TT: What do you think is behind the rise of the independent brokers in algo trading?
Mike Seigne: I can only really talk from our perspective, as I am sure other independents will have different views. For us, there are three main drivers. Firstly, the lack of conflicts as “agency only” has given us a discrete role to most of our competitors. We approach challenges such as liquidity reaggregation from a different starting place. For Redburn, all venues have one thing in common: they are not ours. As such we are incentivised to find the best way to extract or supply liquidity to that ecosystem without having to deal with issues such as “internalisation”. When you’re acting with principal at play all the competing forces skew your incentives towards maximising crossing “against the house”.
The second point is really that because of this “agency only” position we only have one source of revenue: client commission. The result is a heightened focus on being performant in terms of our clients’ experience and results. If we don’t perform, our clients have no reason to come back. It really is that simple. Lastly, our size and the fact we are only an agent offers a great deal of agility to make changes quickly. It also gives us the ability to offer different levels of service to our clients in a way that most of our competitors find challenging – the level of trust a client has in their broker tends to drop as the broker’s size and complexity scales.
TT: How does a smaller firm compete in terms of technology?
MS: We compete very strongly. You can see that reflected in some of The TRADE’s client survey results, but it is sometimes hard for people to get their head around the principle. Redburn gaining the top score in the “order routing logic” category is, in part, a reflection of our technology platform. Having said that Redburn does not set out to compete on technology, we pick the battles we think we can win, so we focus on outcomes in terms of both execution performance and service. The way we approach our technology stack is, wherever possible, to build our IP on top of specialist partners rather than build everything in-house. Working closely with these providers means we have the same level of control but much more flexibility in terms of the technology we use and how we deploy it. It also means we have access to the latest technology and can leverage a range of providers to develop best of breed without the challenge of maintenance and dealing with legacy code. We think about it as you might think about leasing a new car rather than buying one. We want the ability to roll to the next model every few years. If today we have level 2 automation, but we know the solutions to levels 3 through 5 (fully autonomous) are only a matter of time, leasing becomes more attractive.
TT: Does that lead to a fundamental difference in how you develop an algorithm?
MS: Yes, it does, and we focus on functionality first and foremost. You can build “the best” algo in the world but if it doesn’t do what the trader wants it to do, he or she won’t use it. We compete strongly on Algo Wheels but our core Algorithms are designed to work with traders to improve their results, not to replace them. Our clients have experienced traders who look to add value, so we see one of our roles is to build and deliver the tools that allow them to do this, as opposed to an Algo that tries to take over. To do otherwise would be to waste all that experience.
Our execution analytics team are involved in all our developments, and by leveraging their capability it allows us to be “braver” when addressing a liquidity source, thereby enabling us to use multiple sources simultaneously. Using this wider range, whilst trying to use less of the arbitrary controls that restrict the ability to capture liquidity is critical. Features such as optimal resting times, dynamic MAQs and heat maps are important but the bottom line is you will not get liquidity from a venue if you are not present.
As our clients know, being an independent firm means we can access more liquidity than most other firms. From this we derive a greater understanding of how those venues work. This means we can provide a more nuanced and controlled approach compared to firms who are more focused on either internal venues or technology in isolation.
A good example is that Redburn has consistently had #1 share of value traded using LSE LIS Hidden mid-point pegged orders. We have found that developing ways of interacting with liquidity sources others avoid can yield outstanding results. All these incremental improvements quickly add up, and we are grateful our clients recognise this, as we mentioned earlier when talking about the top spots Redburn received in your survey.
TT: So, what is more important, the technology or the people?
MS: I smile because it makes me think of a piece of work our analytics team wrote called Superficial Intelligence, where we talk about the trade off and the interface of people and technology. In answer to your question, we think that to offer a truly differentiated service you need both the people and the technology to be at their best.
There is a trend in the industry to reduce, juniorise and automate coverage and algo trading is normally the place to start. Given it’s already highly automated it’s tempting to reduce the number of people managing the flow.
This is something Redburn tries to avoid. Our algo coverage team work closely with our clients so together we can both understand what is driving the data. This helps us to contextualise and understand intuitively what is happening, trying to limit the risk of being caught in the weeds without recognising what is happening at a higher level. This is critical and links back to what we were discussing earlier about focusing on the outcomes at the parent order level. If your work at the micro level does not lead to a parent order performance improvement then it becomes an academic exercise at best, or a means of justification for things like the greater usage of lower fee liquidity sources or internalisation at worse.
TT: Do you see this trend continuing in the future?
MS: Yes and no. It seems clear the trend towards more analytical monitoring of algos and the use of algo wheels and automation will continue and we think this works in our favour. Where we think things will change is on the service side. It plays back to your earlier question on technology or people: we see through the lens of technology and people.
We have been selected to receive algo wheel flow from a number of the largest asset managers out there. We think the buy-sides who derive the most value from their brokers are those who engage closely in terms of the benchmarks they set and their expectations in terms of how an algorithm works. Firms such as Redburn can deliver custom solutions that perform well on both the statistical and ‘trader feel’ level.
Where we see the bigger change is the service delivered into the client to help their larger “live orders”, especially those multiday orders. It is clear for these orders that liquidity trumps pretty much everything, so the starting point for almost all is some sort of liquidity re-aggregation algo tilted towards “conditional venues”. Clearly everyone would like to be able to create the nirvana of access to infinite liquidity at price points and times our clients want, but sadly that is not possible. So, at this point two things come into play: the broader reach independent brokers have to re-aggregate the rest of the venues; and the trust to engage the “high touch” components of agency brokers to help both with navigating markets over many days and at times sourcing liquidity that only makes it way to the market via the skill and knowledge of the human.