Trading’s Formula 1 Problem
According to a detailed estimate by Quinlan & Associates, trading inefficiencies cost asset managers 1.2% to 2.7% of annual return, equivalent to losses of up to $18 billion for the largest players.
There are three main factors that negatively influence execution performance, according to David Rogers, one of the report authors and a speaker on a panel on Enhancing the Value Proposition of the Buy-side Trading Desk during FIX’s 19th Asia Pacific Trading Summit, held 2nd September.
“The first is a lack of technological proficiency; while most desks use of algos, real-time analytics and so on, they’re not necessarily utilizing them to the full,” explained Rogers. “The second is a poor choice of execution method. The third, and the one the creates the biggest opportunity cost, is a lack of true partnership and collaboration with both internal and external stakeholders.”
The need for collaboration was echoed by Will Psomadelis, Head of Trading Australia; Global Head of Electronic Strategy Research at Schroders Investment Management. “As a buy-side trading community, we can’t just see ourselves as executors of orders. It is deeper than that. As a trading team, we own that liquidity space and can inject meaningful inferences into the portfolio process to improve returns. When viewed through this lens, the value proposition goes upstream.”
“Culturally, we needed to bring people across the line. We can’t just go and put an overlay on and tell people what to do,” added Psomadelis. “That might sound nice and easy in theory, but the point is everyone has different competitive advantages. Not everyone’s an algo expert, not everyone’s a relationship person, not everyone has the instinct to figure out what the right time to do a block is. So we re-designed the entire process to ensure every decision is supported by data. Once we laid the foundation of centralized data and built the appropriate models to solve the problems we face, we’re able to find people’s competitive advantages. Every order is appended with guidance with respect to how to trade it, all generated from internal models, but ultimately the trader is free to use their judgment and intuition. This job is never complete. It’s always evolving.”
Improved communication between the buy-side and sell-side will be critical to realizing marginal and yet significant improvements in trading efficiency, observed Raj Mathur, Managing Director, Co-Head Advanced Execution Services (AES), APAC at Credit Suisse. “The overall slippage in performance that effectively ends up accumulating on your orders is mainly from the market movements rather than trading. Where the market movement is significant, the difference, the algo impact on alpha is marginal. That said, it’s the Formula 1 problem: pole position to fifth position is a quarter of a second difference, but we’re all in this to make that quarter of a second difference to help our clients to improve performance. We’ve still got a lot of research to do, but whatever clarity we can get from the buy-side in terms of giving us more detail, that’s obviously invaluable.”
As Mathur pointed out, buy-side trading desks are now getting a lot more useful context about orders. “If you look at what the buy-side is doing now – and they’ve developed a long way in the last couple of years. They’re getting a handle on some of the mechanics around why the order has been placed and what’s the inception information around it. However, on the sell-side orders, we don’t have a clear line of sight on the investment decisions from the buy-side portfolio managers. So, a lot of what we end up doing is trying to reverse engineer some of these things to try to work out if there is a pattern and if there’s something we can optimize to help to achieve their objectives.”
The speakers also noted the need for ‘triangulation’ to achieve optimal outcomes. “One of the things that I really came to learn as a desk head on the buy side is that you don’t and shouldn’t sit in an operational vacuum,” said Rogers. “The triangulation of trading, investment and sales to deliver singular client outcomes is where you should be at as a desk, as opposed to being a sort of operational adjunct to the broader investment organization. When truly scrutinized, most desks fall short on this front.”
And the single most effective way to achieve that triangulation seems to be fostering dialogue. “One of the things we’ve tried to do internally is, rather than prescribing to the sell-side, we’ve sought to have a relationship with them and encourage them to look at the problem that we’re trying to solve,” revealed Psomadelis. “We do that by providing them some color about what we’re trying to achieve around particular orders. We are fundamentally against ranking algorithms in aggregate, and given QTRMASTER [Schroders’ Quantitative Trading Research Model for Automated Strategy & Timing Execution Routing] forecasts routes on an order by order basis, we provide the brokers with simulations to guide them on how we see their strategies interacting with our flow, allowing them to make specific changes to the behavior of the underlying models.
Another major cause of poor execution performance is under-utilization of technology. “For example, in the fixed income space, while electronic trading (algo) usage is rising, 60% of respondents in a recent survey by Barclays admit they didn’t even use an EMS to enhance automation, analysis and straight-through workflow processes,” noted Rogers. “And in equities, the current Greenwich Associates survey on Asia equity trends showed that less than 10% of firms used algo wheels, a systematized method of broker allocation that has been proven to improve aggregate execution performance. There are of course many desks that do use this and other methods, but across the industry we observe a noticeably slow uptake.”
Sticking with the F1 analogy, Rogers likened it to “getting off the line and staying in second gear” and issued a call to arms: “If you’ve got a really good EMS, you need to ask yourself whether you are really using it well. Are you categorizing your flow correctly, for example? Are you assessing brokers to the fullest extent possible (free of biases)? Are you using data optimally?” Moreover, Rogers added, “Where trading desks on the buy side really need to get to is setting up their own specific commercial value proposition. Then be directly involved with sales teams and the client themselves, where possible: trading desks don’t compete with other asset managers, in the way the funds themselves do, but by demonstrating how trading improves execution performance, the client conversation becomes more enriching, and partnership-based.”