The increasing power of classical computers has changed modern finance. From high
frequency trading to machine learning, new ways of dealing with large amounts of data are
constantly being developed to develop an “edge” over the competition.
Most of these algorithms are not particularly new but can provide an outsized competitive
advantage for firms that can properly apply the ever-increasing computational power that is
being made available.
Recent headlines and reports would indicate a growing interest from major firms in the emerging
technology of quantum computers. Firms such as Goldman Sachs and JP Morgan have gone
so far as to hire their own quantum teams to stay up to date with the underlying principles and
technology so that they may be able to have an advantage over the competition as
advancements are made. The big tech players like IBM, Google, Microsoft, and Amazon have
their own programs and are providing access to interested researchers from many domains.
In the Quantum Computing community this is all very exciting as researchers from ambitious
startups and multi-billion dollar companies figure out new applications, new algorithms and new
uses for hardware that is becoming more powerful and more reliable every year.
These advancements in hardware could be considered analogous to the development of
computers in the 1950s or 1960s. If the comparison holds, a question that requires some
investigation becomes: What action needs to be taken for companies to be able to quickly take
advantage of any advancements in the discipline but at the same time won’t waste valuable
resources on technologies that in the end will not prove to be useful.
A great deal of attention has been given to the concept of “AI Winter” and there is much to be
learned from the advancements in machine learning and the attempted implementations.
Although today “data graveyards” and half completed machine learning projects are common,
the areas of artificial intelligence and machine learning are some of the most productive and
fastest growing fields. By learning from the past, it should be possible to profit from quantum
computing without going down most of the dead ends seen before.
What do clients need?
Deep Tech as it is known, is predominantly dominated by academics as it begins. Most of the
time, ideas are formed with some theoretical insight that is then confirmed by experimentalists
and then is eventually developed into a usable technology by engineers. This is the path that
classical computers have followed. Abstract mathematical ideas from scientists such as Turing
and von Neumann were eventually developed into usable technologies. After some decades,
the modern microprocessor and its associated industries form a non-trivial percentage of global
GDP. Is it possible that quantum processors will be as widely used in the economy as their
classical counter parts?
The answer to this question is unknown and perhaps from a practical point of view, unimportant.
What happens in labs and mathematical journals is of little concern to financial professionals in
the short term. They have benchmarks to hit and can’t be overly worried about the
technological landscape in five years. Although this is certainly the case, there is always the
risk that some unknown technology will arrive on the scene and disrupt current incumbents.
This is not overly rare in history. In the current environment, managers must be able to buy an
option on the future. How is it possible to benefit from rapid technological advancements without
going “all in”? For the many financial organizations that don’t have the resources of a large
bank, is it possible to compete?
Historically the answer to this question would be “no”. The most prestigious banks recruit from
top technical universities, have effectively infinite amounts of money to throw at research
projects and if they want, sit back and watch developments and advancements until clear
winners emerge.
For medium to small organizations it is indeed possible to compete but it requires discipline and
a rigorous framework to evaluate and implement potential projects.
Organizations that are accountable to stakeholder must ask themselves how quantum
computing can solve problems and how it fits into their current stack. To take advantage of a
yet untested technological innovation, great care must be taken to include various departments
and levels so that the processes can be efficiently scaled to the entire organization.
To be effective, quantum computing must be able to scale in the organization. This means a
deep understanding of the current infrastructure and ongoing training of employees that will be
able to quickly understand and implement the latest advances. There must be processes,
workflows, systems, and controls. An academic project moved into a commercial environment
without these elements is a recipe for failure. Research papers published in prestigious
academic journals will create levels of publicity but without an understanding of how the
technology contributes to the organization, it will not be effective.
To fully participate in this new technology, smaller firms should implement phased projects with
clearly defined smaller goals with accountable internal and external stakeholders. A good
starting place would be for a domain expert to have some basic understanding of the various
platforms, and the state of the industry. This individual could be expected to liaise with outside
technical experts to customize solutions to their specific organization.
As quantum computing continues to increase in computational strength, it’s expected to become
a lot more relevant to financial institutions and there will surely not be one single strategy for
implementation and integration. What is important however is that these strategies are
developed keeping in mind some of the intricacies and novelties of this technology. Although
still in the NISQ era, fault tolerant error corrected systems are expected in just a few years and it
will be interesting to see just how institutions prepare to leverage this technology for the best
possible quantum advantage.