A Regression-Based Composite Indicator for Economic Forecasting: Integrating Financial and Real-Sector Signals to Improve Macroeconomic Nowcasting

This paper develops a multivariate regression-based indicator constructed from non-traditional economic variables to improve real-time estimation of aggregate economic output and to identify potential economic downturns and recessions in future years by mimicking GDP at a higher frequency than the official value is released (i.e., quarterly). This indicator reduces economic reporting delays by blending real-activity data with forward-looking financial indicators through a structured lag framework, thereby capturing both general and specific economic dynamics in the global economy. The results of this analysis provide an accurate indicator (89.40% similarity to the industrial production index) and diagnostic insights from residual analysis. The results indicate that combining leading and contemporaneous indicators improves the robustness of macroeconomic nowcasting frameworks. The proposed indicator demonstrates the value of integrating multiple non-traditional economic channels into a unified framework for real-time macroeconomic monitoring, This paper argues that GDP is able to be accurately tracked using non-traditional, high frequency indicators, overcoming reporting delays associated with GDP quarterly release by using a multivariate regression model combining real-activity, financial stress, and labor market indicators and data.