Abstract
This paper investigates the evolving behavior of calendar anomalies (monthly effects) within the Portuguese stock market over a period spanning approximately 120 years. By employing a combination of sub-sample and rolling window analyses, we demonstrate that the performance of these anomalies fluctuated adaptively over time. Additionally, we apply the "Superior Predictive Ability" test to assess whether these anomalies present exploitable profit opportunities, factoring in data-snooping effects. The results for the full sample indicate significantly higher returns in January and lower returns in June and July, while the positive September effect appears to be historically concentrated in earlier decades of the sample. Sub-sample and rolling window analyses reveal that the strength and even the sign of several calendar effects vary across periods. However, bootstrap simulations suggest that once trading costs are considered, calendar-based strategies do not consistently outperform a buy-and-hold benchmark. Overall, the evidence supports the Adaptive Market Hypothesis as a more suitable explanation for the observed dynamics in the Portuguese stock market.
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
Recommended Citation
Lobão, Júlio and Costa, Ana C.
(2026)
"Calendar Anomalies and the Adaptive Market Hypothesis: New Evidence from a Historical Financial Dataset,"
American Business Review: Vol. 29:
No.
1, Article 13.
DOI: 10.37625/abr.29.1.287-308
Available at:
https://digitalcommons.newhaven.edu/americanbusinessreview/vol29/iss1/13
DOI
10.37625/abr.29.1.287-308
