A stress-testing framework for integrating climate risks into financial stability analysis: Evidence from the Dominican Republic

This paper develops an empirical stress-testing framework to incorporate climate physical risks into financial stability analysis for the Dominican Republic. Our objective is to trace the transmission of physical risks to banks’ balance sheets, quantify tail credit losses, and map these losses into supervisory solvency requirements. To this end, we construct three scenarios for 2025–2030—baseline, credit valuation shock, and climate disaster—aligned with the NGFS short-term scenarios. Using an unbalanced panel of 44 financial intermediation entities over the period 2012–2024, we estimate credit losses and capital adequacy ratios, and quantify the value at risk and expected shortfall. The results show that, under the baseline scenario, losses remain largely contained and existing provisions and capital buffers are sufficient to absorb moderate shocks. Under the credit valuation shock scenario, borrower-level vulnerabilities materially erode capital, with capital adequacy ratios declining by as much as seven percentage points, as losses concentrate in commerce, construction, real estate, and other service activities sectors. Although larger banks appear capable of absorbing these losses, second-tier financial intermediaries emerge as important vulnerability nodes. Under the climate disaster scenario, compound hydrometeorological events generate substantially larger tail risks, with expected losses in the most adverse cases approaching half a billion U.S. dollars and capital adequacy ratios declining by at least five percentage points, pushing several institutions toward the threshold for supervisory intervention. Beyond the country case, the framework provides a replicable and policy-relevant tool for calibrating provisions, capital buffers, and risk-based supervision in small economies facing heightened climate uncertainty.