Over the last year, the buy-side has seen a notable uptick in algo usage and looking into The TRADE’s most recent survey findings, this looks set to continue.
The buy-side has long voiced simplification and automation as key focus areas within this environment, and taking the top spots this year in terms of what the buy-side most desires when it comes to algorithms were ease of use (5.97) and customer support (6.03).
Antish Manna and Kendell James
With this in mind, there appear to be diverse tactics being employed in order to achieve the best model. The TRADE’s research demonstrated that alongside a general increase in algo usage, there has in some instances been a marked uptick in the actual number of algo providers being used by buy-side traders. However, a notable subsect do appear to be remaining ‘loyal’ to just one.
While almost half (45%) of buy-side traders reported using five or more providers, 29% of traders reportedly only use one provider, demonstrating a notable split between those who display provider loyalty and those of the school of thought that diversification is best.
The number of long-only managers looking to have exposure to five or more providers continues to rise however – increasing since 2023 – indicating a potential for single provider loyalty to fall.
Speaking to The TRADE at the TradeTech Europe conference, Kendell James, multi-asset trader at Federated Hermes, explained that with the need to mitigate risk being a primary consideration for traders, these findings are not very unexpected.
“Given developments, upgrades and occasional lapses and latency, having ample options helps preserve firm capital and minimises sunk costs in the presence of these issues.”
However, as pertains to variability, he also highlighted that this is somewhat surprising due to the similarity across providers’ algo suits.
Despite this, “every incremental increase in liquidity exposure you can get, counts when trying to achieve best execution,” concluded James.
Read more: Federated Hermes’ Kendell James on optimising algo trading
Interestingly, once The TRADE’s research team applied the veneer of AUM, it became clear that among the larger asset managers, higher numbers of providers were less the case.
Looking specifically at asset managers with AUMs of between $1 and $10bn, a slight decrease from 3.88 in 2023 to 3.04 in 2024 was recorded. Additionally, large, long-only managers, with AUMs of more than $50bn also saw a slight decline. Whereas in 2023, these firms reported using an average of 4.99 providers this dropped to 4.77 this year.
This is the largest decline seen in recent times, likely down to consolidation leading to fewer distinct providers in the market, as well as increased in-house development.
Antish Manna, head of execution analytics, multi-asset, at Man Group, recently sat down with The TRADE to discuss exactly this phenomenon, explaining that his firm does both proprietary algo development and execution analytics in-house for some key reasons.
“With regard to algo development, firstly we think the information we have on our alphas and strategies gives us an edge when we’re planning the optimal path and approach to execution. Another key factor here is that we can control the pace and focus of development, and finally, being actively involved in the process and deep into the details has improved our team’s DNA and understanding of markets.
“The same applies for execution analytics – we’re aiming for a best-in-class platform, so every component needs to shine.”
Elsewhere, The TRADE’s algo survey unpacked exactly which algorithms were favoured by the respondents. Taking the top spot, perhaps unsurprisingly, was VWAP highlighted by 79% of respondents. Closely following VWAP was dark liquidity seeking at 77%.
These findings further highlight how client demand has driven innovation in algo trading, leading to more traditional strategies, such as VWAP, to begin incorporating predictive techniques such as machine learning to remain relevant.
Delving into this, Manna explained that when it comes to which metrics his firm uses to monitor algos and broker selection on algo wheels, machine learning – specifically reinforcement learning – indeed comes out on top.
“We think there are multiple benefits to this model: it is a systematic process and devoid of human bias; it offers a statistical framework to balance exploration and exploitation (because panels allocation doesn’t get stale); it’s complementary to experimentation, and finally, it offers a clear incentive for brokers to improve,” he explained.
Current and previous surveys can be accessed here.