FRACTAL AI AGENT Docs ← Terminal

Weierstrass · Time Series PRO

Decompose price into trend, cycle and fractal noise.

Think of Weierstrass Time Series as a market-noise filter. It separates the chart into its smoother trend, repeating cycle and rough noise so you can tell whether a tradeable move is really present.

What this tool is for

You do not need to understand the mathematics to use it. Read it as a strategy selector: trend-following needs persistence, a swing strategy needs a usable cycle, and mean reversion needs evidence that moves tend to reverse. When the series behaves like a random walk, the most useful answer may be to avoid forcing a directional trade.

The quick decision guide

If you see…What it means and what to do
Hurst above 0.55Price has been persistent: moves are more likely to continue than immediately reverse. Favor trend-following setups that agree with the trend component.
Hurst below 0.45Price has been anti-persistent: extensions have tended to snap back. Favor mean-reversion ideas at extremes and be cautious chasing breakouts.
Hurst near 0.50Price is behaving close to random noise, with no strong directional edge. Wait for cleaner structure or use less exposure.
High roughness / poor fitThe path is jagged and the model does not describe it cleanly. Lower confidence in precise timing; simplify or stand aside.
Clear cyclePrice has a repeating rhythm with a measurable period in bars. Use the cycle phase to time swings instead of chasing every impulse.

How to read the result

OutputWhat it means
Hurst exponentThe main decision value: above 0.55 favors persistence, below 0.45 favors mean reversion, and around 0.50 means no clear directional edge.
Market regimeThe tool’s summary: trending, mean-reverting or random walk.
Fractal dimension / roughnessHow complex and jagged the series is. Greater roughness usually means a less clean path.
Trend componentDirection and strength of the smoother directional layer.
Cycle componentThe current phase, estimated period in bars and amplitude of repeating movement.
Fractal noiseThe character of the irregular residual after trend and cycle are separated.
Fit qualityHow well a Weierstrass-style fractal model describes the loaded series.
Noise signalA directional interpretation only when the measured structure supports one.

Practical workflow

  1. Read the regime and Hurst value before looking at the directional signal.
  2. If trend strength and persistence agree, favor trend-following ideas in that direction.
  3. If the cycle dominates, use its phase and period to time swings rather than chase breakouts.
  4. If roughness is high or fit quality is poor, reduce confidence in precise timing.
  5. If the tool reports a random walk or no edge, wait and re-check after the regime changes.
Why “Weierstrass”? The Weierstrass function is a classic example of a curve that remains jagged at every scale. The tool uses that fractal idea to describe how smooth structure and rough noise combine in price.
Quick interpretation: strong trend plus Hurst above 0.5 favors persistence; a dominant cycle favors swing timing; mostly noise favors patience. See Hurst Exponent for a deeper explanation.
No-edge results are useful: when Hurst is close to 0.5, the tool deliberately returns a neutral, low-confidence read instead of inventing a directional call.