Economic and Financial Review
Publisher
Central Bank of Nigeria
Keywords
Big data, Forecast, Inflation
Abstract
The success of monetary policy is substantially predicated on the availability of reliable forecast of inflation. However, the shocks arising from COVID-19 and the Russia-Ukraine war have brought about significant economic uncertainties; thus, necessitating the fine-tuning of existing forecasting models of the Central Bank of Nigeria. This study explores the usefulness of public sentiments obtained using machine learning methods to improve the predictive power of the existing short-term inflation forecasting model (STIF) in Nigeria. Findings indicate that, for all components of inflation, models that include the computed sentiment index perform better in both in-sample and out-sample forecasts than those excluding the index. Thus, we conclude that sentiment-based inflation forecasting models are useful for improving the headline inflation forecast and suggest the use of forward guidance monetary instruments in the form of “open mouth operations”, to ensure economic agents’ sentiments are well anchored.
Publication Title
CBN Economic and Financial Review
Issue
1
Volume
60
First Page
1
Last Page
23
Classification-JEL
C53, C55, E31
Recommended Citation
Adebiyi, M.A., Adenuga, A.O., Olusegun, T.S. & Mbutor, O.O. (2022). Big Data and Inflation Forecasting in Nigeria: a text mining application. CBN Economic and Financial Review, 60(1). 1 - 23.