⌂ Research Index · tradebotsecrets research record · evidence-first, negatives included
Multiscale Oscillator — Direction vs. Volatility

BTC · 12 timeframes · damped, not pure Fourier

Multiscale Oscillator: Direction vs. Volatility

Each panel is one timeframe's real BTC history, fit as a damped oscillator — an exponentially-decaying envelope modulating a rotation at that timeframe's own natural frequency, drawn as a spiral moving forward through time (radius shrinks as it damps, exactly like a swinging pendulum losing energy, not a Fourier sinusoid that repeats forever at fixed amplitude). Panels run biggest/slowest (Monthly) to smallest/fastest (1-minute), per the operator's own framing.

Blue = direction (fit to raw signed returns) · Orange = volatility (fit to |returns|). Frequency: FFT power-spectrum peak of the real series. Damping: linear regression on the log-decay of the return series' own autocorrelation function. Drag any panel to rotate all of them.

Direction (raw returns) — no real structure found, shown for comparison Volatility (|returns|) — real, calendar-clock periods • drag any panel to rotate · each spiral shows ~3 of its own oscillation cycles
Timeframe Direction period (candles) Direction damping Volatility period (candles) Volatility half-life (candles) Volatility ACF(lag 1)