Volatility & bands

Standard Error Bands · SE Bands

Bands around a linear-regression line at two standard errors — tight in trends, wide in chop.

Works best in trending marketsEngine-computed on a fixed sample series
14612196Price above Upper = strengthPrice below Upper = weaknessUpper acts as support / resistance
UpperMiddleLowerHow to read SE Bands on the chart — the callouts mark what to look for.

The formula

A least-squares line is fitted over the lookback to give the midline. The standard error measures how far price scatters around that line; it shrinks when price tracks the fit closely and grows when price is noisy. The bands sit two standard errors either side, so they narrow in an orderly trend and widen when it fragments.

Midline = Linear Regression line Standard Error = √( Σ(Close − Regression) squared ÷ (n − 2) ) Bands = Midline ± 2 × Standard Error
Worked example
BarCloseRegression(Close − Reg)²
1100.0100.50.25
2102.0101.50.25
3101.0102.52.25
4104.0103.50.25
5105.0104.50.25

Σ = 3.25, SE = √(3.25 ÷ 3) ≈ 1.04. Bands = 104.5 ± 2.08 → 106.58 / 102.42.

What Standard Error Bands are

Standard Error Bands are a volatility envelope built around a moving trendline rather than a moving average. The centre is a linear-regression curve — the mathematical best-fit straight line through the last N closes, updated bar by bar — and two bands are drawn a set distance above and below it. That distance is the standard error, a statistical measure of how tightly the actual prices cluster around the fitted line. In plain terms, the bands answer one question: is price marching in an orderly, straight-line trend, or is it scattering all over the place? When price hugs its own trendline the bands pull in tight, and when price starts whipping around they flare out — which is the reverse of how most traders expect volatility bands to behave.

How it is calculated

First a linear-regression line is fitted over the lookback window (Jon Andersen's original design used 21 bars), and the value of that line at the most recent bar becomes the midline, usually smoothed with a short 3-period average to reduce jitter. The standard error of the estimate is then computed: conceptually you measure the vertical distance from every close to the regression line, square those distances, average them with an adjustment for the degrees of freedom, and take the square root. That single number tells you the typical scatter of price around the fitted trend. The upper band is the midline plus two standard errors and the lower band is the midline minus two standard errors, and those band values are also smoothed by the same short average. Because the whole thing rests on a regression fit, the midline and bands are recomputed on every new bar.

Reading it, step by step

Start with the slope of the midline: rising means an uptrend, falling a downtrend, and flat a range — this is your directional read. Next read the band width, which is the real message of the tool. Narrow, tightly pinched bands mean price is tracking its regression line closely, the hallmark of a clean, high-quality trend worth staying with. Widening or flaring bands mean price is scattering away from the line, a warning that the trend is losing coherence and may be topping, bottoming, or dissolving into chop. A touch of a band during a narrow-band trend is usually just a pullback that snaps back toward the midline, not a reversal. The combination to watch for is a strong slope with tight bands that suddenly begin to widen while the slope flattens — that is the fingerprint of a trend running out of fuel.

Reading the signals on the chart

14612196
UpperMiddleLowerThe ▲/▼ marks flag where price most recently crossed the line — the cues a trend-follower would act on.

Best timeframes

  • Scalping1m – 5mjumpier midline
  • Day trading15m
  • Swing1h – 4hcleaner slope read
  • PositionDaily

The regression midline repaints as new bars arrive, so judge signals on closed bars and give the fit a long enough window to be stable.

Standard Error Bands vs other bands

SE BandsBollinger BandsKeltner Channels
MidlineRegression lineSMAEMA
Width fromStandard errorStd deviationATR
Narrow meansClean trendQuiet marketQuiet market
ReadsTrend qualityVolatilityTrend channel

Common price-action setups

How the signal typically plays out on the chart.

Ride the narrow trend

The midline slopes up and the bands stay tight — trade with the slope and hold while the bands remain narrow, exiting when they flare or the slope flattens.

Ride the slope
Trend continuation
Pullback to the band

In an orderly uptrend price dips to touch the lower band and turns back toward the midline — buy the pullback with a stop just below the band.

Buy the pullback
Bounce to midline
Bands flare — caution

After a run the bands suddenly widen, warning the trend is fragmenting — tighten stops or stand aside rather than adding, since the orderly move is breaking down.

Tighten stops
Trend fragmenting

Best timeframes and settings

The default 21-period regression with a 3-period smoothing and two standard errors is a sensible starting point across daily and 4-hour charts, where the tool is most at home for swing trading. Shortening the regression length makes the midline and bands more responsive and quicker to flag a change, but it also lets more noise through, so the bands flare on minor wiggles. Lengthening it produces a smoother, more deliberate read that ignores small pullbacks but reacts slowly at genuine turns. Intraday scalpers sometimes drop to a 10 to 14 length on 5-minute or 15-minute charts, accepting the extra flaring in exchange for speed. Position traders can push the length toward 34 or 50 on daily and weekly charts to isolate only the dominant trend. The two-standard-error setting is close to a natural container for most price action and is best left alone.

When and where to use it

Standard Error Bands are a trend tool first and foremost, so they shine in markets that actually trend — index futures, large-cap equities, and major currency pairs on daily and 4-hour charts. Their special value is as a trend-quality filter: they do not just tell you the direction, they tell you whether the trend is clean enough to trust. Use them to stay in a strong move while the bands are tight and to step aside when they balloon. They are far less useful in a persistently choppy, directionless market, where the bands stay wide and the midline meanders, giving you little edge. Avoid leaning on them for mean-reversion fades of the outer bands, because unlike Bollinger Bands a tight band here does not signal a coming breakout — it signals a healthy trend that is likely to continue.

Strategies that use it

The core strategy is trend-following with a quality gate: when the midline slopes up and the bands are visibly narrow, enter long on a pullback that tags the lower band and turns back toward the midline, placing a stop below the recent swing low. Hold the position for as long as the bands stay tight and the slope holds, and treat a sudden widening of the bands as your exit cue rather than waiting for a fixed target. A second approach is a warning-based exit overlay on any existing trend trade: keep your primary entry logic, but flatten or tighten stops the moment the Standard Error Bands flare and the midline flattens, since that combination reliably front-runs the end of a move. A third, more conservative play only takes new entries when the bands have contracted after a period of width, using the fresh tightening as confirmation that a new orderly trend has begun.

Combining it with other indicators

Because the bands measure trend quality but not participation, they pair naturally with volume or On-Balance Volume, which confirms whether a tight-band trend is backed by real buying. ADX is an excellent partner because it independently rates trend strength on a 0 to 100 scale, so an ADX above 25 alongside tight bands is a strong two-source agreement that the trend is worth trading. A momentum oscillator such as RSI or MACD adds a divergence check: if the bands are still tight but momentum is fading, you get an early hint the trend is hollowing out before the bands flare. Some traders overlay a longer moving average purely to define the higher-timeframe bias and only take band signals in that direction. The key is to let the bands judge quality while another tool judges strength or participation.

Where it fails

The most common trap is treating them like Bollinger Bands and expecting tight bands to precede a breakout — here tight bands mean the trend is already healthy, and the mistake leads traders to fade moves they should be riding. A second failure is repainting: because the regression line is refitted every bar, the most recent midline and band values can shift as new data arrives, so signals judged on an unclosed bar may vanish, and you must wait for the bar to close. In choppy markets the bands stay chronically wide and the midline whipsaws, producing a stream of low-quality direction changes. The tool also lags at sharp V-shaped reversals, because the regression needs several bars to reorient. Finally, on very thin or gappy instruments a single outlier close can inflate the standard error and flare the bands for reasons that have nothing to do with the trend.

A worked example

Imagine a stock grinding higher from 100 to 115 over three weeks on a daily chart. The 21-period regression endpoint sits at 114.2, and price has been tracking it closely, so the standard error is a mere 0.60; the bands sit at roughly 113.0 and 115.4, a tight one-point-plus envelope that tells you this is a clean, orderly advance you should be holding. A pullback tags 113.1 near the lower band and immediately turns back up — a textbook buy in a tight-band uptrend. Two weeks later price becomes erratic, chopping between 116 and 121 with sharp reversals; the standard error swells to about 2.0, and the bands widen to roughly 111 and 119. That flaring, combined with a midline that has flattened, is your warning that the orderly trend has fragmented, and it is the signal to take profits or tighten stops rather than add to the position.

Common mistakes

  • Expecting narrow bands to signal a coming breakout as with Bollinger — SE Bands behave oppositely.
  • Reading the newest values as fixed; the regression midline repaints, so trust closed bars.
  • Fading a band touch in a strong trend instead of trading with the midline slope.
  • Using a lookback so short that the line and bands become jumpy and meaningless.
  • Ignoring the midline slope, which is the actual trend signal, and watching only the bands.
  • Treating a band touch as a reversal rather than a pullback within the trend.