady market at getting more and larger blocks executed at mid price.
Wood: We’d support that. We made a couple of changes to the way our lit auction order book operates. In particular, we obviously heavily rely on market makers to make the service viable. So Hudson River Trading and Stifel joined us from the start, and we publish their prices in real time. Typically, lit auctions are public at the point that an auction starts, through the publication of an indicative auction quantity and price. What we’ve done is to increase that transparency for retail brokers, in theory, anyone can take that market data feed, but only retail flagged orders can actually interact with those prices. We made no changes to the auction mechanism, so when a retail order comes in, it triggers an auction which generates a notification that gets sent out to all market participants, which can then interact with that order. We display additional market data, but the auction process itself doesn’t change, and if someone is prepared to provide a better price, they can.
How does this style of execution benefit the end user, and who are you expecting to use it?
Lowres: We have a range of wealth managers and retail brokers connected, as well as a range of institutions, hedge funds, long-only funds and outsourced dealers. The main benefit to them is that it’s always mid-price, so most of the time, most participants are achieving a much-improved price. For example, at the moment the bid/offer spread for the FTSE 250 is approaching 40bps: so in that scenario each side gains around 20bps of price improvement. As the academic studies have shown, and the TCA continues to show, there’s very little movement pre- or post-trade, which is more important to institutions who’ve got blocks to trade, but retail also enjoys that lack of signalling.
Wood: I think one of the key things here is the multilateral nature of our order book. So instead of providing buy or sell signals to a group of market makers, the retail order participates directly in the price formation mechanism within that multilateral order book; you’re not triggering any information leakage. You’re minimising reversion after the trade as well.
A second benefit is that we’re also waiving the fees on our real-time market data feed for retail brokers, so they can provide real-time market data for all securities – UK, Swiss, European – to their end customers at no additional cost. We’re really trying to make it easier for the retail broker and the retail trading community to get access to good quality, real time market data across the European spectrum.
What about usability – how do you connect to T-REX and Turquoise Retail Max™?
Lowres: For T-REX, we’re connected to pretty much all the mainstream OMS and EMS. It took a good couple of years to connect them all, but that now allows connection to the whole of the European and UK market.
Wood: We don’t onboard end users directly. Retail brokers can become members of Turquoise and trade directly, and some of the bigger ones are doing so. But what we typically see are retail brokers utilising other brokers to access the markets where they don’t necessarily have the technology or the connectivity to be able to connect directly. We certainly see this when talking to Spanish and Italian retail brokers, who generally connect to their home market and then rely on other brokers to access other marketplaces.
Following a successful launch for Turquoise Retail Max™, and a successful first few years for T-REX, what changes have you made based on client feedback, and how have you adapted or innovated to meet evolving demands?
Lowres: We’ve recently done a collaboration with Infront, which owns NETbuilder. They have a three second version of T-Rex, all or nothing order style, still pegged at mid-price that is connected to their RSP offering, so dark and lit auction venues can be checked automatically prior to the order flowing through to the RSP module. That’s been quite popular. We’ve also seen customers asking us to incorporate the close at 4:30pm in Europe and the UK – because it has so much volume going through it now.
Wood: When we were working with market participants prior to designing the offering, including analysing the quality of execution and the ability to interact within an order book, one thing that became very apparent was that when interacting with a retail order, market participants were keen for that execution to occur, even if the market moves against them during the auction. We therefore introduced a deferred cancellation function, which you can opt into as a liquidity provider, or as a market participant, to ensure auctions don’t fail once triggered. We are also focusing on how we can extend the service to UK and Swiss securities, and how we bring this service to the UK. Based on feedback over the next three, six and 12 months, we’ll look at any further enhancements we need to make on the European side.
What are your plans for the future, and what are the next steps?
Lowres: Straight-through processing has become a major part of T-REX, where retail and other flow automatically uses the algorithm suite. We also envisage a massive opportunity in continental Europe, because the way we’ve calibrated the smart order router means we can transfer all the benefits UK clients have to EU customers that we’ve never seen before. So the two main growth areas for us will be automation, and increased volumes of trading in continental Europe, which ties in very nicely with Turquoise Retail Max’s™ focus on trading EU names. We want to be front and centre of that initiative.
Wood: For us, at this stage, it’s really about providing the data and analytics to show the benefits of being able to trade using a lit auction mechanism. We will be focusing very heavily on price improvement and reversion. The lack of change after the trade and the positivity of the price improvement are areas that T-REX has done a great job of demonstrating. We just want to take that information and share it more publicly, to highlight that if you’re a retail broker and you’re not providing these services to your customers, then they are really missing out. Along with expanding our geographical reach, that’s what we will be concentrating on over the next few months.
What will the key trends for retail be over the coming year?
Wood: There’s obviously a lot of regulatory attention being focused on the retail community. In the UK there have been a number of papers released over the last four or five years on the subject, and we’ve seen something similar in Europe as well. Regional regulators have done some deep diving into how retail orders are handled – while issues like payment for order flow are of course always on the radar. It’s something that neither T-REX nor Turquoise Retail Max™ participate in, because ultimately we don’t believe it leads to a good outcome for retail orders, so we are hopeful it will be addressed from a regulatory perspective sooner rather than later.
Lowres: Over the next 12-24 months we’re going to see a lot of scrutiny on how retail orders are handled, both in the UK and Europe. We are in regular contact with the FCA and run regular webinars for our retail clients so that they can stay abreast of any upcoming changes. The real question is whether the RSP model is truly fit for purpose. Is it the best execution route for retail? Well, T-REX and Turquoise Retail Max™ are here to say that actually, we think there’s a complementary alternative.
©Markets Media Europe 2023
