At Tradetech 2024, conversation revolved around how to determine which algo to utilise given the range of investment strategies and asset classes.
Aviva Investors global trading analytics manager Ash Sharma said the firm utilises different TCA vendors for each asset class. “I think traditionally when vendors come into the market, they focus on one or two of the asset classes. But now we’re seeing a lot of the vendors branching out into other asset classes converging. Down the line, I think the optimum scenario is to have one vendor that covers all the asset classes we trade.”
Phil Lemmon, head of EMEA Sales, ISS LiquidMetrix, said clients just want an answer to the question, What is the appropriate algo? “There is really no straightforward answer. It depends on market conditions.”
Lemmon likens an algo to a black box. “To a buy side trader, they have a high level of understanding of how it works and how it interacts with liquidity in the market. But the most important thing to understand is how it behaves, each algo behaves under different market conditions. You need to know which algo is going to yield the best results. Only then can you tie it in choosing the right algo and the right strategy to your investment objectives. That’s really the question we are trying to answer for our clients.”
BMLL’s chief product officer Elliot Banks thinks algo selection rests on getting high quality data. Determining which algo is best “all starts with good input from a market wide perspective.”
Questions around liquidity, context, which algo, if you have good quality data, that allows you to build out from there, Banks said. “It’s not TCA or algo execution or algo monitoring for the sake of it, it is actually value add, it is adding alpha into your whole process. That is where a lot of firms are taking it, moving away from a purely box-ticking exercise.”
Data insights
LSEG handles 10 billion rows, including trades and quotes, of data every day. Every year, the figure stands at 6 trillion rows of data. “That amounts to around 50 petabytes of data that is stored in our database,” LSEG’s Harish Komalam Gopalakrishnan, director, tick history product and PRS solutions, said.
This data is fragmented and sourced from different exchanges, and consolidating and normalising it is key, as well as having a platform where it is accessible and available.
“Think about the amount of storage that you need to process this data. People are spending millions every year on an average to large scale mine; they spend at least 10 million on storing and processing the data. We are removing that burden of processing and storing the data, making it available in a shared storage, like in AWS, as well as in GCP. So, that is actually making a difference in the space.”
Given all that, how do you convert this hosepipe of data into better outcomes? Each line requires context and colour.
Joe Collery, head of trading at Comgest, thinks the responsibility lies with the firm and its team to add that context around the order. “You have to have the facilities to add colour around the order.”
“This trader, actually, unbeknownst to the trader, is quite passive, and is that effective? They may not have noticed his unconscious bias and then another trader could be more aggressive, is that effective?,” Colliery added.
Behavioural analytics
The potential of behavioural analytics to understand market behaviour and decision-making processes is, to a certain degree, untapped.
In conversation, both Northern Trust’s senior data analyst Victoria Bryan and PGIM Quantitative Solutions’s chief investment officer George Patterson recommended adopting tools that can better analyse behavioural data and provide more actionable feedback to traders, without requiring an expert to interpret. This could be done by TCA providers or new vendors, Patterson noted.
Bryan highlighted the importance of understanding market decision-making and the impact of factors like sleep and meals on performance. Interestingly, it was looking outside of financial services that could provide the most insight – something that may be spurred by the influx of younger talent.
Both Bryan and Patterson anticipate younger generations entering capital markets to increasingly demand the use of behavioural data and analytics in investment processes, given its use in almost every other technological domain that touches data in some way or another.
“There are other industries that do this a lot [behavioural analytics] and they do it very, very well. I do feel like the financial industry is a little bit behind the curve when it comes to actual behavioural analytics because there is that nervousness around adoption. Ultimately, if you understand the market, understand the decision making, you can improve,” Bryan said.
©Markets Media Europe 2024
