Skip to content

Content Library

Back to Articles

Article February 23, 2026

Geopolitics Breaks Models Before Markets Break

Geopolitics Breaks Models Before Markets Break

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

Krishan Sharma of Citi argues that traditional risk models are structurally unfit for fast-moving geopolitical shocks. From regime-switching stress tests to Bayesian overlays and dynamic model maps, he outlines how firms must rethink calibration, connecti
  • Traditional models assume stationarity and fail under structural geopolitical breaks
  • Stress testing must focus on velocity persistence and regime switching
  • Reverse stress testing identifies liquidity breaking events
  • Integrated scenarios needed to capture cross risk amplification
  • Bayesian overlays reduce false precision in qualitative risk
  • Model dependency mapping strengthens enterprise resilience

Geopolitical shocks are exposing the structural limits of traditional risk models built on historical stability and linear assumptions, but one senior executive at a leading New York bank argues that firms must shift toward adaptive, integrated, and sensitivity-driven frameworks that capture velocity, regime shifts, and cross-risk amplification.

Krishan Sharma, a Senior VP Model Risk - Regulatory Stress Testing and Capital Forecasting at Citi says only this approach will enable organizations to move beyond prediction toward true enterprise resilience.

“The fundamental issue is that our traditional models, whether VaR, Expected Shortfall, or PD and LGD models, are empirically calibrated,” he explains.

“They rely on the assumption of stationarity – that the future statistical properties of a system will resemble the past.”

That assumption collapses when sanctions regimes are imposed overnight or supply chains fracture within weeks.

“Geopolitical shocks are, by definition, structural breaks,” Sharma says. “A sudden sanction regime or a blockade does not behave like a standard deviation move in a normal distribution – it changes the distribution entirely.”

The result is a dangerous lag in detection.

“Traditional models are rearview mirror driven,” he notes. “When a geopolitical shock hits, the correlation matrix often goes to one, and historical calibration windows fail to capture the velocity of the transmission.

We are trying to model unknown unknowns using a dataset of known knowns.”

In today’s environment of stagflation risk, trade fragmentation, and policy volatility, Sharma believes stress testing must evolve beyond linear degradation assumptions embedded in standard regulatory exercises.

“We need to move beyond standard regulatory exercises like CCAR or DFAST which often assume a linear degradation of variables over nine quarters,” he says.

Instead, firms should focus on velocity and persistence. He advocates regime-switching scenarios that explicitly model relationship changes.

“In a stagflation scenario, the typical negative correlation between equities and bonds might flip to positive, nullifying diversification benefits,” he explains. That structural shift, not just the magnitude of GDP decline, drives portfolio impact.

Reverse stress testing is another critical tool. “Instead of asking what happens if rates rise 500 basis points, ask what geopolitical event would break our liquidity position and work backward,” Sharma recommends. He also warns against assuming stabilizing policy responses.

“We need to stress test for policy errors or volatility – where central bank reaction functions become unpredictable or contradictory to fiscal policy.”

Tariffs, sanctions, and supply chain disruptions rarely occur in isolation. Sharma argues that siloed modeling obscures their combined impact. “A tariff is not just a market risk factor – it is a transmission mechanism,” he says.

He outlines an integrated scenario framework mapping causal chains from political event to market volatility, credit deterioration, and operational disruption.

“You cannot simply sum the market VaR, credit VaR, and operational risk capital,” Sharma emphasizes. “You must quantify the cross-risk correlations which tend to spike during these specific events.”

Tools such as network theory and Bayesian belief networks are increasingly useful in capturing these interdependencies.

Yet geopolitical modeling also risks false precision. “Assigning a 12.4 percent probability to a trade war is meaningless,” Sharma cautions.

Instead, he recommends formal qualitative overlays supported by governance: “Accept that the core model cannot capture the risk. Apply a governance-approved management adjustment based on expert judgment.”

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.