Volatility & bandsAcceleration Bands · AB
Volatility bands whose width scales with each bar's range — built to catch accelerating breakouts.
Works best in trending marketsEngine-computed on a fixed sample series
What it is
Acceleration Bands are a volatility envelope developed by Price Headley of BigTrends that plot a moving average of price flanked by an upper and a lower band. Unlike a simple percentage envelope, the width of these bands is not fixed — it expands and contracts with each bar's own high-to-low range, so the channel physically opens up when a market starts to move faster. The question the indicator answers is a practical one for a breakout trader: has price moved outside its recent volatility-adjusted range with enough conviction to signal a real acceleration, rather than just an ordinary wiggle? Because the bands react to the speed of price rather than a static percentage, they are designed to keep a trader in a move that is genuinely gathering pace and to keep them out of the aimless drift of a quiet market.
How it's calculated
The middle line is usually a 20-period simple moving average of the close, the same backbone many envelope indicators use. The upper band for each bar is the high multiplied by (1 + 4 × (High − Low) / (High + Low)), and the lower band is the low multiplied by (1 − 4 × (High − Low) / (High + Low)); both raw band values are then smoothed with the same 20-period average. The heart of the formula is the fraction (High − Low) / (High + Low), which is simply the bar's range expressed as a proportion of its price level, and the factor of 4 amplifies it. When a bar's range is wide relative to its price the multiplier grows and the bands flare outward; when the range is tight the bands pinch back toward the average. The result is a channel whose distance from price is proportional to recent volatility rather than a constant.
Reading it, step by step
A single close poking outside a band means very little on its own — intrabar noise pushes price beyond the envelope routinely, and Headley was explicit that these one-bar pokes are traps. The signal he built the tool around is price closing outside the same band for two or more consecutive bars, which he read as confirmation that a genuine acceleration is underway. Once price is riding along or hugging the outer band, that persistent contact is the sign of a strong directional run rather than a reason to fade. When price finally closes back inside the bands and then reaches the opposite band, the acceleration has spent itself. Watching the bands' width also tells you something: a sudden flaring open often coincides with the start of the very move you want to catch.
Best timeframes and settings
The classic setting is a 20-period lookback on daily bars, which is where Headley originally applied the tool for swing and position trades. The same default translates cleanly to intraday charts — a 20-bar Acceleration Band on a 5-minute or 15-minute chart works for day traders hunting momentum breakouts — but the shorter the timeframe the more single-bar noise you must filter. Shortening the lookback below 20 makes the average and bands hug price more tightly, producing earlier but far more frequent and less reliable signals; lengthening it past 20 smooths the channel and demands a bigger, more committed move to trigger. The multiplier factor of 4 is the other lever, though most traders leave it alone; raising it widens the bands and requires a more violent breakout, while lowering it makes the tool trigger-happy. As a rule, keep the two-consecutive-bar confirmation regardless of timeframe, because that rule is what separates the tool from a naive envelope.
When and where to use it
Acceleration Bands are fundamentally a trending-market and breakout tool; they earn their keep when a market is coiled and about to expand, or already running. They work across liquid asset classes — index futures, large-cap equities, major forex pairs and liquid crypto — anywhere clean range expansions occur. In a persistently rangebound, low-volatility market the two-bar breakout rarely triggers, and when it does the failure rate climbs, so the honest use of the tool in a chop is to stay flat. They are at their best right after a period of contraction, when the bands have narrowed and a fresh expansion is more likely to mark the birth of a trend than its exhaustion. Avoid using them as a mean-reversion fade tool: fading a market that is closing outside the band for several bars is fighting exactly the signal the indicator was built to confirm.
Strategies that use it
The core Headley strategy is the acceleration breakout: go long when price closes above the upper band for two consecutive bars, stay in as long as price keeps closing above the middle line or riding the upper band, and exit when price closes back below the middle line or tags the lower band. The mirror applies for shorts using the lower band. A second, more conservative approach uses the bands as a squeeze-and-release filter: wait for the bands to contract to a multi-week narrow, then take the first two-bar break in whichever direction it resolves, treating the opposite band as the trailing stop. A third approach uses the middle line as a trend filter and only takes upper-band breaks while the 20-period average itself is rising, which screens out counter-trend pokes. In every version the two-bar close is the trigger and the opposite band or the midline is the objective exit.
Combining it with other indicators
Because Acceleration Bands only measure volatility-adjusted price extension, they pair naturally with a trend-strength or momentum gauge that confirms the break has force behind it. ADX above 25, or a rising MACD histogram, gives independent evidence that the acceleration is a real trend rather than a one-off spike. Volume is the other classic partner: a two-bar band break on expanding volume is far more trustworthy than one on fading volume, since genuine acceleration should draw participation. Some traders overlay a slower moving average or an anchored VWAP to define the larger trend and only trade band breaks in that direction. Pairing the bands with a simple support-and-resistance read also helps, because a breakout that clears a well-watched level and the band at the same time carries more conviction.
Where it fails
The signature failure is the single-bar head-fake — price closes outside the band once, sucks in breakout traders, then snaps back inside — which is precisely why the multi-bar confirmation exists and why abandoning it is the classic mistake. In a rangebound market the bands generate a stream of small false breaks in both directions that chew up capital through repeated whipsaws. Because the bands are anchored to a 20-period average, they lag: by the time two closes confirm the break, a fast move may already be extended, so chasing without a plan for the pullback leaves you buying near the top of the first leg. The bands also say nothing about the quality of a move — a break driven by a single news spike looks identical to one driven by steady accumulation. The remedy is discipline: demand the two-bar close, confirm with volume or trend strength, and size positions so the inevitable false breaks are survivable.
A worked example
Imagine a stock trading around $100 with the 20-period middle line flat at $100, an upper band at $103 and a lower band at $97, the bands recently pinched narrow after weeks of quiet. On Monday the stock closes at $103.40, its first close above the upper band; a disciplined trader waits rather than chasing. On Tuesday it closes at $104.80 — a second consecutive close above the band — which triggers a long entry near $105, with the middle line, now curling up toward $101, serving as the stop reference. Over the next several sessions price rides the upper band from $105 to $112 as the bands flare wider, confirming the acceleration is real and letting the trade breathe. The exit comes two weeks later when price finally closes back below the rising middle line at $110, banking roughly $5 of the $7 move while sidestepping the one-bar pokes that trapped less patient traders on the way up.