[ T H O U G H T
The evolution of the execution
technology stack
The result of these “macro” trends
yielded better efficiencies in the
execution value chain, while im-
proving overall execution quality
and thus saving millions of dollars
for investors. By some estimates
investors have witnessed a 70-bps
cost reduction in execution impact
cost as a result of the more efficient
execution value chain, even with-
out taking reduced commission
rates into account.
However, from a purely oper-
ational cost perspective, these
efficiencies favor scale (i.e. large in-
vestment banks and brokers) as le-
veraging automation and high-tech
trading for competitive advantage
requires significant technology in-
vestments. The cost for developing
custom, AI-based execution algos
or an algo wheel is fixed, unrelated
to the trading volume (and thus
commission revenues) it gener-
ates. We see some evidence for the
impact of these economies of scale
in recent mergers of smaller Tier 2
& 3 execution brokers globally.
To understand the costs for
developing an automated exe-
cution value chain, one needs to
examine the components involved
in creating it. Beyond the invest-
ment in developing an “in-house
secret sauce” for the algo engine, a
firm must also build a connectiv-
ity framework which enables its
customers to connect to the algo
and the algo to connect to target
market(s). In addition, a moni-
toring framework is required to
manage risk and compliance, and
to provide IT and low touch sales
traders the ability to manage and
interact with the algos.
However, while the algo “brain”
can be used as a differentiating
factor for the firm, it is not as clear
that the connectivity nor mon-
L E A D E R S H I P
|
S P O N S O R E D
itoring framework are; yet they
carry significant development and
maintenance costs. Until recently,
most large firms rarely considered
off-the-shelf software as a solution
or as “building blocks” to acceler-
ate and reduce internal develop-
ment costs. But, as we are entering
the second phase of the automation
revolution, this is rapidly changing.
Firms are adopting a more critical
view on what should be built
in-house vs. tools that can be ac-
quired, and do so for good reasons.
Build? Buy? Both?
Examining why firms have opted to
“Build”, i.e. developing proprietary
full-stack algo trading technology
rather than “Buy” the frameworks
needed around the algos from soft-
ware vendors, we have identified
three main factors:
1) Availability - the fast pace
of change in early years and the
novelty of the automated execu-
tion desk meant that there were
no built-for-purpose, off-the-shelf
solutions available in the market.
2) Customisation - off-the-shelf
solutions were not flexible or
adaptable enough to easily on-
board customers or manage chang-
ing exchange interfaces.
3) Performance - vendor products
were lacking the throughput,
capacity and latency required by an
automated execution value chain.
These shortcomings are ex-
plainable, as most vendors have
designed their offerings with
serving the high touch execution
value chain in mind. Such solutions
were then augmented to address
low touch monitoring as well, but
implementations were typically
awkward.
However, over the past three
years, the automated execution
value chain has matured, with a
new generation vendor solutions
B Y
I T I V I T I ]
emerging. Itiviti, for example, has
developed a connectivity and mon-
itoring framework, designed from
the ground for low touch environ-
ments. With ample flexibility in the
connectivity framework, onboard-
ing has become much simpler and
more configurable.
This framework is FIX-based but
supports multiple protocols and is
built to translate and enrich order
flows. The monitoring framework is
geared towards an “exception man-
agement” workflow, built to support
automated execution systems with
hundreds of thousands of orders
and millions of executions.
Conclusion
With growing market complexity,
increased competition and cost
pressure, firms are looking to
use their in-house development
resources more efficiently. It
no longer makes sense to build,
maintain and manage a full execu-
tion technology stack unless this
enables the firm to add clear value,
draw competitive advantage, or
leverage economies of scale.
Given the options at hand today,
a more cost effective and busi-
ness-focused solution is to focus
development on areas where the
firm can add value to customers,
e.g. through intellectual property,
and to source components where it
does not.
There is a new generation of
built-for-purpose tools which offer
the flexibility, capacity, latency
and throughput for supporting an
automated execution value chain.
This is a key reason why we are
witnessing a rising demand for
connectivity and low touch solu-
tions from firms across different
sizes. We expect this trend to con-
tinue over the coming years as the
second phase of automation of the
execution value chain matures.
Issue 59 // TheTradeNews.com // 21