UK household-sector money demand during Brexit and the pandemic*

Conventional monetary aggregates assume perfect substitutability between monetary assets. We use monthly household-sector data to estimate two versions of the dynamic Fourier demand model to determine the degree of substitutability between U.K. monetary assets and find strong evidence against perfect substitution throughout our sample. Evidence of inelastic substitution is stronger after the 2008-09 recession and some elasticities of substitution are relatively low either following the Brexit vote or during the onset of the pandemic. Our results strongly support using Divisia monetary aggregates to gauge the effects of monetary policy in the U.K., since they do not impose any assumptions about substitutability. Annual growth rates of comparable household-sector Divisia and simple sum monetary aggregates had converged just before the start of the pandemic. Subsequently, money growth increased sharply and the simple sum increasingly understated the annual money growth rate compared to Divisia with the difference exceeding two percent by early 2021.