Bank rate surprises, quantitative easing regimes and volatility connectedness in UK financial markets

Type of the article: Research ArticleAbstractThis study examines how conventional Bank Rate surprises and quantitative easing (QE) regimes relate to volatility connectedness across UK gilts, equities, sterling and oil. Monthly realized-volatility proxies for the 10-year gilt yield, FTSE 100, GBP/USD and Brent are constructed from public daily data for 2010–2025. A vector autoregression with generalized forecast error variance decomposition estimates total, directional, and net connectedness. Inference uses 2,000 bootstrap replications that re-estimate the decomposition and a moving-block bootstrap for policy regressions. Total connectedness is 39.65%, and corrected total connectedness is 52.87%. The FTSE 100 is the dominant net transmitter at 42.11 (95% interval 35.48 to 53.31), followed by gilts at 13.25 (4.62 to 24.19). Brent is a net receiver at –54.00 (–70.10 to –51.20); sterling, at –1.36 (–5.71 to 7.05), is indistinguishable from zero. Narrow-window Bank Rate surprises do not predict connectedness across five monthly lags, with robustness across alternative identifications and rolling windows and at 59 of 64 specification-grid points. A close-to-close two-year gilt proxy shares only 31% of its variance with the identified surprise and produces significant coefficients under Newey-West errors, but moving-block bootstrap inference removes them. None of the fifteen QE-active minus QE-inactive differences is distinguishable from zero, and the aggregate difference changes sign with crisis exclusions. Neither policy instrument shows a measurable monthly association with volatility allocation after accounting for surprise identification and estimation uncertainty.