SVE Bands

SVE Bands builds Sylvain Vervoort's SVE bands, an envelope that denoises its input before any band math runs. Typical price — optionally rebuilt from Heikin-Ashi-style averaged candles — is exponentially smoothed and averaged again, and the bands offset the double-smoothed base by an exponentially averaged, range-flavored deviation measure, so a single erratic bar barely deflects the pair.

How to Trade the SVE Bands?

• Close beyond a band: a stretch past the smoothed envelope — fewer print than raw-band tags, each meaning more.
• Close back inside a band: the calmer mean-reversion trigger this construction is built for, aiming back at the centerline as its natural target.
• Centerline slope: the line tints with its direction — a low-flicker bias read, since the heavily smoothed center rounds through turns instead of jittering.

Band Width plots to the data window: narrowing marks compression, widening release. Alerts cover excursions, reversions, centerline crosses and slope turns — a natural fit beside the rest of the volatility toolkit.

SVE Bands Settings

• Input Series (default Typical Price): raw typical price, or Heikin-Ashi Typical to blend each candle with its neighbors first.
• Smoothing Length (default 8): length of the exponential smoothing and of the second average run on it — the double pass is the signature.
• Volatility Length (default 13): lookback of the average on the deviation measure.
• Deviation Multiplier (default 3.55): averaged deviation units the upper band sits above center.
• Lower Band Adjust (default 0.9): fraction of the upper offset used below; set 1 for symmetry.
• Style toggles cover slope coloring (on), the band gradient fill (on) and reversion markers (off).

Frequently Asked Questions

How do SVE Bands differ from Bollinger Bands?

Bollinger Bands wrap a raw moving average in standard deviations, so one wide-range bar deflects both center and width at once. SVE Bands smooth the input twice before centering and derive width from a range-based measure, trading immediacy for touches that filter one-bar spikes.

Why is the multiplier 3.55 instead of 2?

Because the offset is not a standard deviation. The measure compares typical price against neighboring lows across adjacent bars — a range gauge on a different scale — and the published 3.55 places the rails where genuine stretches, not routine noise, reach them.

What is the cost of all this smoothing?

Lag. Band excursions, re-entries and slope turns all land later than their raw-band equivalents — the deliberate trade for signals less prone to one-bar head fakes. Judge the balance on your own timeframe.