The amount of data that artificial intelligence can handle dwarfs the capacity of the human brain
CC
The next wave of
artificial intelligence (AI) will likely be used by hedge funds making
long-term macro plays around things like oil prices, involving more data
than an analyst could crunch in a lifetime.
Although habitually secretive, the use of machine learning and AI among the hedge fund community has been well publicised.
Bridgewater
Associates, the world's largest hedge fund with about $154 billion
(£109bn) under management, has been vocal about its use of AI. And it's
not uncommon to hear about investment firms hiring data scientists with
PhDs in neural networks, or physicists and astronomers who can remove
the noise from data signals.
Publicis.Sapient AI leader Josh Sutton, who has worked in
financial services for 15 years and has some hedge funds among his
clients, expects to see a confluence of causal analytics and long-term
macro strategies.
Using Bridgewater as an example, Sutton told IBTimes UK:
"If you look at their historic trading strategies, it's been very much
long-term bets around what's happening at a macro level. They have built
their entire business on having some of the best research and analytics
in the industry and some of the smartest minds thinking on that.
"When you combine those two things, I would definitely
expect artificial intelligence to be applied to identify large-scale
trades that might not be evident to an individual researcher.
"Look at how a lot of funds are making large bets around
what is going to happen to the price of oil, for instance, how is that
going to impact different industries, what does that mean from a
portfolio investment?
"Based on this set of five or 10 things that I believe to be
true about what's going to happen over the next five to 10 years, how
does that translate into different bets that I can make across various
industry performance?"
Sutton said he expects machine learning to be overlaid with
more common sense AI technologies on top to mimic the role of an
analyst.
"That's an area where I think you are starting to see a lot
of very private interest, but interest none the less, from a number of
the leading financial firms. How you can deploy AI tech to mimic the
role of analysts at scale. That's to do everything from putting together
valuation models around companies that you would normally hire an
analyst to do, to doing entry-level macro analysis of what's happening
in various industries, based on digesting large amounts of data and
coming up with hypotheses."
A report out today from Citigroup predicts some 30% of
banking jobs could be lost to digitalisation over the next 10 years. So
are analysts' jobs going to be under threat in the future?
"I think they absolutely are going to be," said Sutton, who
has seen variety of firms pushing to automate a lot of the entry level
analyst work that's being performed, and building AI platforms
accordingly.
"And that's going to present a little bit of a quandary because this is going to do two things, I believe.
"Firstly, it's going to dramatically increase the coverage
that companies can have, where historically they have been constrained
by the number of analysts that they pay and the amount of research that
an analyst can perform during a normal work week. That constraint goes
away as you look at leveraging AI platforms.
"But a separate, longer-term challenge is that the analyst role has
typically been a grooming role for talent. It enables people to come in,
learn the business and progress through the organisation to things that
require a greater degree of insider trading and portfolio management
capability.
"So you are taking away some of that training ground. It's a
little bit of a Catch 22, in that companies are having greater
capability to cover more and generate more insights from wider coverage
area. But at the same time will suffer a little bit from a lack of
grooming for future talent."
Sutton agreed we will see an evolutionary shift in the role
of the human workforce. In terms of how this relates to financial
services, he said it would ultimately be a good thing.
"I think we are going to be deploying our talents and intellect at more challenging and intriguing problems."
This data analytics headspace includes things like using AI
to analyse the data signals of individuals that work for various
companies – everything from social media to shopping habits – to
generate interesting insights about how those companies are performing.
"We are working with a couple of hedge funds on this area
right now; using AI platforms to read all of their various social media
and public tweets and other types of information that they make
available. To actually get a true sentiment analysis track of the
company.
He said regulators in Europe have been considering such methods with a view to how AI might be used to self-police companies.
"Being able to read and understand lots of data about people
that work at different organisations can help identify indicators for
potential illegal or at least borderline activities."
Post a Comment