Stress Test (Bank)

A bank stress test is a forward-looking analysis used by regulators and banks themselves to estimate losses, revenues, and capital ratios under hypothetical shocks — typically a sharp recession, spike in unemployment, collapse in asset prices, or specific financial-market disruptions. The goal is to determine whether an institution holds enough high-quality capital (notably Common Equity Tier 1) to keep lending and absorb losses without taxpayer support.

Modern stress testing emerged as a core supervisory tool after the 2007–2009 global financial crisis. In the United States, the Federal Reserve launched the Supervisory Capital Assessment Program (SCAP) in 2009, which examined the 19 largest bank holding companies and required several to raise additional capital. SCAP was institutionalised under the Dodd-Frank Act (2010) as the annual Dodd-Frank Act Stress Test (DFAST) and the related Comprehensive Capital Analysis and Review (CCAR), which also evaluates planned dividends and share buybacks.

In the European Union, the European Banking Authority (EBA) runs EU-wide stress tests in cooperation with the European Central Bank, European Systemic Risk Board, and national supervisors, with exercises conducted in years including 2011, 2014, 2016, 2018, 2021, and 2023. The Bank of England runs a separate annual cyclical scenario for UK banks.

Tests typically include a baseline and one or more adverse scenarios specifying paths for GDP, unemployment, interest rates, housing and equity prices, and sometimes country-specific shocks. Banks project losses on loans, trading positions, and counterparty exposures, then compare resulting capital ratios against regulatory minima.

Critics argue stress tests can become predictable, underweight tail risks (such as rapid interest-rate rises or non-bank contagion), and rely on models opaque to outsiders. Defenders note that no major US or EU bank that passed recent tests has failed without idiosyncratic causes — though the 2023 collapse of Silicon Valley Bank, which was below the largest-bank testing threshold, prompted renewed debate over scope and scenario design.