Back to Docs
T
Traderoid

MarketEntropy Node

Information content and disorder measurement

StatisticalInformationChaos

Overview

Market Entropy measures the degree of order or chaos in price movements using Shannon Entropy from information theory. High entropy = random/chaotic price action (low predictability). Low entropy = ordered/organized patterns (high predictability). It quantifies market efficiency: perfectly efficient = maximum entropy; mispriced = low entropy.

When entropy is low, patterns are exploitable. When entropy is high, noise dominates. This helps traders adapt strategy complexity: in high-entropy (random) markets, use simple strategies; in low-entropy (ordered) markets, use complex feature-rich strategies for edge.

Formula & Calculation

Shannon Entropy
H = -Σ(p_i × log₂(p_i))
p_i = probability of price in bin i
Sum over all bins of histogram
Normalized Entropy
H_norm = H / log₂(n_bins)
Ranges 0-1: 0=ordered, 1=maximum disorder/random

Parameters

ParameterDefaultDescription
lookback100Periods for entropy calculation
bins20Price histogram bins

Common Use Cases

1. Market Predictability

Low entropy = more predictable patterns. Trade momentum/patterns. High entropy = random = tighten stops, reduce position size.

2. Regime Detection

Rising entropy = transition from ordered to chaotic. Falling entropy = transition from random to patterned (often precedes trends).

3. Model Complexity

Low entropy: Use complex ML models (they find patterns). High entropy: Use simple strategies (overfitting risk too high).

4. Position Sizing

Size inversely to entropy: High entropy (chaotic) = smaller positions. Low entropy (patterns clear) = larger positions.

Advantages & Limitations

Advantages

  • Measures predictability directly
  • Theoretically sound (info theory)
  • Guides strategy selection
  • Independent of price level
!

Limitations

  • Bin selection affects results
  • Slow to change (lookback lag)
  • No directional information
  • Requires adequate samples

Related Nodes