What Max Verstappen and Red Bull Racing can teach capital markets about risk calculations

Oracle Red Bull racing has won seven World Drivers’ Championships, and six World Constructors’ Championships since it began its Formula One participation in 2004. The team relies on technology to help it with every facet of the game, including car design, analyzing massive quantities of data (4 billion simulations every race weekend!) ?for race insight, driving simulation and communication.

Given the transient nature of Formula 1,? twenty four weekends in twenty four different cities over the course of nine months, the underlying technology that powers the billions of simulations, ?is required to be portable (the race teams require the ability to spin up computing grids quickly and spin them back down once the races conclude) and highly performant (in an environment where wins and losses are measured in hundredths of seconds strategy and decision are measured in microseconds).???

In comparison, capital markets firms (both buy/sell side) as well as market infrastructure organizations and exchanges operate in environments requiring similar levels of simulation requirements to accurately price, and calculate key risk metrics on the financial instruments and the markets that they operate in. ???In contrast, very few capital markets firms have adopted the technology that is currently available to deliver the business outcomes that their businesses, shareholders, and regulators require.

According to a study by Accenture in 2023 (Banking Cloud Altimeter Volume 6 | Accenture),? only 15% of ALL eligible workloads had migrated to the cloud and only 4% of capital markets workloads had migrated to the cloud.?? ?The cloud offers significant price and performance benefits compared to alternative on-premise based architectures that are slow to scale, have a fixed set of limited capacity in backlogs and run times, and limited compute options.

?

A typical day for high performance compute platforms looks like this at most capital markets firms:

Imagine being able to automatically provision resources at times of bottlenecks and then shut down close to zero when no longer required (very similar to race day weekends for Formula 1). ??This technology (high performance grid computing) is available in the cloud and can be leveraged by capital markets firms to not only solve the problem above but to improve resiliency, fail over processing all at a ?fraction of the cost.?? In other words, ?with the right solution the move can often be self-funding.

To examine potential fit of the right solution, it is helpful to examine the the potential solutions through the lens of price and performance.

?

Not all clouds are created equal- Price

While high performance grid computing is the answer to the key risk calculation problems that capital market firms face, choosing the right grid computing solution and cloud are critical in achieving the price performance advantage capital markets firms need to unlock.?? The hyperscalers all offer different pricing models, architectures, and pros and cons in their approach to their high performance grid solutions leading to a large disparities in both price and performance.?? The below represents list pricing ?for compute, storage, and networking requirements and associated costs for a Tier 1 financial institution to implement high performance grid computing for their capital markets needs between two hyperscalers:

Not all clouds are created equal- Performance

STAC-A2?is the technology benchmark standard based on financial market risk analysis based on technology from some of the world’s largest banks, STAC-A2 reports the performance, scaling, quality, and resource efficiency of any technology stack that can handle the workload. ??STAC-A2 noted that compared to all publicly reported, non-accelerated cloud solutions to date, this solution (Oracle Cloud) set numerous records:

??????? The best price-performance (19,788 options / USD) over 1 hour and 3 days of continuous use[1]

??????? The highest throughput (119.2 options / second)[2]

??????? The fastest warm time (24.3 ms) in the baseline Greeks benchmark[3]

??????? The fastest cold (2.98 s) and warm (1.96 s) times in the large Greeks benchmarks[4]

??????? The most correlated assets (145) and Monte Carlo paths (7,000,000) simulated in 10 minutes[5]

?

In summary,? given the external and internal pressures facing banks in general and capital markets firms in particular,? Chief Risk Officers and their technology counterparts can apply the same tools (high performance compute grids) and techniques to deliver winning outcomes to their capital markets business partners, shareholders, and regulators.

?