This time the IIF UK Chapter hosts the Quarterly Forecasting Forum under the theme “Planning in times of uncertainty: A forecasting perspective on Covid-19”. This forum will be held in the form of an online meeting hosted via Zoom. We would appreciate if you could register, so that we know how many participants to expect, and to provide you the necessary reminders. The registration is totally free.

You can join the meeting using the link: https://us02web.zoom.us/j/83165209102 or by search Zoom using Meeting ID: 831 6520 9102

The agenda of the event

Research presentations:

• 14:00 – 14:15 Short-term forecasting of the Coronavirus Pandemic, Dr Jurgen A. Doornik, Senior Research Fellow, University of Oxford, UK.

• 14:15 – 14:30 Forecasting in retail during uncertain times, Trevor Sidery, Lead Data Scientist, Tesco PLC, UK.

Question and Answer Panel

• 14:30 – 15:00 Theme: The role of forecasting in responding to the Covid-19 Pandemic.

• Panel: Dr Jurgen A. Doornik, Trevor Sidery, and two others TBC.

Online Pub Quiz

• 15:00+ Online Forecasting-themed Pub Quiz for those who want to stay longer and chat. Don't forget to bring your beer!

The speakers and their abstracts

Jurgen A. Doornik is James Martin Fellow, Institute for New Economic Thinking, University of Oxford. He moved to Oxford to work for David F. Hendry as a research officer, did his PhD there, and has worked since on several research projects at Nuffield College and the University.

He has published widely in professional journals on econometric methods, modelling, as well as software, numerical methods, and mathematics. Jurgen also created the Ox language, and the Autometrics algorithm for automated model selection. He has recently developed new forecasting techniques, as reflected in the real-time COVID-19 forecasts at www.doornik.com/COVID-19.

Short-term forecasting of the Coronavirus Pandemic

We have been publishing real-time forecasts of confirmed cases and deaths for COVID-19 online at www.doornik.com/COVID-19 from mid-March 2020. These forecasts are short-term statistical extrapolations of reported data, assuming that the underlying trend is informative, without requiring other assumptions of how the SARS-CoV-2 virus is spreading. The forecasts are based on extracting trends from windows of the data, applying machine learning, and then constraining the flexible extracted trend. The methods have been effective in the COVID-19 setting, with lower absolute errors than some epidemiological models.

Trevor Sidery did his PhD in astrophysics in the University of Southampton. After a couple of post-doctoral positions that lead him from studying neutron stars to signal processing for gravitational wave astronomy, he joined the grocery retailer Tesco 6 years ago. He now leads the area of forecasting within the data science team, looking for ways to improve the current algorithms, and to incorporate best practices in forecasting across more of the business.

Forecasting in retail during uncertain times

Significant events that impact forecasts are not unheard of, but in the realm of grocery retail these have been largely trending behaviour. The recent pandemic has not just created a sudden stocking up ‘shock’, but has changed our buying patterns for an unknown period of time. I present some of the challenges of forecasting that we face in the grocery business, both in the near and farther future, as well as some ways we have tackled these issues.