No. 1026 - Nowcasting the Italian consumer price index using online prices and machine learning
The paper uses machine learning techniques to analyse online prices collected from the websites of 20 supermarkets belonging to a major retail chain across different cities, with the aim of forecasting in real time the inflation dynamics of selected food product categories in Italy.
Online prices can improve short-term forecasts of food inflation in Italy, providing timely and reliable estimates during periods of high price volatility as well, such as those observed after the pandemic and the Russian invasion of Ukraine. Machine learning models produce particularly accurate forecasts for product categories that are less affected by seasonal fluctuations, such as meat. In addition, using a limited number of representative products allows accurate results to be obtained in a highly efficient way. The findings suggest that the data available online can complement traditional data sources in inflation monitoring.
Instagram
YouTube
X - Banca d'Italia
Linkedin
RSS
E-mail Alert
