Financial Rogue Waves
Dr Rosie Hayward explores what rogue waves can teach us about financial crises – using physics models to identify early warning signs in markets and assess whether extreme shocks can be predicted.
Abstract
Rogue waves are a type of extreme event observed in many physical systems, from optics to ocean physics. They are defined by their extreme heights relative to surrounding waves, their seemingly sudden appearance, and their frequency – occurring more often than Gaussian statistics would suggest. These are all features they share with extreme financial events.
This talk asks whether an equation that describes optical and water waves can offer insights into financial crises. By studying the linear modes of this nonlinear equation, we find evidence of Anderson-like localisation in volatility time series prior to the peak of a rogue wave, as well as potential early indicators of extreme events.
We explore the reliability of these signals in real time, when no future information is available to the signal processing algorithm, and assess whether they can predict the largest financial rogue waves.
About the speaker
Rosie Hayward is a postdoctoral researcher at the Austrian Supply Chain Intelligence Institute. Her research focuses on complex economic and financial systems, with a particular interest in extreme events, cycles, and self-interacting behaviour.
She also studies global multiplex networks of trade and financial flows, including development aid, remittances, and debt.
Rosie completed her PhD in Physics in 2021 as part of the CM-CDT at Heriot-Watt University, Edinburgh. Her research explored analogies between different theories in theoretical physics, drawing parallels with classical nonlinear optics.
From 2021 to 2024, she worked at Cambridge Econometrics, where she played a key role in developing and maintaining the global macroeconometric model, E3ME. She also contributed to the development of the Future Technology Transformation (FTT) model suite – a set of open-source, dynamic technology diffusion models used to evaluate decarbonisation policies.
This event is part of the Financial Computing and Analytics Research Group seminar series at UCL Computer Science.
