There are many reasons why banks need to improve their management of risk. Not least of these is the rapidly approaching deadline for the implementation of the Fundamental Review of the Trading Book (FRTB) standard, a regulation detailing the capital that banks must hold against market risk exposures. The ability to report to the required level of detail is likely to prove a struggle for those banks that do not have appropriate measures in place. There are also a host of post-Brexit tax changes to negotiate, as well as upcoming requirements in the areas of open banking implementation and operational resilience policy, all of which require data handling and computational ability of the highest order. 

With the bar on risk management and compliance set so high, what are banks to do? For one thing, they need to be sure they have the right level of compute power at their fingertips, and for that they will need to look further than the isolated mainframe they may well still be relying on. Timely access to information is the key to success with risk management, and these days that means grid, in short the pooling of compute resources from not one or two but potentially hundreds of different machines located in any number of widely distributed places. The complex regulatory reporting required by FRTB necessitates this kind of approach, as does the need to run calculations at a more granular level in order that banks achieve competitive advantage by understanding their position ahead of rivals.

High performance computing (HPC), managed with a grid platform, can give a bank the processing power and scalable infrastructure they need to run millions of tasks using tens of millions of data points in parallel. This allows for faster results than any single machine could deliver, and more precise analysis of just the kind that today’s data-heavy challenges call for. It is useful not just for financial risk modelling and regulatory compliance, but a host of other areas besides, such as front office trade pricing, profit and loss calculations, model back-testing, financial fraud detection, customer engagement, credit risk assessment and cyber security. Grid has many business benefits, including: 

• A proven ability to enhance operational efficiency

• The power to transform real-time catastrophe modelling

• The capability to reduce a firm’s carbon footprint and associated overheads

• The significant reduction of operating costs

Grid is already the computational backbone of businesses operating in the world's most demanding markets. It is something that TIBCO has delivered for seven of the 10 largest names in banking – moving approximately 750k cores of compute to the cloud in the last 3 years. Now, grid is moving to the next phase. Allied to the power of cloud, it is becoming an even more formidable tool.

Unleashing grid’s power into the cloud already looks like a game changer, enabling banks to dynamically scale compute capacity up and down as required, only paying for what is really needed at any given moment in time through bursting and other automated techniques. It spells a permanent end to the need for expensive, under utilised, inflexible on-premise servers and data storage.

Grid, tied to the power of major cloud platforms like Microsoft Azure, Google Cloud Platform and AWS, is helping major financial institutions to tap into transformative capabilities across a range of mission-critical business areas. Banking giant HSBC, for example, was able to migrate its TIBCO HPC environment to Google Cloud Platform (GCP) with assistance from partner GFT. This meant it was able to process and store more data in a more secure fashion, while focussing on what it is best at, which is innovating in finance.

It’s about a more service-oriented approach to grid computing, ensuring you can burst workloads to the cloud when the demand arises. It is perfect for mission-critical value at risk management, complex pricing, capital adequacy and other computation-intensive applications, carrying out the required processing in minutes or seconds rather than hours. Grid is also a good fit with the multi-cloud deployments currently in service with so many enterprises.

Let’s examine a few of the use cases for grid-powered HPC deployment into cloud:

• In the capital markets arena, on-premise compute power is often sized to be just big enough to handle fluctuating overnight risk calculation batch runs. While there might be excess capacity during a normal trading day, there might not be enough during highly volatile trading sessions. This  can seriously impact business performance. Running HPC in the cloud enables capital markets firms to ensure there is always appropriate capacity available, even during periods of high market volatility.

• Alongside improved operational efficiency, the deployment of HPC into the cloud helps with the seamless implementation of FRTB and its complex and processing-intensive regulatory reporting requirements.

• Institutions can analyse their risk positions across different asset classes. As we’ve seen, this is great from a reporting and regulatory perspective as well as for capital adequacy and risk.  It is also important for providing optimal customer service. The customers that banks serve in capital markets are also dependent on high levels of compute power and the timely delivery of critical information.

• Grid, allied with cloud, is a useful component of other transformational technologies such as digital twinning and digital resilience. 

• It is great too for industries with exponential growth in data, such as enhanced segmentation in quant pricing and risk models in markets like insurance.

It is no secret that many banks are struggling with digital transformation because of the large amount of legacy technology they still operate. Grid gives them the ability to de-risk the change over from their on-premise data centre if they choose

HPC adoption in the cloud has been driven at different speeds by the institutions according to their cloud strategy, regulatory and risk profiles and market exposures.  Having taken this step for such a key deliverable, now is the right time for banks to be evaluating this technology and its potential to change their businesses.