Predicting Turning Points in the UK Inflation Cycle

Working Paper: CEPR ID: DP880

Authors: Michael J. Artis; Robin C. Bladen-Hovell; Denise R. Osborn; Graham W. Smith; Wenda Zhang

Abstract: We set out a reference chronology for annual UK inflation, identifying nine complete cycles between 1958 and 1990. Inflation over this period is asymmetric, falling more quickly than it rises. Leading indicators are also proposed, with composite shorter and longer leading indicators constructed. These exhibit leads at peaks which are generally longer than at troughs. An evaluation of the composite series is also conducted to indicate how they might perform in real time. Overall, they give clear prior turning point information with relatively few `false signals'. Nevertheless, we find that inflation troughs are more difficult to predict than peaks.

Keywords: inflation; leading indicators; forecasting; Bayesian; cycles

JEL Codes: E31


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
Composite leading indicators (C43)Inflation turning points (E31)
Unemployment (J64)Inflation rates (E31)
Retail sales (L81)Inflation rates (E31)
Inflation turning points (E31)Asymmetrical relationship (troughs harder to predict than peaks) (E32)

Back to index