Exploring the Use of Anonymized Consumer Credit Information to Estimate Economic Conditions: An Application of Big Data

Abstract

The emergence of high-frequency administrative data and other big data offers an opportunity for improvements to economic forecasting models. This paper considers the potential advantages and limitations of using information contained in anonymized consumer credit reports for improving estimates of current and future economic conditions for various geographic areas and demographic markets. Aggregate consumer credit information is found to be correlated with macroeconomic variables such as gross domestic product, retail sales, and employment and can serve as leading indicators such that lagged values of consumer credit variables can improve the accuracy of forecasts of these macro variables.

Keywords: Consumer credit information, Administrative data, Big data, Real-time data, Nowcasting, Forecasting

JEL Classification: C53, C55, D12, D14