Stephane Rio, CEO and Founder of The Independent Calculation Agent (ICA), and Jérôme Maetz, Head of xVA quants at Santander, Discuss The Future Of CIBs In Capital Markets

Do you believe that CIBs (corporate & investment banks) need to innovate to remain relevant in today's financial value chain?

[JM] In the past few years, market has evolved towards a significant increase of analysis and reports requirements, especially on the regulatory side, pushing information systems to their limits. The amount of data and monitoring constraints as well as real time requirements are driving number of innovations today, in particular if one wants to capture the full economics of a deal. New technologies are also opportunities to provide tools to assist trading in a better risk management of their books, leveraging on the potential of artificial intelligence in analysis and prediction as well as cross validation.

[SR] In the past few years, market has evolved towards a significant increase of analysis and reports requirements, especially on the regulatory side, pushing information systems to their limits. The dimensions of the problems to solve have changed but yet most of the systems did not. To address the explosion of data and the exponential increase of computational power requirements, banks have the options to use shortcuts, approximations, impoverish their dataset to reduce it to a more manageable size, or endorse new technologies which are well adapted to those new dimensions.

What technologies do you believe will be most influential in the transformation of CIBs in the near future?

[JM] Big data technologies are definitely a game changer opening new fields of research. Combined with machine learning / artificial intelligence algorithms, they are powerful tools with an extremely large spectrum of application, from basic automation of reporting to market analysis, risk management etc… These technologies also break the traditional frontiers between departments reducing the time to delivery and the way people exploit the information making necessary to rethink general architecture and platforms.

[SR] Two key aspects that have direct application on today’s pricing and risk environment are the migration towards big data manipulation (big data viz and calculation) and cloud computing, and more generally endorsing a big data architecture.

Hardware requirements have grown exponentially over the last 10 years and the only real way to stop it is to start to use resources on demand. If a bank is not comfortable putting all their data on the cloud, one can implement an architecture that leverages a hybrid infrastructure where the confidential information stays in a private, secure area and core calculations (still representing the bulk of the computational burden) are sent to a public cloud.

Similarly, the amount of data to manipulate has been multiplied by a huge factor: XVA and FRTB are 2 prime examples of this. Big Data technologies allow to handle hundreds of billions of data as easily as when we were dealing with hundreds of thousands in the past: this is a radical game changer that calls for re-architecturing many legacy systems as it allows for a central source of common intermediary data that, aside from being calculated only once, enables more interactivity, analysis, etc. and can be used for multiple usage: for instance, a what-if scenario can be optimised by just re-computing the piece that changed, if any hanged at all (accelerating what-if scenarios from minutes to near real time).

What are the key elements holding CIBs back from implementing new technologies and operational initiatives?

[SR] While ICA had the benefit of starting from a blank page (with a “day one” fintech start-up spirit and with fresh ideas brought from BigTech experts from outside the banking world), banks need to take their existing legacy set-up into account and deal with numerous urgent short term constraints while the lack of innovation and post-crisis lower attractiveness of the industry has minimised the impact of new blood.

From a pure practical viewpoint, a parallel run of the banks’ legacy systems with cost sensitive new generation solutions would be ideal but is challenging on a big scale.

Looking at certain silos (starting with the most demanding ones like XVA and FRTB), this is now made possible with the emergence of Software as a Service (“SaaS”) in capital markets, many years after the Silicon Valley retired installed software. Installation are made in multiple of weeks instead of multiple of semesters, upgrades are seamless, overall TCO are very significantly reduced, response time to new regulations are no headaches anymore.

The other approach is to break down existing set-ups into smaller components and replace them with new generation building blocks. This is what our “custom approach” is all about and allows to modernise legacy architectures with a step-by-step approach.

[JM] Traditionally, corporate & investment banks incentivized a high level of specialization leading to a Taylorism of tasks, vertically and horizontally. Today, to keep on evolving, there is a need for a comprehensive transversal approach based on maximizing the overall outcome without necessarily looking at individual benefits. This leads to the development of pools of expertise rather than specific mandates. This tendency coincides with the large democratization of coding and data analysis we are experiencing these days. As a direct consequence, hierarchy and tools evolve, raising a number questions on how to deal and adapt to these fundamental changes. Regulation in this context acts as a booster and a breaker. This also raises the point of ongoing training, competences and adaptability to change. Additionally, banks need to keep on attracting talents in a context where different industries are fighting for the same skills.

Why are you involved in CMIS 2018 and what is your message to the industry?

[JM] This year, I will be commenting on a specific initiative Santander took on the XVA space, working with a specialized fintech, ICA. I have the pleasure to jointly present with Stephane, the CEO of ICA, some of the challenges we are facing on the technological aspect as well as on the key question of how to build a fruitful partnership with these fintechs.

[SR] In an environment where innovative initiatives are scarce, I often hear innovation officers or business managers ask how to convince their internal stakeholders to get going and test. I therefore thought it could be of value for many people to hear on which set up we chose with Santander to work and how fruitful it has been for both parties. Finally, I’d very much like to encourage decision makers to support and incentivise their staff towards internal or external innovations and allocate some resources to invest on those new initiatives which will overall drive their competitiveness post 2020. A number of very mature big data technologies are available and used in many other industries; it is very frustrating not to see them at more at work in an environment which they are very well suited for.

Make sure to download the Capital Markets Innovation Summit agenda to check out all of the great activities, speakers, and sessions planned for this year.

You can hear more from Stephane Rio in his Case Study: How to establish innovative partnerships with Fintech companies in order to develop new IT models and solutions that compliment in-house legacy systems on Day two at 11:30.

Learn more about The Independent Calculation Agent.

The opinions expressed in this interview are the author's own and do not necessarily reflect the view of Grupo Santander.