Composite acceleration
— σ
SPX velocity regime
PACE accel / decel
How quickly S&P momentum is changing—not merely whether price is up or down. Daily, weekly, and monthly pace are measured independently, normalized against their own histories, then combined only after each is expressed in standard deviations.
1D · MAX FULL-SPAN RESOLUTION
Ten years cannot honestly be reconstructed from intraday candles. Weekly and monthly bars are aggregated from the same daily ES continuous-futures store used by
/apps2/basics.
Regime
—
SPX / ES reference
—
Extreme percentile
— %
Model controls
L smooths each timeframe · Final smoothing smooths the white composite · Z-score lookback changes the baseline, not smoothness.
Normalized acceleration history
Rolling σ · positive = acceleration · negative = deceleration
Composite
Daily
Weekly
Monthly
Not enough history for the selected settings.
Each extreme, measured in σ
One dot per breach episode · peak standard deviations from normal
No extremes at this threshold.
Extreme event ledger
—
| Peak date | Direction | Composite σ | Daily σ | Weekly σ | Monthly σ | SPX / ES | Episode |
|---|---|---|---|---|---|---|---|
| Waiting for data… | |||||||
Exactly what “acceleration” means here
A transparent transformation, not a forecast or trade signal.
Direction and magnitude stay separate
For each timeframe, X sets the return span, L averages that per-bar pace, and K measures how much the averaged pace changed versus K bars earlier. Dividing by K makes the result a true rate-of-change of velocity: points per bar² or percent per bar².
The rolling z-score then asks how unusual that acceleration is relative to the same timeframe’s preceding normalization history. The current observation is excluded from its own mean and standard deviation.
High positive σ means momentum is accelerating upward—or a decline is losing speed quickly. High negative σ means upward momentum is decelerating—or downside velocity is intensifying. Read it with the price context above the oscillator.
delta Δₓ(t) = C(t) − C(t−X)
pace v(t) = Δₓ(t) / X
average pace v̄(t) = SMAₗ[v(t)]
acceleration a(t) = [v̄(t) − v̄(t−K)] / K
normalized z(t) = [a(t) − μₚᵣᵢₒᵣ] / σₚᵣᵢₒᵣ
pace v(t) = Δₓ(t) / X
average pace v̄(t) = SMAₗ[v(t)]
acceleration a(t) = [v̄(t) − v̄(t−K)] / K
normalized z(t) = [a(t) − μₚᵣᵢₒᵣ] / σₚᵣᵢₒᵣ