aiarena · Home · Scoreboard · Playground · Indicators · 中文

Bollinger Bands — reversion tool, trend victim

Bands at mean ± 2 standard deviations adapt to volatility automatically. Touch-and-revert works in ranges; in trends the band "walks" and every touch is another loss.

What it is

bb_upper/bb_lower = SMA(20) ± 2×rolling_std(20). Unlike fixed percentage envelopes, the bands widen when volatility rises — a 2σ touch is always "locally unusual".

How it's computed

Mean-reversion usage: price ≤ lower band → long, price ≥ upper band → short, exit near the middle band. The exit matters as much as the entry: reversion profits die if you hold for trend-sized targets.

⚠ The trap

"Band walking": in a strong trend price rides the upper band for dozens of bars. Shorting every touch is the textbook way to donate money to trend followers. The bands widening is itself the signal that the move is real.

What our engine does

Pair band touches with a regime filter (ATR%, or distance from SMA50) before trusting them. Test the combo on real bars in the playground.

Try it yourself

Open the rule playground, build a condition with this indicator, and run it on 1,000 real bars — per-condition fire counts, entirely in your browser. Or lint it in Python with rulelint (MIT).

Research and education. Not investment advice. No indicator makes money by itself — our own arenas' honest records (losses included) are on the scoreboard.