Cloud takes credit risk analytics to new heights.

High-performance computing (HPC) offers twice the speed at one-third the cost.

As part of its digital transformation, this financial giant needed to migrate its credit risk data processing and analytics to the cloud. Working against strict regulatory deadlines, a team of 10 from GFT delivered a cloud-native solution to replace the bank's legacy architecture. The new, cutting-edge system increased calculation speeds markedly, processing up to 220 billion calculations overnight, thus significantly reducing processing times and maintenance costs.

A breathtaking view of skyscrapers towering above the clouds in a misty cityscape, symbolising innovation and modern architecture. — A stunning image of skyscrapers emerging through a layer of clouds, representing the convergence of technology and innovation in urban development. This visual embodies the concept of cloud-native architecture and its potential to transform credit risk processing in the financial industry, aligning with GFT's leadership in cloud solutions and high-performance computing.

Scalable

The cloud-based solution is scalable and provides a 10-fold improvement over the legacy system, reaching up to 220 billion calculations overnight.

Massive cost reduction

Operating twice as fast, at one-third the cost of the legacy system, this has enabled the bank to reduce processing costs in perpetuity, and is now deploying the new platform across a further 38 locations worldwide.

Maintenance-free and compliant

Working against a strict regulatory deadline, a team of 10 from GFT delivered a maintenance-free cloud-native solution to replace the bank's legacy architecture.

Challenge

Risk is an opportunity

As one of the world's most respected banks, this global giant relies on modern technology to stay ahead of the competition. To support its digitalization strategy, it needed to migrate mission-critical credit risk processing to a modern cloud environment to boost performance, reduce costs and increase flexibility.

Calculating counterparty credit risk is central to the bank understanding, monitoring and controlling its counterparty exposure. This is critical to facilitate effective trading, risk management and accounting, and to meet regulatory compliance obligations.

With new regulations effective in January 2022, the bank urgently needed to implement a new system that would boost its processing capacity five-fold with minimal interruption to business-as-usual. After careful consideration, the bank chose GFT as its partner for this transformation.

Engagement

The power of partnership

GFT accepted the challenge of delivering a cloud-native credit risk data streaming and batch analytics architecture. Starting with a team of 10, GFT built a proof of concept within 12 weeks. This proved that Google Cloud Dataflow could process vast credit risk pricing and aggregation workflows.

The new cloud-native architecture replaced a legacy on-premise technology stack which struggled to process the vast amounts of data on time. From the outset, GFT designed the new solution to be flexible, scalable and fast. The new approach quickly demonstrated the speed of cloud processing with a 10-fold improvement over the legacy system, reaching up to 220 billion calculations overnight.

In addition to a HPC architecture for counterparty risk calculations, GFT expanded the scope of the platform to include what-if simulations for on-demand risk analysis to improve risk management.

Benefit

A future-ready bank

With the new architecture deployed in London and Hong Kong, the bank enjoys increased levels of business agility and responsiveness. Now it can develop bespoke analytics tools, rather than relying on external vendors. This means it can shorten time-to-market to maintain its competitive edge and confidently meet regulatory compliance deadlines.

The new platform offers a step-change improvement in risk management. Traders and risk managers can run multiple rounds of accurate risk exposure numbers throughout the day, free from the worry of system constraints. The bank can model potential global stress scenarios, for example, the impact of climate change or inflation. This empowers the bank to undertake more active hedging and risk management, and to align the trading book accordingly.

The bank is now future-ready and positioned for growth. The new architecture supports dynamic scaling to support rapid trade volume growth. Critically, the solution also enables intraday risk and improved computations. The bank has reduced processing costs permanently and has deployed the new platform across a further 38 global locations.

High-performance computing (HPC) | What is HPC and why does a bank need it?

Technical details

Cloud/GCP services and OSS (Open Source Software) deployed

Apache Beam – an open source unified programming model to define and execute data processing pipelines, including ETL, batch and stream processing
Google Dataflow – Unified stream and batch data processing that is serverless, fast and cost-effective
Google Cloud Storage – manged service for storing unstructured data offering features such as Object Lifecycle Management (OLM) and AutoClass
Google Kubernetes Engine (GKE) – the most scalable and fully automated Kubernetes service that puts your containers in autopilot, eliminating the need to manage nodes or capacity and reducing cluster costs with no Kubernetes experience required
Google Compute Engine – Secure and customisable compute service that lets you create and run Virtual Machines in Google Cloud
Google Cloud Load Balancing – High Performance Scalable load balancing on GCP Google Cloud Filestore-High Performance fully managed storage
Redis – an open source, in memory data structure store, used as a database, cache and message broker