Analysis and Forecast of Leading Sector Rotation in the US Stock Market for 2026
DOI:
https://doi.org/10.61173/1v3xdm79Keywords:
Sector rotation, Markov-switching VAR, Regime forecasting, Macroeconomic indicatorsAbstract
Sector rotation is central to equity allocation because relative industry performance changes with macro conditions, volatility, and investor risk appetite. This study examines the U.S. equity market with monthly data from January 2007 to December 2025 and models the joint dynamics of market variables and five clustered sector groups through a three-state MSIH(3)-VAR(1) framework. The sample combines the S&P 500, VIX, unemployment, credit spreads, an economic growth factor, and industry return clusters constructed from 49 Fama-French industries. Results show that the bull market regime dominates the sample, has the longest average duration, and is associated with lower volatility and stronger technology performance, whereas bear states are shorter and defensive sectors remain relatively resilient. Out-of-sample forecasts for January 2023 to March 2026 achieve high regime consistency but much weaker sector hit rates, implying that macro regimes explain broad market environments better than high-frequency leadership shifts. The findings support regime-switching models as useful tools for strategic allocation, risk monitoring, and structured discussion of sector rotation.
References
This study uses a three-state MSIH(3)-VAR(1) framework 2017, 19(2): 116-132. to connect market returns, macroeconomic variables, and [3] Skare M, Stjepanovic S. Measuring business cycles: A
clustered sector returns, then extends the information set review. Contemporary Economics, 2016, 10(1): 83-94. through March 2026 to produce a conditional forecast for [4] Tuaneh G L, Essi I D, Etuk E H. Markov-switching
April 2026 to April 2027. The forecast indicates that the vector autoregressive (MS-VAR) modelling (mean adjusted): aggregate market environment remains expansionary, but Application to macroeconomic data. Archives of Business
the internal leadership pattern does not stay fixed. Instead, Research, 2021, 9(10): 261-274. the projected path shifts from renewed strength in basic [5] Perlin M. MS Regress: The MATLAB package for Markov materials and energy to a later rotation toward cyclical regime switching models. 2012.
consumer and industrial groups in early 2027. This pro- [6] Kim C J, Piger J M, Startz R. Estimation of Markov regimejected sequence suggests that the coming year is more switching regression models with endogenous switching. St.
likely to feature internal reallocation under a still-stable Louis: Federal Reserve Bank of St. Louis, 2005. (Working Paper macro regime than a broad reversal of market state. In that No. 2003-015C, revised 2005) sense, the model is most useful as a medium-horizon envi- [7] Zhang F P, Zhang Y L, Xu Y X, et al. Dynamic relationship ronment and rotation framework rather than as a high-fre- between volume and volatility in the Chinese stock market: quency rule for picking the single best-performing sector Evidence from the MS-VAR model. Data Science and
each month. Its main value lies in organizing the direction Management, 2024, 7: 17-24. of rotation and the structural logic behind the forecast. [8] Zha H F, Ruan S M, Li W. The characteristics of the new Dean&Francis Meixi Jin dual-cycle development pattern and systemic financial risk based A-share investment conditions and sector rotation: From the
on TVP-SVAR and MS-VAR model analyses. Heliyon, 2024, perspective of behavioral finance. New Economic Research, 10: e34943. 2025-12-02. [9] Chen C Q. Analysis of the impact of China's new monetary [12] Adongo F A, Lewis J A J, Chikelu J C, et al. Principal policy tools on the term structure of treasury bond yields: Based component and factor analysis of macroeconomic indicators.
on dynamic Nelson-Siegel and MS-VAR models. Hangzhou: IOSR Journal of Humanities and Social Science, 2018, 23(7):
Zhejiang University of Finance and Economics, 2024. 1-7. [10] Yang B W, Zhang Y K. Is the impact of digital [13] Bazzi M, Blasques F, Koopman S J, et al. Time varying industrialization on macroeconomic growth persistent? A study transition probabilities for Markov regime switching models.
based on nonlinear MS-VAR model. Frontiers of Engineering Amsterdam: Tinbergen Institute, 2014. (Discussion Paper TI
Management and Technology, 2025, 44(3): 83-89. 2014-072/III) [11] Huang Q C, Qin J L. Analysis of the relationship between
Downloads
Published
Issue
Section
License
Copyright (c) 2026 by the authors.
This work is licensed under a Creative Commons Attribution 4.0 International License.
