With regulatory compliance such as FRTB and demand for innovation from financial Institutions, the pressure on quants has never been higher.
Models need to be optimized for Front to Risk consistency, and often need to be re-validated. The ability to deliver and test quickly is important and therefore performance, continuous delivery, agility and real-time feedback loop are hot topics in the industry today.
Adding to the increasing pressure, trying to scale with more quants can also be difficult. Platforms used within financial institutions are often not scalable outside those organizations, and results in long learning-curves.
We have created a cloud-based platform - FusionFabric.cloud, to allow open innovation in the financial industry between various parties such as banks, FinTechs, and students to create a collaborative ecosystem.
In fact, we have already tested it with some partners and universities - our recent engagement with University College of London, under the guidance of Professor Donald Lawrence, demonstrates the value of such a platform.
In early December 2017, we started to discuss the possibility of have students using FusionFabric.cloud, and in particular the valuation module, called Valuation as a Service (VaaS). After discussing the idea with Pr. Donald Lawrence, UCL, during the Finastra London Spark Day event, he found the concept brilliant and thought it would be great to include it in the Compliance, Risk and Regulation course starting in January 2018 with its 40 students.
We first listed all required tasks to be achieved, from the Cloud-based environment required, to the quantitative exercises to be run on VaaS. Then, we prepared a dedicated environment running on Microsoft Azure which was able to support 40 concurrent users. For the quantitative part, quants from Finastra prepared the presentation overviews, as well as the theoretical and practical exercises for each group to produce.
With students arranged into nine groups, we agreed with UCL on a format of 6 sessions, starting early February and finishing end of March:
- Presentation of VaaS, details on the project and complete overview
- Onboarding follow-up, collaboration deep-dive and communication
- Market risk regulation, FRTB, Value-at-Risk and Expected Shortfall valuation adjustments (xVAs), allocation methodology, collateral overview, impacts on valuation
- Introduction to Machine Learning applied to Finance
The course has been led by Laurent Chollet, Product Manager on FusionFabric.cloud/VaaS, Maha El Boury, Senior Quant on FusionCapital and Hughes Jan, Product Manager and Senior Quant on FusionCapital.
Exercises were also given, followed by corrections and feedback provided in the following session. Reviews of the exercises were done using the cloud platform and for the communication and live support, a dedicated channel on Slack was put in place.
At the end of the project, each group received its final assessment which accounted for 30% of their total grade. Below is some feedback from some of the participants:
“UCL Graduate students have gained tremendous exposure to business and quant teams at major global FIs and risk management solution providers like Finastra. Students have had hands-on experience pricing and modelling risk in a constantly evolving regulatory environment. Their first hand understanding often becomes apparent in interviews when students describe use cases they have worked on. We are proud and cherish these win/win partnerships.”
- Professor Donald Lawrence
“It has been a very rewarding experience to combine financial mathematics concepts with coding in the fusion fabric platform. The incorporation of git bucket with the platform makes it very efficient to work in groups. Overall, we learned a lot, and thank you for this opportunity”
- Group 2
“We highly appreciate the opportunity of working directly with people from the industry as it helps us get a better understanding of the real world. Thank you Finastra team once again.”
– Group 3
A similar experience is currently under preparation with New-York University and Fordham University (expected to start end of May).
The platform leverages behind the scenes SIMD, which means the code can leverage multi-core CPUs such as Xeon, or i7/i9 processors or GPUs like V100 from NVidia, which are available on Microsoft Azure which supports our infrastructure. Performance needed for FRTB or xVAs calculation can be achieved using a simple usage based cost model.
It allows banks and financial institutions to collaborate with academics, FinTechs and their own quants to speed up model development and validation thanks to a platform that can be shared.
FusionFabric.cloud also features a low code development platform to create UX business logic to showcase the work to end-users such as traders. With everything being real-time, feedback can be gathered and changes done "on the fly" which allow reactivity and shortened development phases.
To learn more, and see a live demonstration of the platform, please join me in the Quant 2.0 stream where I will be glad to present our work, and do come and visit us at our Booth #10, where you will have opportunity to meet Laurent, Maha and Hughes for in-depth discussions with the experts!
Finastra unlocks the potential of people and businesses in finance, creating a platform for open innovation. solutions enable customers to deploy mission critical technology on premises or in the cloud. Our scale and geographical reach means that we can serve customers effectively, regardless of their size or geographic location – from global financial institutions, to community banks and credit unions. Finastra will be at QuantMinds International this May.