ALICE: A New InflationMonitoring Tool

Abstract

This paper develops Area-wide Leading Inflation CyclE (ALICE) indicators for headline and core inflation in the euro area. The main aim of the two leading indicators is to provide early signals about turning points in inflation. The applied methodology follows that of traditional composite leading indicators and the underlying indicators are carefully selected from a large dataset. An asymmetric band pass filter is used to extract the cyclical components of the time series. Nine series are selected for the headline ALICE with an effective lead time of 3 months.The core ALICE includes seven series and is designed to effectively lead the reference cycle by 5 months. Extensions of these indicators increase the effective lead times to 6 and 9 months, respectively. Both indicators identify main turning points in the inflation cycle ex post and they also perform well in a pseudo real-time exercise. ALICE performs better than other forecasts in terms of forecasting the direction of future inflation developments. In terms of quantitative forecast accurracy, random walk forecasts often performed better than ALICE but that was over some periods less clearly the case for forecasts from private and public sector organisations.

Density Forecasts of Inflation: A Quantile Regression Forest Approach

“…The data is obtained from the ECB Statistical Data Warehouse (SDW) and comes from a variety of original sources. Broadly speaking, the dataset is inspired by the Phillips Curve framework, covering different areas of the economy, and the choice of the variables is similar to de Bondt et al (2018).…”