Forecasting with Jennifer Castle (Magdalen College, Oxford)

"Forecasting UK inflation using evidence on the role of energy in productivity and prices," joint work with David F. Hendry and Andrew B. Martinez.

Abstract: The recent rise in energy and food prices has resulted in inflation levels not seen in decades in the UK. Many models trained on data since the 1990s could not capture the degree of variability experienced in 2022. But high and variable inflation is not a new phenomenon; the last century and a half has seen inflation levels at more than double those of 2022. We use a long span of data to build congruent models of wage inflation, unemployment, productivity and price inflation, allowing for dynamics, relevant variables, non-linearities and location and trend shifts using indicator saturation estimation. The production function reveals a major role for energy inputs additional to capital and labour. Combining these models results in a system in which price inflation is determined by many exogenous variables which can be used to generate scenario forecasts based on underlying assumptions about the conditioning variables. We produce forecasts using higher frequency data of the conditioning variables to result in projections for price inflation over 2023 and 2024. The substantial rise in oil and gas prices seen by mid-2022 contribute half of the increase in price inflation, and suggests that price inflation is likely to fall rapidly if energy prices fall in the next two years.

The seminar will be live-streamed in the Kendrick seminar so that attendees have the option to attend in-person.

Please sign up for our seminars listserv to receive the details.