When Risk Moves Faster Than the Model

Quantitative models remain essential to bank risk management. They give structure, consistency, and discipline to decisions across credit, market, operational, and enterprise risk. 

They also have limits. In a volatile environment shaped by geopolitical fragmentation, fast-moving technology, cyber threats, and financial crime, historical data may not fully capture what comes next. Qualitative tools, meanwhile, can help firms prepare for tail risks by presenting plausible scenarios, early warning indicators, and response options. 

In a ProSight Q&A, Joe Iraci, founder and CEO of Atlas Quotient Inc., said quantitative models became central because they matched the data, computing power, and analytical methods firms had available. Their limitation, he said, is that they are strongest “when the future resembles the past.” 

A more balanced approach can help leaders pressure-test what the numbers may miss. 

Use qualitative tools where risks are harder to measure. Qualitative assessment is especially valuable for risks that are difficult to observe, measure, or reduce to historical data. Iraci specifically pointed to non-financial risks such as operational resilience, culture, conduct, and governance. Elsewhere in the Q&A, he applied similar logic to reputation, human capital, geopolitics, and emerging risks. 

Treat geopolitics as a scenario problem. Shifting alliances, trade barriers, sanctions, regional conflicts, and competing spheres of influence can change markets, operations, and strategy. Iraci said firms need to consider possible outcomes and second- and third-order effects, rather than relying only on known-risk playbooks. 

Use AI to strengthen judgment. AI can help risk teams gather information faster, detect patterns across larger data sets, test scenarios, and improve monitoring and reporting. But Iraci cautioned that AI introduces model risk, governance questions, explainability concerns, and the risk of overconfidence in outputs that look precise. “AI should be viewed as an enabling technology rather than a replacement for risk judgment,” he said. 

Make reputation and people risk part of planning. Reputation can be difficult to define and measure, but Iraci said its effects can linger across customers, employees, counterparties, and franchise value. Scenario analysis and stress testing can help leaders examine how reputational issues might develop, spread, and be contained. Human capital also deserves a wider lens, since attracting and retaining talent can shape resilience, innovation, and long-term value. 

Bring discipline to emerging risks. For risks such as supply-chain disruption, ESG, and cyber threats, Iraci recommends defining the risk clearly, identifying when it could become material, and assessing likelihood and impact before translating the analysis into management guidance. Over the next 12 months, he said institutions should also watch global debt levels, geopolitical fragmentation, and political risk in the markets where they operate. 

The takeaway: Better risk measurement starts with knowing what each tool can and cannot do. Quantitative models bring discipline and structure. Qualitative assessment helps leaders add context, judgment, and preparedness when the past is an incomplete guide to the future.