A DV E RTO R IAL
Will smarter machines make
smarter traders?
Although the use of AI-led systems, such as algo wheels, automation
and analytics solutions, have been in use for some time, the buy-side
trading desk is primed to develop these tools even further.
A
rtificial intelligence (AI) and ma-
chine learning technologies have the
potential to fundamentally improve trader
performance. However, in order to be suc-
cessful, trading desks must adopt a flexible
and data-driven approach, according to a
recent webinar hosted by Refintiv.
Speakers on the Smarter Traders with
Smarter Machines webinar said that
although the use of AI-led systems, such as
algo wheels, automation and analytics solu-
tions, have been in use for some time, the
buy-side trading desk is primed to develop
these tools even further.
“Simple trade automation, the idea of
creating rules to take some of the more
liquid or easier to trade orders off the
books makes sense,” said Ian Mawdsley,
head of buy-side trading EMEA & APAC at
Refinitiv.
“One of the more interesting ways to look
at that would be if there was some way of
setting certain parameters around tracking
stocks or volatility, to be able to look into
the myriad of different trading applications
and bring them all together to form a story.”
The use of AI on trading desks should
ultimately be geared towards providing as
much relevant and condensed informa-
tion as possible to the trader, said Ashwin
Venkatraman, global head of equity trading
automation and execution at JP Morgan
Asset Management, and the best way to do
so is to adopt a data-driven approach as
early as possible.
“There’s a vast amount of data that a
trader is having to take in, to try to decide
what to do with a for a given order, so it’s
about trying to condense that as much as
possible,” commented Venkatraman.
“It’s about aiding the trader with con-
densed information which the model can
14
THETRADETECH DAILY
do behind the scenes to assist with that
decision-making. If you look at the medical
world, people don’t necessarily talk about
automation, but they do talk about AI and
models that are there to assist the clinician.
It’s not different to trading.”
Webinar audience listeners were also giv-
en their chance to provide their opinions on
the use of AI on trading desks. A poll of lis-
teners to gauge the current use of AI as part
of the trading process founds that around
27% were currently using the technology
to optimise trading, while just over-half of
respondents indicated that they were still
considering future implementation.
Optimising trader performance was
highlighted as the area where AI technol-
ogies possess the most value to the trading
desk, according to half of webinar listeners,
while streamlining trader workflows (20%)
and improving alpha generation (16%) were
also noted as key factors.
Michael Broadbent, principal consultant
at Ergo Consulting, added that while AI
can present trading desks with a significant
amount of potential, firms should be aware
of their current technology capabilities and
how that will form the basis of their future
front-office technology stack.
“You are embedding your future with
present technologies, so whatever your
present technology solution is, it has to
have a certain amount of future embed-
ded in it,” Broadbent said. “The point is to
have a rolling view of what might come up
in the industry in the future and to start
investigating how you might be able to
take advantage of that while you use your
current set up.
“This is happening now. It is time for the
buy-side desks to take control of their own
trading and their own technology.”
“There’s a vast
amount of data that
a trader is having
to take in, to try to
decide what to do
with a for a given
order, so it’s about
trying to condense
that as much as
possible.”
ASHWIN VENKATRAMAN,
JP MORGAN ASSET MANAGEMENT