During bouts of intense volatility, risk managers must wrestle with billions of data points. To illustrate, in the immediate aftermath of Liberation Day earlier this year, our RiskSmart clients ran 100 million stress tests a day.
Extreme financial market events during the 2008 Financial Crisis led to government mandated stress testing. While regulation has eased since then, once in a century market shocks now occur with even greater frequency. In 2025, stress testing is no longer about meeting regulatory demand – it is about ensuring the survival of your business.
Risk compounds with crypto, heated geopolitical dynamics, and advances in AI
As financial institutions run billions of stress tests, data fuels the risk management engine. Processing the amount of data necessary for rigorous stress testing requires greater computing power than ever before. The breadth of inputs and combinations of factors they need to analyze at any given point are infinite.
Models therefore need to be constantly fed structured data so they can adapt to ever-changing parameters – building on existing, highly complex models. As data consumption increases exponentially, risk engines require more processing power to analyze and instantly deliver data in a digestible format for rapid, increasingly automated, decision-making.
Finding clarity in the cloud
In 2025, sound risk management is virtually impossible without cloud technology. Gone are the days of cumbersome on-premises data centers. The largest, most sophisticated financial institutions are rapidly migrating to the cloud.
While compute is invisible, it is more powerful than ever, and the attributes of SaaS architecture have become essential for financial institutions.
- The ability to scale and adapt: As market shocks become more frequent, data consumption will increase over time. Through SaaS-based risk platforms, models, tools and datasets continuously evolve, unbound by physical technology.
- Keeping up with the AI revolution: AI and machine learning are disrupting financial markets, and data is the lifeblood of automation. Greater compute and storage give firms the basic materials needed to automate vital but time-consuming processes – whether modeling, monitoring or reporting.
- Operational efficiency: Physical hardware is expensive to maintain and scale. Cloud compute solves this problem, enabling teams to increase processing power on demand while creating cost efficiencies.
- Connectivity: Global firms need their international offices to operate in harmony. Cloud is borderless and timeless, connecting teams from New York to London to Hong Kong. In a world where risk never sleeps, this continuity and connectivity is essential.
Resiliency is king
Moore’s Law stated that processing power per microchip will double every two years. This prediction has largely been proven true – the insatiable demand for and applications of data, however, were hard to predict.
To remain resilient, risk managers must harness platforms with significant scalable compute that will support the inevitable increase in processing power required to sustain their business and capitalize on opportunities during market shocks.
Cloud is not a silver bullet. Firms still need to weigh latency, data residency, cost optimization, and vendor lock-in when evaluating and building their architecture. But compared to on-premises infrastructure, cloud offers the only practical path to scaling risk management for the volatility we are facing. Financial institutions that have not migrated to the cloud face an existential threat to their business.
