International Journal of Analysis and Applications ISSN 2291-8639 (Online)

Research Article

Stochastic Discount Factors under Uncertainty States: Theory and Evidence from Equity Portfolio Pricing

  • Nguyen Hoang Ngoc Anh
  • Huynh Phuong Thuy

Abstract

This study examines whether uncertainty variables improve stochastic discount factor (SDF) performance in cross-sectional asset pricing beyond the Fama–French five-factor (FF5) model. Using monthly U.S. data from January 1985 to March 2026, including FF5 factors, the VIX index, the Economic Policy Uncertainty (EPU) index, and twenty-five size–book-to-market portfolios, the paper develops an uncertainty-augmented SDF framework. Pricing performance is evaluated through time-series regressions and the Gibbons–Ross–Shanken (GRS) efficiency test. The analysis further distinguishes between persistent uncertainty states and transitory uncertainty shocks by comparing uncertainty levels and first-difference measures. The results indicate that the benchmark FF5 model continues to exhibit economically meaningful pricing errors. Incorporating uncertainty variables improves pricing efficiency, particularly for value-weighted portfolios. VIX generates the greatest improvement, reducing the FF5 GRS statistic by more than 60% and eliminating rejection of the joint pricing restrictions in several specifications. Persistent uncertainty states contribute more strongly to pricing efficiency than short-run uncertainty shocks, highlighting the importance of state-dependent pricing mechanisms in modern asset pricing.

How to Cite

N. H. N. Anh, H. P. Thuy, Stochastic Discount Factors under Uncertainty States: Theory and Evidence from Equity Portfolio Pricing, Int. J. Anal. Appl. 24 (2026), 370. https://doi.org/10.28924/2291-8639-24-2026-370

BibTeX
@article{Anh2026,
  author  = {Anh, Nguyen Hoang Ngoc and Thuy, Huynh Phuong},
  title   = {Stochastic Discount Factors under Uncertainty States: Theory and Evidence from Equity Portfolio Pricing},
  journal = {Int. J. Anal. Appl.},
  volume  = {24},
  year    = {2026},
  pages   = {370},
  doi     = {10.28924/2291-8639-24-2026-370}
}