Direct Answer
The fractal dimension of a price path measures how jagged it is. A value near 1 indicates a smooth, nearly straight path, and values closer to 2 indicate a rough, plane-filling path. Katz's estimator is a commonly used version.
Fractal Dimension: Formula, Meaning, Signals, Examples and How to Use It
How Is Fractal Dimension Calculated?
The fractal dimension of a price path measures how jagged it is. A value near 1 indicates a smooth, nearly straight path, and values closer to 2 indicate a rough, plane-filling path. Katz's estimator is a commonly used version.
Formula: Katz: D = ln(n) / (ln(n) + ln(d / L)), where L is the total path length, d the largest distance from the first point, and n the number of steps
Inputs:
- Close prices
- Window length
- Path length L and diameter d
Worked example. A path length of 30, a diameter of 20 and 20 steps give a Katz dimension of 1.157. The inputs are illustrative numbers, not market data.
Reading it. Lower values indicate a more directional, trending path and higher values a choppier one. Other estimators exist, for instance Higuchi's, and they do not give identical values.
Limits. The result depends on the estimator and on how price and time axes are scaled. Short windows give unstable values.
What is Fractal Dimension?
The fractal dimension of a price path measures how jagged it is. A value near 1 indicates a smooth, nearly straight path, and values closer to 2 indicate a rough, plane-filling path. Katz's estimator is a commonly used version.
How to use Fractal Dimension
Lower values indicate a more directional, trending path and higher values a choppier one. Other estimators exist, for instance Higuchi's, and they do not give identical values. A reading alone is never a buy or sell signal; it needs context from price structure, volume, volatility and the broader market regime.
Strengths
- A single number between 1 and 2 describes how rough the path is, with no assumption about direction.
- Katz's form uses only total path length and the maximum distance from the start, so it needs a few sums and one maximum.
- Reproducible when the formula and data convention are fixed.
Weaknesses and limitations
- Path length depends on the sampling interval, so finer data gives a longer path and a different dimension for the same market.
- A path that is nearly straight is pulled toward 1, so a strong trend and a quiet drift can look alike.
- Correlated indicators can create false confidence by repeating the same underlying information.
Editorial and risk note
This page is educational content, not individualized investment advice. It should not imply guaranteed prediction, accuracy or outperformance.
Frequently Asked Questions
Is Fractal Dimension a buy or sell signal?
No. Lower values indicate a more directional, trending path and higher values a choppier one. The result depends on the estimator and on how price and time axes are scaled. A trade built on it still needs a trigger, an invalidation rule, a position size and realistic execution assumptions.
What is the best setting for Fractal Dimension?
Fractal Dimension takes these inputs: close prices; window length; path length L and diameter d. There is no universal best setting for the tunable ones, so test them across instruments, regimes and out-of-sample periods and favor stable parameter regions over one historical winner.
Can Fractal Dimension be used by itself?
The result depends on the estimator and on how price and time axes are scaled. Short windows give unstable values. For that reason it is read next to measures of direction, participation, valuation, liquidity and risk, not alone.
Does Fractal Dimension work on every timeframe?
The calculation (Katz: D = ln(n) / (ln(n) + ln(d / L)), where L is the total path length, d the largest distance from the first point, and n the number of steps) can be run on any bar size, but behavior is not portable: a reading on a five-minute chart describes a very different market window from the same reading on a daily chart. Validate the exact timeframe before relying on it.
Why does Fractal Dimension give false signals?
Path length depends on the sampling interval, so finer data gives a longer path and a different dimension for the same market. A path that is nearly straight is pulled toward 1, so a strong trend and a quiet drift can look alike. Lag, noise and regime changes add to this, because market participants react to new information after the reading is calculated.
How should Fractal Dimension be backtested?
Rebuild Fractal Dimension from close prices using point-in-time data, write the rules before testing, include delisted securities where relevant, model realistic fills and costs, reserve a validation sample and compare with a simpler baseline.