This study examines the effectiveness of Value-at-Risk (VaR) models in UK banks from 2015 to 2024, spanning pre- and post-Brexit periods. Brexit is conceptualised as a political shock with systemic effects comparable to financial crises. Guided by Extreme Value Theory (EVT), the study evaluates Parametric and Historical VaR models against four objectives: predictive accuracy, statistical adequacy, cross-firm variation, and practical implications. Using daily returns of nine FTSE 100 banks, VaR was estimated at the 95% confidence level and validated through Kupiec’s Chi-Square backtesting. Findings reveal that Parametric VaR performed adequately in stable markets but underestimated tail risks post-Brexit due to Gaussian assumptions. Historical VaR more effectively captured fat-tailed volatility but varied by firm size and EU exposure. Larger internationally integrated banks faced greater exceedances, while domestically focused banks showed resilience. The study recommends hybrid frameworks integrating EVT-based approaches and urges regulators to embed multi-model validation into supervisory regimes.
Keywords: Brexit, Risk Management, UK banking, Value-at-Risk, extreme value theory, political shocks.
