Using Approximate Entropy to Filter Trading by Market Predictability

Summary

The article presents Approximate Entropy (ApEn) as a measure of recurring structure in rolling sequences of closed-bar log returns. It explains how the calculation embeds returns into short pattern vectors, counts similar patterns using Chebyshev distance and a tolerance scaled to standard deviation, and compares recurrence at successive embedding lengths. Lower ApEn indicates more repeated structure; higher ApEn indicates less predictability over the measured window. The accompanying MQL5 calculator, indicator, and synthetic-data script are intended to implement and check this measure.

ApEn is proposed as a regime gate for an existing directional strategy, not as a standalone buy or sell signal: an EA can suppress entries when the reading exceeds a chosen threshold. The article recommends using completed bars and tuning window, tolerance, and thresholds to the instrument and trading costs. Its synthetic example checks that periodic data produces lower ApEn than pseudo-random data, but that does not establish trading profitability. The calculation is computationally expensive, can be biased on short samples, depends on parameter choices, and provides no significance test in the described implementation.

Key ideas

• Approximate Entropy measures how often patterns in a rolling return series recur as each pattern is extended by one observation.
• Low readings suggest serial structure, while high readings suggest that recent returns provide less predictive information.
• The article positions ApEn as a filter for directional entries rather than an independent trade signal.
• The described implementation uses standard-deviation-scaled tolerance and closed-bar log returns.
• Short samples, parameter sensitivity, quadratic computation cost, and the absence of significance testing limit interpretation.

Tags

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