FRANKFURT AM MAIN — Opening a joint conference held by the European Central Bank and the Bank of England, ECB Executive Board member and Supervisory Board Vice-Chair Frank Elderson warned financial leaders that traditional risk management models are rapidly becoming obsolete in an increasingly interconnected and unpredictable global economy.
Addressing multidisciplinary experts from 18 countries and three continents gathered to discuss climate risk supervision in a non-linear world, Elderson emphasized that the financial sector must fundamentally evolve its approach to risk. For generations, modern finance has relied on a foundational assumption: that with sufficient historical data, comprehensive borrower behavior insights, robust modeling, and a solid grasp of past patterns, financial institutions can estimate potential losses with a high degree of precision. These calculations have traditionally guided how banks price risk, establish loss provisions, and determine capital reserves.
While these conventional methods remain vital and are frequently supplemented by stress testing, underwriting standards, and risk-appetite frameworks, Elderson noted that they were built for stable relationships. Today, the global financial system is confronting a rapidly shifting landscape marked by geopolitical fragmentation, demographic shifts, artificial intelligence, quantum computing, and the escalating crises of climate change and nature degradation.
Although these macro trends differ significantly in their origins, scopes, and speeds, they share a critical characteristic: they emerge within complex, interconnected systems exhibiting non-linear behavior. In such environments, minor adjustments can trigger disproportionately large disruptions, amplifying shocks through feedback loops and driving abrupt transitions across critical tipping points. Tail risks that once seemed remote are rapidly becoming tomorrow’s baseline realities.
The Evolution of Risk: From Known Knowns to Radical Uncertainty
Navigating this new era requires categorizing challenges through a framework of what is known and unknown. Elderson broke down this approach into three distinct tiers: the known knowns, the known unknowns, and the unknown unknowns.
In the realm of known knowns, significant progress has been made over the past decade. Banks and regulatory authorities have established foundational architectures to identify, quantify, and manage climate and nature risks. The physical impacts of rising temperatures are no longer just visible; they are increasingly measurable. Extreme weather events over recent summers have demonstrated severe tolls on economic activity, labor productivity, and international supply chains. Furthermore, the financial transmission channels are becoming clearer. Enhanced climate hazard and geo-location data allow institutions to model the impact of droughts and floods on standard financial metrics such as loss given default.
However, average exposures remain high. Data shows that since 2022, the average share of interest income earned from carbon-intensive sectors by significant supervised banks has remained consistently above 70 percent. While transparency has improved through public data sources covering various physical risk hazards, these known factors only reveal part of the picture. Because forward-looking climate and nature risks diverge sharply from historical precedents, they remain inherently difficult to quantify fully.
This limitation leads directly to the second category: known unknowns. These are risks recognized to exist, yet their exact timing, magnitude, and probability remain uncertain. Economic forecasting faces a similar dilemma, as both the economy and the climate function as complex adaptive systems. Scientists have identified Earth system components where tipping points may already be approaching, including the dieback of the Amazon rainforest and the potential collapse of the Atlantic Meridional Overturning Circulation. Similarly, compound events—such as simultaneous heatwaves and severe droughts, or concurrent agricultural disruptions and energy system pressures—complicate forecasts because they fall outside historical datasets.
The Role and Limits of Scenario Analysis
To address these known unknowns, scenario analysis has emerged as an indispensable tool for structured exploration rather than precise forecasting. The Network for Greening the Financial System (NGFS) scenarios have established a global baseline, offering a consistent and transparent framework for central banks and financial institutions to assess long-term impacts.
Yet, Elderson cautioned against falling into a scenario "tunnel vision," where capital planning becomes anchored to a narrow set of assumptions. Current climate models often reflect scientific caution rather than exhaustive worst-case realities. For instance, while the Intergovernmental Panel on Climate Change confirms that further warming will thaw permafrost and release additional carbon, most global climate models do not directly represent this feedback loop, nor do they fully account for Arctic wildfires or warming wetlands. Research indicates that such natural emissions could add up to 0.6 degrees Celsius of warming in an adverse scenario.
Within the internal capital adequacy assessment process, climate stress tests are often utilized primarily to calibrate buffers for otherwise uncovered risks, yet banks currently incorporate all material risk drivers in only about 15 percent of cases. Elderson stressed that scenarios must be actively challenged, adapted to local contexts, and combined with expert judgment rather than treated as simple box-ticking exercises.
Radical Uncertainty and Building Institutional Resilience
Moving beyond measurable risk brings financial supervisors into the territory of genuine unknown unknowns—what economists term radical uncertainty. In this sphere, future events, probabilities, and even relevant outcomes cannot be measured with confidence. Institutions cannot map every complex interaction, hidden feedback loop, or emergent transmission channel before a shock occurs.
Evidence suggests that climate and nature crises are unfolding faster and more severely than anticipated. The Arctic is warming at nearly four times the global average, Europe is warming at twice the average, and ocean heat levels have reached record highs while ecosystems lose resilience at unprecedented rates.
When facing radical uncertainty where future shocks cannot be accurately modeled, the primary defense cannot rely solely on scenarios. Instead, banks must invest in resilience before disasters strike. This involves maintaining adequate capital, liquidity, operational capacity, and strategic flexibility, alongside a robust and risk-sensitive reaction function. Financial institutions must regularly test their decision-making structures, refine their underwriting and pricing strategies, evaluate insurance protections, and integrate climate adaptation and mitigation investments into their core business models through prudent transition planning.
These lessons extend far beyond environmental risks. Cross-cutting challenges like demographic shifts, rapid advancements in artificial intelligence that heighten cyber and operational risks, and the long-term vulnerabilities posed by quantum computing to data encryption all share the same fundamental denominator: radical uncertainty. Because backward-looking data cannot adequately capture these structural shifts, supervisors and financial institutions must upscale their risk management frameworks.
Concluding his address, Elderson likened modern risk management to driving along a winding mountain road. Preparedness requires more than simply reacting when a cliff suddenly appears; it requires sturdy guardrails and contingency measures for the moments when the road veers dangerously close to the edge. For known knowns, the task is measurement; for known unknowns, preparation; and for unknown unknowns, resilience.
