NON-CUSTODIAL·AES-256-GCM ENCRYPTION·BUILT ON FREQTRADE

[ Mean Reversion ]

Bollinger Bands Mean-Reversion: Fade the Lower Band

Viewing pinned version v5 · SPOT · 1h · USDT

A rule-based mean-reversion strategy using Bollinger Bands (20): enters when price closes below the lower band, exits at the middle band.

Return
-2.9%
Max drawdown
3.0%
Win rate
56.7%
Sharpe
-6.40
Sortino
-6.23
Calmar
-6.52
Profit factor
0.58
Trades
552
Backtest window: 2025-08-28 – 2026-06-08
Moderate overfit risk

Simulated backtest results on historical data, self-selected by the publisher — no orders were ever placed and no capital was ever at risk. A backtest can overfit to the past no matter what its robustness band says, and past performance does not predict future results. Live trading differs from simulation — slippage, fees, latency, liquidity and exchange outages all apply — and losses can exceed anything shown here. This is not investment advice or a suitability assessment: forking a strategy runs it with your own capital, on your own exchange account.

Forking copies “Bollinger Bands Mean-Reversion: Fade the Lower Band” into your own workspace after a free sign-up. It never runs automatically, and your exchange keys are never touched.

Report this listing →
This strategy implements a classic Bollinger Bands mean-reversion technique: it treats a close below the lower band as a statistically stretched, higher-probability dip, and exits once price reverts to the middle band (the basis moving average) rather than waiting for the upper band or a fixed profit target. How it works: - Entry: close price below the lower Bollinger Band (20-period, 2 standard deviations). - Exit: close price above the middle Bollinger Band (the 20-period SMA basis). - Timeframe: 1 hour, run across BTC/USDT, ETH/USDT, ADA/USDT, DOGE/USDT, DOT/USDT and LINK/USDT. This is an original implementation built with the platform's visual Strategy Builder for this listing. Bollinger Band mean-reversion is a well-known, publicly documented trading technique and no third-party code was used. The pinned backtest is the in-sample half of an 80/20 out-of-sample split covering roughly nine months of historical data -- see the Overview tab and the performance disclaimer for what a simulated, self-selected backtest does and does not tell you before forking.

Strategy structure

Indicator types and condition shape only — every threshold and tuned parameter is masked. Fork this strategy to see the real values.

BB
Entry: 1 condition (all of)
Exit: 1 condition (all of)
  • long:
    • entry:
      • children:
          • left:
            • field: close
            • type: PRICE
          • operator: lt
          • right:
            • field: lower
            • indicator: BB
            • type: INDICATOR
          • type: COMPARE
      • type: AND
    • exit:
      • children:
          • left:
            • field: close
            • type: PRICE
          • operator: gt
          • right:
            • field: middle
            • indicator: BB
            • type: INDICATOR
          • type: COMPARE
      • type: AND