THETRADETECH DA I LY
highlights
THE OFFICIAL NEWSPAPER OF TRADETECH 2019
THETRADETECH DAILY
highlights
THE OFFICIAL NEWSPAPER OF TRADETECH 2019
• The way the market has evolved with
asset managers now having to trade so many
bonds, you can’t have the expertise that you
had 20 years ago. Firms must reconstitute the
experience that the trader used to have; what
has traded, what was the liquidity and what
was the market impact.
ALL STAR PANEL: Evaluating dark liquidity –
what are the similarities and differences and
proven strategies for interacting with them?
• Despite major industry concerns pre-im-
plementation, the unintended consequences
of MiFID II have been positive. We’ve seen
positive innovation driven by commercial
need, but ultimately the objective to move
market liquidity to lit venues has forced the
industry to come up with new and innovative
trading venues, such as periodic auctions.
• The industry is still focused on navigating
the UK’s impending departure from the Euro-
pean Union. While some panellists agreed the
long-term impact of Brexit could be positive
in terms of increasing competition, others
said that the trick will be to find sufficient
harmonisation with some form of equiva-
lence to preserve the UK’s right outside of the
EU to be more creative.
DISCUSSION PANEL: AI and automation in
trading – how to approach it in a sensible and
cost-effective way?
• The combination of human hypothesis
and ideas with the power of machines to
store, process and analyse data is the most
successful approach to adopting artificial
intelligence and machine learning systems.
• Automation will lead to more automa-
tion; while there are many manual processes
that have been automated for a number of
years already, there are further areas where
automation can benefit capital markets
participants, such as communications and
compliance functions.
• However, compliance does present a dif-
ficult issue for AI-based activities and firms
should be careful to document all steps they
undertake and all data processes.
• The transcription and recording of voice
communications is one specific area where
AI technology can bring tangible benefits. Al-
though significant advances have been made
in this space, the technology is not perfect
and there is still much work to do before it
can be adopted on a more widespread basis.
DISCUSSION PANEL: How to achieve scal-
ability to enhance trade performance with
future-proofed technology investments
• The availability of resources for technol-
ogy investments can be a difficult issue to
navigate and business cases must be airtight
to ensure allocation.
• Recruiting talent has become a far harder
prospect for buy-side firms, particularly
22
THETRADETECH DAILY
technology talent, due to a change in culture
and the emergence of large tech firms such as
Google and Amazon.
• Technology is a necessity and demands
investment, but should not be an end goal
in and of itself. Many firms are not correctly
structured to leverage significant value from
these investments.
• Agile development is not always beneficial
for capital markets firms and, in some cases,
longer-term strategies for large-scale projects
are far more suitable.
the line.
• Although the sell-side has attempted to
provide a single, holistic service desk for
buy-side firms with multi-asset desks, this
approach hasn’t been hugely successful, as
traders will always need specialists on the
sell-side with deep, asset-specific knowledge.
As the buy-side, what are your biggest changes in demands from your broker and their
services post-MiFID II?
Sample size: 124
More customised algos
Better quality execution data
Increased demonstration of compliance and governance
DAY TWO
KEYNOTE INTERVIEW: The future of trading
– how leveraging AI tactics and new data sets
can improve trading and investment perfor-
mance
FIRE SIDE CHAT: Critical success factors for
• Machine learning projects can be applied
implementing a multi-asset trading desk
to a range of trading processes; MAN Group
– overcoming skill set, infrastructure and
has developed a framework for broker alloca-
operational challenges
tion and order flow that went live last year.
• Addressing the challenges that are inher-
• Machine learning is primarily aimed at
ent with a multi-asset class trading
removing human bias from deci-
desk are unique to the business
sion-making and implementing
and different approaches will
a system that goes beyond
be necessary depending on
prediction models to a sys-
what the firm is attempt-
tem that can take actions
Assuming current cost,
ing to achieve by trading
based on predictions and
do you believe AI can
different products.
then learn from the re-
deliver cost benefits to
As such, context is
sults of those decisions.
running a fund?
extremely important for
• Building a machine
each individual trade.
learning system is all
• The complexity for
about the team; while
Yes 55% No 45%
each trade will also differ
it is a difficult and long
and a like-for-like compar-
task to recruit the right
Sample size: 104
ison will yield little insight
expertise, having this founda-
or allow firms to improve their
tion in place enables to firms to
processes going forward. Data plays
pivot technology systems across the
a crucial role here and the approach to this
business and towards specific challenges or
will most likely follow the equities path, if
objectives.
not in the actual application further down
• Patience is a vital element for successful
Q
Q
Proof of embracing and investing actively in new tech
True transparency on order routing and venue analytics
0%
25
machine learning projects. It cannot be done
in a short time frame and requires significant
buy-in from the leaders of the company. It is
not the same type of journey that most firms
undertake for other technology projects.
ALL STAR PANEL: Taking AI and predictive
data analytics to the next level – how to access
liquidity systematically whilst navigating
complex markets to gain a competitive edge in
a global marketplace
• AI is limited by the human that is coding
it. Tools so far have been very good at describ-
ing what is happening, but not predicting.
When you go to what will happen tomorrow,
that’s when most of these processes fail.
• On predictive analytics: Axa’s global trad-
ing head said if you’re in trading you have no
choice but to invest into a serious amount of
data analytics – advanced data analytics; you
have to somehow be able to turn something
into you understand.
50
75
100
• All panellists agreed that the development
of innovative trading venues under MiFID
II, specifically periodic auctions, has been
a positive development for the industry.
Periodic auctions are just at the start of their
evolution, with the promise that other major
players could enter the space.
KEYNOTE INTERVIEW: Which new disruptive
technologies are likely to transform the way
the equity trading ecosystem operates and
how can you structure your trading division to
benefit?
• Financial services organisations need to
keep up with the pace of life of the consumer
in their everyday life. What people are expe-
riencing in their everyday lives are instant
decisions and transfers – take that into the
world of asset management where a big fund
says I need to look at where my holdings are
– it used to be ‘okay, you’ll get it through the
post in a couple of weeks and then we’ll have
a meeting about it’.
• The industry needs to keep evolving.
SSGA has set up an AI research centre in
China, investment centre in Poland to look
at disruptive processes and AI, and change
the orthodox of processes in terms of doing
things differently.
• Aggregate of AuM will continue to grow,
it’s just a matter of whether we can be re-
silient to price pressure on the managing of
assets, it’s the competition of managing those
assets. The threat to SSGA is if they’re not
relevant in technology, and that goes back to
the Charles River Development acquisition.
GLOBAL ECONOMY GUEST KEYNOTE: What the
future holds - Gerard Lyons predictions on the
future state of the global economy and how to
stay ahead
• World economy in size: beginning of the
century $32 trillion, day of the financial crisis
$62 trillion, end of last year $86 trillion.
• Three different ‘S words’ sum up the glob-
al economy: Scale, sequence and shocks.
• As we move through this year, the world
economy will stablise. Policy plays a big role,
with central banks putting their tightening
on hold, this is giving another prop to the
global economy from deteriorating further.
Debt levels are high - now moving into the
early stages of an early economic and political
cycle.
• China is slowing but its scale is now huge.
They are committed to reduce pollution in
China, while there is a concern debt level is
rising.
• We are about to have a fourth industrial
revolution through artificial intelligence. This
will lead to more goods becoming available at
a lower price, which creates more jobs.
ALL STAR PANEL: How are the buy- and
sell-side leveraging technology innovation to
enhance performance and generate alpha from
research and trading?
• The introduction of research and execu-
tion fee unbundling as a result of MiFID II
has thrown up opportunities for technology
innovation to improve research production
and consumption. Natural Language Process-
ing (NLP) and mining alternative data sets
are potentially game changing technologies
here.
• The proliferation of data sources presents
a challenge for firms seeking to embrace
innovation. Intelligent tagging presents one
solution to vast quantities of unstructured
data sets.
• Ubiquitous data can only provide so much
value to trading firms and will more likely
lead to beta, rather than alpha. Proprietary
data holds far more potential to achieve
alpha, but larger firms with bigger resources
hold a significant advantage in this respect.
• The industry has now embraced collabo-
rative projects and strategic partnerships to
achieve innovation, with these efforts now
Issue 2
TheTradeNews.com
23