Google Correlate and Google Trends as nowcasting tools for retail sales

Authors

  • Florencia Camusso Center for Studies and Services (BCSF), Argentina; National University of the Litoral, Argentina
  • Ramiro Jorge Center for Studies and Services (BCSF), Argentina; National University of the Litoral, Argentina

Keywords:

Big Data, Cycles, Google Tools, Nowcast

Abstract

The paper proposes a nowcasting model for Santa Fe’s supermarkets retail sales, an indicator that is released within two months of delay, internalizing information from Google Trends and Google Correlate. The procedure identifies an array of proxy variables with high predictive ability and then uses the data in order to estimate the target series considering searching patterns. Estimations computed by the model are compared to X13-ARIMA-SEATS’s forecasts. Obtained output suggests that results are not only consistent but also more opportune that official statistical releases.

Date of presentation: 02-19-2020

Date of approval: 01-11-2021

JEL classification: E27 ; E32

Downloads

Download data is not yet available.

Downloads

Published

2024-04-10

How to Cite

Camusso, F. and Jorge, R. (2024) “Google Correlate and Google Trends as nowcasting tools for retail sales ”, Ensayos Económicos, (76), pp. 26–45. available at: https://bcra.ojs.theke.io/ensayos_economicos_bcra/article/view/126 (accessed: 27 February 2025).

Issue

No. 76 (2021): Ensayos Económicos

Section

Articles