Entropy

What it measures

Entropy, borrowed from information theory (Shannon entropy), measures how random or unpredictable a security's recent price movements have been. Low entropy means price action has been following a more ordered, patterned structure — the kind of behavior technical analysis is generally trying to identify and exploit. High entropy means price action has been closer to pure randomness, where historical patterns carry less predictive value.

How readings are interpreted

Low entropy suggests recent price behavior has structure that pattern-based and trend-following methods are more likely to have a genuine edge on. High entropy suggests price has been behaving close to a random walk, where technical patterns are more likely to be coincidental noise rather than genuine signal.

Conventional levels

• Low entropy: more ordered, structured price behavior — historically a more favorable environment for technical/pattern-based strategies
• High entropy: closer to random price behavior — technical signals are less trustworthy in this state
• Entropy shifting from low to high (or vice versa) signals a change in the underlying character of price action, worth noting even without a specific directional read attached

Where it works, and where it does not

A regime-characterization tool in the same broad family as the Choppiness Index, though built from a genuinely different (information-theoretic) foundation. Most useful as a meta-signal about how much to trust other technical signals right now, rather than a trading signal itself.

Commonly read alongside

• Choppiness Index: Both are regime-classification tools from different mathematical angles — agreement between them strengthens confidence in the regime read
• Any pattern-based or trend-following strategy: Entropy can serve as a gating filter — trust signals more when entropy is low, discount them more when entropy is high
• RWI: Conceptually related (both ultimately ask "how random is this"), from different statistical approaches

Known limitations

• Like Choppiness, gives no directional information on its own — purely a "how trustworthy is technical analysis right now" read
• The information-theoretic foundation is less intuitive to reason about than more familiar price-based indicators
• Requires a reasonably sized rolling window to be statistically meaningful, so it changes relatively slowly

In practice

• Use as a confidence gate on other signals: weight technical signals more heavily when entropy is low, less heavily when entropy is high
• Particularly useful for systematic/algorithmic approaches that want an explicit, quantified "is this a good environment for my strategy" check
• Combine with Choppiness and RWI for the fullest regime picture, since each captures a slightly different statistical angle on the same underlying question
• Not intended as a standalone signal — it functions as a meta-layer sitting above the entry and exit logic itself