Reliable real-time economic information is essential for policy making but difficult to obtain: publications of official statistics are often delayed, and autoregressive estimates fail to capture unanticipated events. Economic nowcasting can fill the data gap. We present a granular nowcasting method for macro- and industry-level GDP by using a network-based approach, that relies on a novel data set of real-time monthly inter-industry payments in the UK covering 2016-2023. The data captures shocks travelling along the supply chain. Our model is an extended version of generalised network autoregressive model, tailored for networks with time-varying edge weights and nodal time series. This framework allows us to track the impact of network effects arising from input-output dynamics in supplier and customer industries. Supply chains are important for understanding economic outcomes, as powerfully illustrated by recent supply chain disruptions.
About the speaker
Kerstin is an academic visitor at INET and a postdoctoral researcher in the Finance and Economics Programme at The Alan Turing Institute in London.
She studied economics at the Universities Tübingen and Bonn in Germany, and obtained her PhD in a joint degree programme from the universities Paris-1 Sorbonne-Panthéon and Bielefeld. In 2020-2022 she was a postdoctoral researcher at the Oxford Martin School before joining the Turing Institute.
Kerstin's research focuses on technological change and, more specifically, on technology transitions when incumbent technologies are replaced by emerging alternatives. Major applications of her work are climate technologies and substitution patterns in economic and technological networks.
She has published empirical and theoretical work using macroeconomic, agent-based simulations, patent citation and input-output networks, and other applied empirical analyses. An overview of her research can be found on her Google Scholar entry. In her current role at the Turing Institute, Kerstin is exploring a novel data set based on granular financial transactions data of businesses in the UK.
This event will be online via Zoom and in person.
To join, please register using the button above, after which you will be sent joining instructions and the seminar room location via calendar invite.
For further information please contact: events@inet.ox.ac.uk