Monthly Seasonality Testing with In-Sample and Out-of-Sample Ratings

Summary

This indicator studies monthly seasonality using separate in-sample and out-of-sample periods on monthly charts. For each calendar month, it calculates average open-to-close return and the share of observations that rose or fell. It then ranks average returns relative to the strongest and weakest months, reports a reliability percentage based on the historical frequency of gains or losses, and combines the rankings and reliability measures into a monthly rating. The rating gives twice the weight to the more recent out-of-sample period and is intended to indicate preferred direction and possible position size by month.

The author emphasizes that short samples contain few observations for each calendar month and that the current month must be set before running the analysis. The example of ten observations illustrates how reliability is calculated, while the suggested cutoffs and position sizes are illustrative rather than validated rules. The rating system is explicitly described as unproven and dependent on the chosen sample lengths. The indicator reports historical patterns, but the document supplies no performance results establishing that those patterns persist or predict future returns.

Key ideas

• The indicator compares monthly returns across separate in-sample and out-of-sample periods on monthly charts.
• For each calendar month, it reports average return and the historical frequency of gains or losses.
• Rank and reliability measures are combined into a directional score, with greater weight assigned to out-of-sample data.
• The author says short samples provide few observations per month and requires the current month to be set before analysis.
• The rating is unproven and sensitive to sample-period choices, so historical seasonality should not be treated as established predictive evidence.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.