We show that payment system data can be used to improve forecasts of short-term changes in economic activity in Denmark. We compare the predictive performance of payment system data to a set of high‑frequency variables that includes industrial production and consumer prices, among many others.