Research
Work in Progress - drafts coming soon
Firms pricing in the UK: survey expectations and industry-level determinants (joint with Cristina Griffa, Nottingham University)
Abstract: We exploit the CBI survey that collects UK firms’ quarterly expectations and perceptions about their own-price inflation and about inflation in markets the firms compete in (own-industry) since 2009. Own-price inflation expectations are found to be robustly positively associated with firms’ price changes. There is evidence for inattention to industry-wide inflation, since own-price and own-industry expected inflation is often reported to be identical. However, when firms expect higher inflation in their industry compared to their own-price expectations, this is associated with additional price increases. The effects are asymmetric, i.e., non-existent when own-industry inflation is expected to be lower. This can reflect that firms receive information for their own pricing from industry-wide signals, and that they might be more willing to adjust price upward when they catch up with their industry and hence worry less about losing market shares. This can have aggregate implications since the effects of inflationary shocks could be amplified when they affect entire industries and firms are catching up to each other in their price setting.
Coverage: Catherine L. Mann: "Mind the gap(s): Inflation data and prospects". Speech given at the Official Monetary and Financial Institutions Forum.
Boosted Inflation: a machine learning inflation model with economic restrictions (joint with Marcus Buckmann and Philip Schnattinger)
Abstract: We enhance a machine learning method, boosted trees, with economic restrictions to separate out the (non-linear) role of supply and demand factors driving inflation. A large number of indicators and a number of identified shock series are fed directly into the model, which imposes sign restrictions for the association between indicators and inflation. Out-of-sample forecast performance for US inflation is competitive against an AR and a random forest. Initial results suggest that sign restrictions help isolating a positive (negative) association between demand (unemployment) and inflation, akin to a Phillips curve. Feeding the model with oil supply news shocks and a measure of global supply constraints helps separating out a supply contribution that has pushed up inflation recently, albeit to a lesser extent than demand factors. A financial block helps accounting for the “missing disinflation” during the global financial crisis. The learnt functional forms for the association between inflation and demand are close to linear, whereas for supply side and financial indicators non-linear functional forms suggest stronger effects during periods of large shocks or tail events.
Working Papers
"Flexible Bayesian MIDAS: time-variation, group-shrinkage, and sparsity" (with David Kohns) - BoE Staff Working Paper No 1025 - R&R at Journal of Business and Economic Statistics
"Forecasting UK inflation bottom up" (with Andreas Joseph, Eleni Kalamari, Chiranjit Chakraborty, George Kapetanios) - Forthcoming in Journal of International Forecasting
Publications
Potjagailo, G. and M.H Wolters (2023). Global financial cycles since 1880. Journal of International Money and Finance, Vol. 131, March 2023.
Jannsen, N., Potjagailo, G. and M.H. Wolters (2019). Monetary policy during financial crises: Is the transmission mechanism impaired? International Journal of Central Banking, Vol. 15, No. 4, October 2019.
Potjagailo, G. (2017). Spillover effects from Euro Area Monetary Policy across Europe: A Factor-Augmented VAR approach. Journal of International Money and Finance (72), 127-147.
Policy blogs
"Dissecting UK service inflation with a neural network Phillips curve" (with Marcus Buckmann and Philip Schnattinger), Bank Underground blog post, July 2023
"How broad-based is UK inflation?" (with Boromeus Wanengkyrtio and Jenny Lam), Bank Underground blog post, October 2022
"Global Financial Cycles since 1880" (with Maik Wolters), Bank Underground blog post, August 2020