Enhancing Business Cycle Forecasting in Pakistan: A Composite Leading Indicator Approach with PLS-SEM

Abstract

Objective: This study aims to forecast the business cycles in Pakistan by developing Composite Leading Indicators (CLI) that capture multi-dimensional interactions between real, monetary, and external sectors of the economy.

Research Gap: Although the econometric techniques, like OLS, VAR, and ARIMA, proved to be valuable but they failed to capture more complex and multi-dimensional interactions between real, monetary and external sectors. This paper fills that gap by constructing Composite Leading indicators to forecast business cycles for Pakistan’s economy

Design/Methodology/Approach: Based on quarterly data between 2011 and 2025, the model incorporates major indicators like narrow money, export volumes, household debt, household prices, policy rates, real effective exchange rates, and global economic conditions and constructs a CLI model through Partial Least Squares Structural Equation Modeling (PLS-SEM).

The Main Findings: The findings indicate that the short-run liquidity and export performance have the highest direct impact on the volatility of GDP whereas the medium-run financial imbalances, such as household debt and household prices, have a key countercyclical role in spurring downturns provided they are not sufficiently reduced. The model captures a considerable proportion of the variance of GDP growth, as non-linear.

Theoretical / Practical Implications of the Findings: This study demonstrates the practicality of a multi-horizon framework of CLI estimation using PLS-SEM in that it can be used to construct more efficient early warning mechanisms and can aid policymakers in the smarter macroeconomic planning against domestic and external shocks.

Originality/Value: This research offers a novel application of PLS-SEM in business cycle forecasting for a developing economy, providing new insights into non-linear macro-financial linkages rarely explored in Pakistan’s context.