The Aligned Economic Index & The State Switching Model

Ilias AarabarXiv 2025 · DOI · arXiv

Introduction

Predicting the equity premium—the excess return of the stock market over risk-free assets—remains one of the most challenging tasks in financial economics. While numerous variables have shown historical predictive power, they frequently fail in out-of-sample tests, suffering from parameter instability and model uncertainty. This paper introduces a novel framework to address these issues by combining a state-switching model with the Aligned Economic Index ($E^{PLS}$).

Methodology

The Aligned Economic Index

To distill information from a vast array of economic indicators without introducing excessive noise, the study employs Partial Least Squares (PLS). Unlike the more common Principal Component Analysis (PCA), which identifies factors that explain the most variance among the predictors, PLS extracts components that maximize the covariance with future stock returns. This "alignment" ensures that the resulting index is specifically tuned for forecasting the target variable.

Time-series of the Aligned Economic Index and Equity Premium (1960-2017)

The State-Switching Framework

Recognizing that market dynamics shift across different economic regimes, the author implements a state-switching model. The model transitions between "up" and "down" states based on a real-time macroeconomic signal: the slope of the yield curve (the spread between 10-year and 3-month Treasury rates). A flat or inverted curve serves as a trigger for a "down state," allowing the model's intercept and slope coefficients to adapt to the changing economic environment.

Key Results

The empirical analysis, spanning from 1960 to 2017, demonstrates the significant predictive power of the proposed framework. The combination of the state-switching model and the $E^{PLS}$ predictor achieves an out-of-sample (OOS) $R^2$ of 4.12%, a remarkably high figure for monthly stock return forecasting.

Predictive performance comparison between One-state and State-switching models

As shown in the results table, the state-switching $E^{PLS}$ model substantially outperforms several benchmarks:

  • One-state $E^{PLS}$ model: 2.60% OOS $R^2$
  • PCA-based factors ($E^{PCA}$): 1.58% OOS $R^2$
  • Forecast combinations ($E^{FC}$): 2.64% OOS $R^2$

Crucially, the model maintains positive predictive power during economic expansions, challenging the conventional wisdom that stock returns are only predictable during recessions. The yield curve proves to be an effective real-time proxy for regime shifts, capturing changes in how investors process information.

Conclusion

The Aligned Economic Index and the state-switching model provide a robust solution to the twin problems of model uncertainty and parameter instability. By focusing on return-relevant information through PLS and adapting to economic states via the yield curve, this framework offers a significant advancement in equity premium forecasting. Future research may explore the impact of transaction costs on these strategies and the evolving nature of the yield curve signal in an era of unconventional monetary policy.

Summary updated

Abstract

A growing empirical literature suggests that equity-premium predictability is state dependent, with much of the forecasting power concentrated around recessionary periods (Henkel et al., 2011; Dangl and Halling, 2012; Devpura et al., 2018). I study U.S. stock return predictability across economic regimes and document strong evidence of time-varying expected returns across both expansionary and contractionary states. I contribute in two ways. First, I introduce a state-switching predictive regression in which the market state is defined in real time using the slope of the yield curve. Relative to the standard one-state predictive regression, the state-switching specification increases both in-sample and out-of-sample performance for the set of popular predictors considered by Welch and Goyal (2008), improving the out-of-sample performance of most predictors in economically meaningful ways. Second, I propose a new aggregate predictor, the Aligned Economic Index, constructed via partial least squares (PLS). Under the state-switching model, the Aligned Economic Index exhibits statistically and economically significant predictive power in sample and out of sample, and it outperforms widely used benchmark predictors and alternative predictor-combination methods.
BibTeX
@misc{aarab2025aligned,
  author = {Ilias Aarab},
  title = {{The Aligned Economic Index \& The State Switching Model}},
  year = {2025},
  eprint = {2512.20460},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2512.20460},
}
Figure