Skip to main content

How It Works

From Raw Data to
Actionable Insights

Detect hidden market cycles using Goertzel DFT, Bartels significance testing, and the Hurst exponent. Turn market noise into structure you can read.

[01] The Challenge

Markets are complex. Traditional tools fall short.

Markets Are Noisy

Price data contains countless overlapping signals, trends, and random fluctuations. Traditional analysis struggles to separate meaningful patterns from noise.

Cycles Are Hidden

Recurring market cycles exist at multiple timeframes, but they are invisible to conventional charting tools. You need specialized methods to reveal them.

[02] Our Solution

A systematic four-step process

Transform raw price data into statistically validated cycle projections.

01

Upload Your Data

Import OHLCV from any market: stocks, FX, crypto, commodities, indices, or 800,000+ FRED economic series. CSV upload or live provider fetch.

  • All timeframes (1m to monthly)
  • Live data: Yahoo, FRED, more
  • CSV upload for private datasets
02

Remove Market Noise

Three detrending methods (linear, HP filter, first-difference) strip the long-term trend and random noise so the spectrum sees the cyclic component clearly.

  • Linear, HP filter, or first-difference
  • Choose per analysis, no black box
  • Preserves the underlying cyclic structure
03

Detect Hidden Cycles

Goertzel DFT scans hundreds of candidate periods and surfaces the dominant ones. Bartels test then asks: is each cycle statistically real, or is it random noise dressed up as a pattern?

  • Goertzel DFT spectrum
  • Bartels p-value per cycle
  • Hurst exponent for regime classification
04

Identify Turning Points

Combine the cycles that pass the test into a composite wave projected forward. Convergence zones show where multiple cycles bottom or top together based on historical phase.

  • Pick which cycles to include
  • Composite wave projection
  • Convergence zones (Nest of Lows / Highs)

[03] Why FractalCycles

What sets us apart

Statistically Validated

Every detected cycle gets a Bartels p-value. You see the number, you decide what to trust. The math is shown, not hidden.

Static and Dynamic Engines

Pick how the composite behaves: Static locks magnitudes at analysis time. Dynamic re-runs the math fresh on every new bar so magnitude and phase breathe with the chart.

Full Control

You pick which cycles get weight in the composite. Toggle them on or off, crop the data, swap the detrender. The math runs your way.

[04] Who Uses FractalCycles

Built for serious investors, analysts, and researchers

Swing Investors

Identify potential cycle turning points to inform your analysis. Gain additional context on positions held for days to weeks.

Position Investors

Use longer-term cycle analysis to understand market rhythms. See where you are within larger cycle patterns spanning weeks to months.

Technical Analysts

Add a quantitative cycle dimension to your existing analysis. Complement chart patterns and indicators with statistically validated cycles.

Quantitative Researchers

Explore cycle phenomena across different markets and timeframes. Export data for further analysis in your own systems and models.

Run your first analysis in the next two minutes

Pull live S&P, BTC, gold, or upload your own CSV. If the math does not change how you read your charts, just walk away. Nothing lost.

Free tier is permanentNo credit card to startCancel any paid plan in two clicks