Skip to content

Content Library

Back to Articles

Article April 20, 2026

Geopolitical shocks expose flaws in bank risk models

Geopolitical shocks expose flaws in bank risk models

Krishan Sharma, SVP, Model Risk - Regulatory Stress Testing and Capital Forecasting, CitiLinkedIn profile for Krishan Sharma

Krishan Sharma warns that traditional risk models are failing to capture fast-moving geopolitical shocks. Banks must adopt dynamic stress testing, integrated modeling, and qualitative overlays to better manage rising uncertainty and systemic risk.
  • Traditional models rely on historical stability that breaks during geopolitical shocks
  • Correlations converge under stress reducing diversification benefits
  • Stress testing must focus on velocity persistence and regime shifts
  • Integrated modeling needed to capture cross risk transmission
  • Firms should prioritize impact sensitivity over false precision probabilities 

Geopolitical risk has moved from the margins to the center of financial modeling, exposing fundamental weaknesses in how banks assess and manage risk, according to Krishan Sharma, SVP Model Risk, Regulatory Stress Testing and Capital Forecasting at Citi.

Sharma argues that the biggest limitation of traditional models lies in their reliance on historical data and stable relationships that no longer hold in a rapidly shifting global environment.

“The fundamental issue is that our traditional models rely on the assumption of stationarity,” he said, noting that these models are calibrated on the premise that future conditions will resemble the past.

That assumption breaks down under geopolitical stress. Events such as sanctions, trade disruptions, or military conflict do not behave like standard market shocks. Instead, they fundamentally alter economic relationships and correlations.

“A geopolitical shock changes the distribution entirely,” Sharma said, adding that models built on historical patterns struggle to capture the speed and scale of such disruptions.

One of the most immediate consequences is a breakdown in diversification. During periods of stress, correlations across asset classes tend to converge, reducing the effectiveness of traditional hedging strategies.

Sharma noted that “when a geopolitical shock hits, the correlation matrix often goes to one,” leaving portfolios far more exposed than expected.

To address these challenges, Sharma believes banks must rethink how they approach stress testing. Rather than focusing solely on the severity of shocks, firms should pay greater attention to how quickly and persistently those shocks unfold.

“We need to move beyond standard regulatory exercises which assume a linear degradation of variables,” he said, emphasizing the need to capture the velocity and persistence of change.

This shift requires more dynamic scenario design. Sharma highlighted the importance of modeling regime changes, where relationships between variables shift abruptly.

In a stagflation scenario, for example, the traditional inverse relationship between equities and bonds may break down, undermining portfolio diversification.

He also pointed to the growing importance of reverse stress testing. Instead of asking how portfolios respond to predefined shocks, banks should work backward from potential failure points.

“We need to ask what geopolitical event would break our liquidity position,” Sharma said, arguing that this approach better reflects the unpredictability of current risks.

Another key challenge is capturing the interconnected nature of geopolitical risks. Tariffs, sanctions, and supply chain disruptions rarely occur in isolation, yet many models still treat them as separate factors.

Sharma warned that “siloed modeling is the enemy,” advocating for integrated approaches that map how shocks transmit across market, credit, and operational risk.

He described tariffs as a clear example of a transmission mechanism, triggering market volatility, credit deterioration, and operational disruptions simultaneously.

To better capture these dynamics, firms are increasingly using techniques such as network models to map dependencies and quantify cross-risk correlations.

The difficulty of modeling geopolitical risk is compounded by its inherently qualitative nature.

Assigning precise probabilities to events such as trade wars or geopolitical conflicts can create a false sense of accuracy. Sharma cautioned against this approach, saying, “The danger of false precision is real.”

Events you may be interested in

View all of our upcoming events

Sign Up to Our Newsletter

Sign up

Connect Plus

Unlock more with Connect Plus membership

This content is available to Connect Plus members.