Bank of Ghana researchers propose AI-driven e-Inflation framework

Researchers at the Bank of Ghana have proposed a new framework named e-Inflation that leverages big data and artificial intelligence to monitor consumer prices in real-time. Developed by Francis Mawuli Abude, Collins Kwesi Appiah, and Julius B. Dasah, the system is designed to provide central banks with faster, more frequent inflation estimates at a significantly lower cost than traditional survey methods. The initiative aims to address the inherent delays in current data collection practices, which often rely on outdated estimates for official reporting.

The e-Inflation framework functions by automating the collection of data from over 120 websites, allowing researchers to analyze approximately 500,000 price observations. This vast data pool enables the system to estimate the Consumer Price Index (CPI) and differentiate between headline, food, and non-food inflation. The researchers have been refining this AI-driven approach since May 2021, focusing on how alternative price sources align with the official basket of consumer goods used by national statistical agencies.

Preliminary findings indicate a strong correlation between the e-Inflation estimates and official inflation data, suggesting the tool could be a reliable precursor to formal announcements. However, the authors noted several challenges, including the need to ensure the accuracy of online prices compared to physical store prices and the necessity of maintaining the freshness of online price listings across various e-commerce platforms. Despite these hurdles, initial surveys suggest that online and physical prices remain statistically similar in most contexts.

The proposal represents a significant technological shift in monetary policy analysis within Ghana. By adopting such an AI-driven approach, the Bank of Ghana could gain a more agile perspective on economic shifts, allowing for more responsive policy decisions. The researchers encourage other central banks to consider similar frameworks to enhance their economic monitoring capabilities while remaining mindful of the technical requirements and data reliability issues associated with big data analytics.