The Bank of England published Staff Working Paper No. 1,140 by Giovanni Covi and Tihana Škrinjarić, setting out a stochastic balance sheet microstructural model that jointly quantifies solvency and liquidity risks in stress testing, including endogenous bank reactions, feedback loops and amplification mechanisms. The approach is applied to granular loan, securities and funding exposure data for seven major UK banks over 2015 Q1–2024 Q1 to estimate “capital and liquidity at risk” metrics and a systemic default probability indicator, with results implying a historical average one-year bank default probability of 0.7% and a reading of 0.3% as of 2024 Q1 after peaking at 1.2% in 2020 Q1. Amplification effects driven by balance sheet constraints and behavioural responses account for around one third of default risk on average. Conditional Liquidity at Risk at the 99th percentile declines from about 10% to 6% of withdrawable deposits between 2015 and 2024, but counterfactual adverse stress analysis indicates outflows could rise to levels comparable to the Liquidity Coverage Ratio assumption (GBP 850 billion). On the solvency side, direct credit and market losses average around GBP 87 billion of Conditional Capital at Risk at the 99th percentile, with fire-sale spillovers averaging around GBP 23 billion; bank defaults most often occur via simultaneous breaches of both capital and leverage requirements, while liquidity constraints are found to be least binding in the period studied. Counterfactual exercises also test prudential tools, including the early-2020 dividend restriction, with results indicating that absent this tool average default probability would have been higher by 32 basis points between 2020 and 2021. The Bank of England notes the paper is research in progress published to elicit comments and does not represent Bank of England policy.
Bank of England 2025-08-01
Bank of England working paper models solvency–liquidity interactions and estimates UK banks’ one-year default probability at 0.7%
The Bank of England's Staff Working Paper No. 1,140 presents a stochastic balance sheet microstructural model to assess solvency and liquidity risks in stress testing for major UK banks. It estimates a historical average one-year bank default probability of 0.7%, declining to 0.3% by 2024 Q1. The research highlights the impact of balance sheet constraints and behavioural responses on default risk and evaluates the effectiveness of prudential tools like the early-2020 dividend restriction.