By Kenneth McLeish, Global Head of Equities Trading Technology, J.P. Morgan Asset Management
Financial firms that put technology at the core of their business, use agile development processes and leverage the cloud will gain a competitive edge.
“I hope our feathered messengers will have brought you in due time our good prices…A B in our pigeon dispatches means: buy stock, the news is good; C D…means sell stock, the news is bad”, Nathaniel de Rothschild, Paris, August 1846
Finance firms have long been at the forefront of pioneering new technology to give them an edge. From Baron Rothschild’s innovative use of carrier pigeons in the 1800s to improve the timeliness of information coming from the Battle of Waterloo, to hedge fund managers of the 2000s who could benefit from the millionths of a second differences between sending messages over microwaves rather than via optic fibres.
During the 1990s and 2000s technology investment in finance soared and technology talent flocked to join finance firms. Finance was at the forefront of technological innovation and home to some of the biggest data centres, most sophisticated networks and most complex software ecosystems.
Yet, during the past decade, newer firms such as Google, Facebook and Netflix appear to have overtaken investment firms as the ones pushing the boundaries of innovation. This is a trend that delegates at last December’s Equities Leaders Summit in Miami recognized.
How did finance fall behind and what can be done to bring finance back to the forefront of innovation?
Technical debt is a term frequently used by technologists. Simply put, it is avoidable complexity which, if allowed to build up, will eventually slow things down and become burdensome. Many finance firms were good at building new systems but less efficient at removing dated ones. The result has been large and complex ecosystems composed of tens, or even hundreds, of applications that are expensive to maintain or change, slowing down a number of finance firms quite considerably.
Increased regulatory reporting, cybersecurity and resiliency requirements have had a profound impact. Building more resilient systems has tended to require more than one data centre and the doubling of equipment (which tends to higher quality in nature to reduce the risk of component failure) while having to keep everything in sync. With the ever-growing threat to cybersecurity, maintaining a bigger estate has become an increasingly expensive exercise. Multiply this by a large number of systems and one can begin to see what’s been consuming the technology budgets of many finance firms.
The increasing cost of delivering technology coincided with the financial crisis of 2007-2008. Technology budgets came under pressure, with many firms struggling to do much beyond meeting regulatory requirements. As a result, many firms looked to reduce technology costs by outsourcing or offshoring which, over time, has become harder to manage from a distance. To compensate, management processes have become more rigorous with distinct phases of analysis, build, test, release and support, often done by separate teams in separate locations.
What can be done to increase innovation?
Innovative companies tend to operate in a notably different fashion to more traditional ones whereby how things are developed is as important as what is developed. Upon closer inspection, we’d argue that innovative firms tend to have three themes in common: they see technology as core to their business, they use agile development processes and they use the cloud.
Within finance, business and technology were, for a very long time, typically viewed as distinct functions. They sat in different locations, had different reporting lines and were often poorly represented on operating committees and in boardrooms. Limited comprehension of emerging issues and misaligned goals (typical by-products of such segregation) would often drive both functions even further apart.
But, as was later learned, when technology is integrated into the business, efficiency gains arise and it creates an environment that’s generally more conducive to innovation. Having technologists sit alongside their business users can, for example, lead to these efficiency gains:
If a developer is sitting beside a business user, the feedback loop tends to become instantaneous. In the event that a business user encounters problems with a particular application, having technologists on standby to ascertain and fix the issue quickly enhances the overall ecosystem.
Technologists are widely regarded as natural problem solvers. Being exposed to daily business activities, they often build a deep understanding of the business which, in turn, increases their overall effectiveness and leads to new and innovative solutions.
Traditional software development would often be run like a construction project. Phases were divided into requirement gathering, analysis, design, build, test and release. The process was lengthy and often led to results that didn’t necessarily match expectations. But iterative and incremental development processes have grown in popularity, especially since being formalized as “agile” in the early 2000s.
Delivering a little and often enables teams to get feedback, learn and amend along the way. With small steps and multi-skilled technologists working in close proximity to the business, layers of business analysts and project managers have often been replaced. Coupled with automated testing and deployment processes, the need for manual testers and release managers has also reduced.
A modern development approach results in a higher proportion of technologists producing code. In an agile team, more than 80% can code. That’s a doubling of producers. In traditional finance firms, one might expect to see less than half of technology teams writing code. With the efficiency gains of automation and continual feedback, the time spent producing also increases.
By way of comparison, Amazon releases new code to production every few seconds. While this degree and pace of change may feel uncomfortable in the highly regulated world of finance, it is possible and can potentially be advantageous. In reality, a periodic release bundles a number of changes together, which increases complexity and the risk of something going wrong.
By investing in automated testing, systematic release processes and a flexible architecture, releasing becomes easier and more predictable. When releases are easier and more predictable, changes can be introduced incrementally. The risk of each release is reduced, impact is minimized and diagnosing the issue becomes easier. Teams then spend less time resolving problems and more time adding value. The pace of change and ability to innovate is accelerated.