As the use of trading algorithms has become endemic in equity markets as during the past decade, correctly assessing when and how to use them, given the potential they have to improve execution, is crucial, delegates at TradeTech in Paris were told.
“Ten years ago use of algos was sub 20% of total equity trading for a firm, now there are plenty of people in this room for whom it’s north of 70%,” said Neil Bond, trader and equities market specialist at Newton Investment Management.
Fellow panellist, Lucienne Lao, trader at Boussard & Gavaudan Asset Management, noted her firm’s use was above 80% and Oskar Wantola, head of listed execution technology at Man Group, concurred noting the access to venues and ability to process huge amounts of data that they provide.
When employing these tools, trader need to be conscious of the context tin the market and the limits or abilities of an algorithm pre-trade.
“You have to ask a few basic questions starting with the type of flows,” said Lao. “So I would say first, start with the type of flows you’re getting from your your path. So I know if it is a quant flow, it’s a reduction or liquidity seeking type of flows. And then you can try to choose between features such as scheduled algos, or liquidity seeking algos. Then you need to assess the urgency of the order and the market conditions that you’re currently facing, and you can choose from there, would be more appropriate for your flows.”
There are no hard and fast rules that travel across geographies warned Bond, which also created a need for vigilance.
“I switch between European and US trading and I can tell you that the market structures in both places are very different,” he said. “If you’re using one provider in both regions, the same algo might work in very different fashions so you have to be very careful, particularly at the close, because that is obviously a peak of volume in the day. The closing strategies in the States are completely different from the UK.”
Although many algorithmic types are standard, there are risks when relying on two ostensibly similar algos delivered by third parties noted Wantola.
“Even though brokers offer the same algo by name, they might not necessarily be the same algos beneath,” he said. “Brokers usually have different tech infrastructure, different teams building regional algos and then there are those microstructure differences in different regions as Neil mentioned.”
Lao added, “I would say on less liquid markets like emerging or small and midcap, you do have to look at the spreads and the environment before using any algos. Pven when you have a simple percentage of volume (POV) algo it could be counterproductive to use, and maybe you should concentrate on a ‘target the close’ but still then being careful on the spreads towards the close. I would say using providers and the brokers research is key on that to be aware of what your equity is doing when the spreads are wide and when the volatility is high.”
©Markets Media Europe 2023
