Holding Periods: Measuring the Inverse of Money Velocity from Transaction Records

Abstract

This paper defines the average holding period of money and proposes a methodology for fully disaggregated measurement. Our measure is shown to be the inverse of the transfer velocity of money under stationary conditions, which is implicitly assumed in conventional aggregate measurement. Our methodology does not require stationarity. We leverage a recent computational technique to extract empirical holding periods from micro-level transaction data as recorded by real-world payment systems. This enables novel empirical analyses of money velocity under non-stationarity conditions. We illustrate several such analyses on Sarafu, a small digital community currency in Kenya, where transaction data is available from 25 January 2020 to 15 June 2021. Our measure implies faster circulation than does the aggregate transfer velocity; we can say that 58% of Sarafu was effectively static. We also disaggregate by geography to study the heterogeneous impact of economic disruptions related to the COVID-19 pandemic on Sarafu. Finally, we consider an ad-hoc currency management operation that took place in October 2020. Measuring the average holding period of money makes it possible to track money velocity during ongoing monetary interventions and on other important occasions when conditions are not stationary.