Bollinger Bands Backtesting: Test Squeeze and Bounce Strategies

Bollinger Bands are one of the most versatile technical indicators available to crypto traders. They adapt to volatility, identify squeeze conditions that precede explosive moves, and highlight potential mean-reversion opportunities when price touches the bands. But which Bollinger Bands strategy actually works on your chosen trading pair? The only way to know is rigorous backtesting against historical data.

For foundational knowledge on Bollinger Bands trading approaches, read our Bollinger Bands trading strategy guide. This article focuses on the backtesting process: setting up BB backtests on Sentinel, optimizing period and deviation parameters, testing squeeze detection strategies, and combining Bollinger Bands with complementary indicators.

Bollinger Bands Refresher for Backtesting

Bollinger Bands consist of three lines calculated from price data:

Middle Band: A simple moving average (default: 20-period SMA)
Upper Band: Middle band + (standard deviation x multiplier, default: 2)
Lower Band: Middle band - (standard deviation x multiplier, default: 2)

The bands expand during high volatility and contract during low volatility. This dynamic behavior creates two primary trading opportunities:

Bounce/Mean-Reversion: Price touching or crossing the outer bands tends to revert toward the middle band.
Squeeze/Breakout: When bands contract to unusually narrow width, a significant directional move often follows.

Each strategy type produces fundamentally different backtest results and requires different optimization approaches.

Setting Up a Bollinger Bands Backtest

Step 1: Choose Your Strategy Type

Before configuring parameters, decide which BB strategy you are testing:

Bounce Strategy (Mean-Reversion):

Buy when price touches or crosses below the lower band
Sell when price touches or crosses above the upper band
Works best in ranging, sideways markets

Squeeze Strategy (Breakout):

Identify when bandwidth contracts below a threshold (squeeze)
Enter in the direction of the breakout when bands expand
Works best at the transition between range and trend

These are opposing philosophies. Do not mix them in a single backtest. Test each independently, then compare performance metrics.

Step 2: Configure the Bollinger Bands Entry Block

In Sentinel's block-based builder, add a Bollinger Bands entry block. Core parameters:

Period: Number of candles for the SMA and standard deviation calculation (default: 20)
Standard Deviation Multiplier: How many standard deviations from the SMA define the bands (default: 2.0)
Signal Type: Band touch, band cross, or squeeze detection

For a bounce strategy, configure the signal to trigger when price crosses below the lower band (long entry) or above the upper band (short entry).

Step 3: Select Timeframe and Pair

Bollinger Bands work across timeframes, but behavior differs:

15-minute to 1-hour: Frequent band touches, many mean-reversion opportunities, higher noise
4-hour: Good balance, popular for BB strategies on crypto
Daily: Fewer signals, higher quality, best for swing trading
Weekly: Very few signals, useful only for position trading context

Start with 4H or daily on BTC/USDT or ETH/USDT for your baseline backtest.

Step 4: Configure Exits

Bounce strategy exits:

Middle band exit: Close the position when price returns to the SMA (middle band). Conservative, captures the mean-reversion move.
Opposite band exit: Hold until price reaches the opposite band. Aggressive, captures the full range but risks reversals.
Fixed stop-loss: Set a stop at -3% to -5% below the lower band touch to protect against breakdowns.
Time-based exit: Close after N candles if the mean-reversion has not occurred.

Squeeze strategy exits:

Trailing stop: Use a 2-4% trailing stop to ride the post-squeeze breakout.
Band expansion exit: Close when bandwidth reaches a target expansion level.
Fixed take-profit: Set based on the expected magnitude of post-squeeze moves.

Step 5: Set Realistic Costs

Commission: 0.1%, slippage: 0.05%, starting capital: 10,000 USDT. Run the backtest and analyze results.

Backtesting the Bounce Strategy

Expected Results

Bollinger Band bounce strategies are mean-reversion by nature:

Win rate: 55-70%. Price does revert to the mean most of the time.
Average win/loss ratio: 0.6-1.2. Wins are moderate since you are targeting the middle band, not a trend continuation.
Profit factor: 1.2-2.0. Solid but not spectacular.
Max drawdown: -10% to -25%. Drawdowns occur when price breaks through the band and trends strongly.
Sharpe ratio: 0.8-1.5. Consistent, moderate risk-adjusted returns.

The biggest risk is a strong trend where price runs along the lower or upper band without reverting. During the 2022 crypto bear market, BTC rode the lower Bollinger Band for weeks, punishing bounce buyers repeatedly.

Identifying When Bounces Fail

Examine your backtest's losing trades. Most losses in bounce strategies occur during:

Strong trends: Price "walks the band" instead of bouncing
High-impact news events: Fundamentals override technical levels
Low liquidity periods: Thin order books cause exaggerated moves

Add a trend filter to reduce these losses. An EMA or ADX condition can prevent bounce entries during established trends. See our EMA crossover backtesting guide for setup details.

Backtesting the Squeeze Strategy

Detecting Squeezes in Backtesting

A Bollinger Band squeeze occurs when the bandwidth (distance between upper and lower bands relative to the middle band) contracts to an unusually low level. This indicates a period of low volatility that typically precedes a significant price move.

Bandwidth formula:

Bandwidth = (Upper Band - Lower Band) / Middle Band

To detect a squeeze, you need to define a threshold. Common approaches:

Absolute threshold: Bandwidth below a fixed value (e.g., 0.04 or 4%)
Relative threshold: Bandwidth at its lowest level in the last N periods (e.g., lowest in 120 candles)
Percentile threshold: Bandwidth below its 10th percentile over the lookback period

In Sentinel's block builder, configure the Bollinger Bands block for squeeze detection. The signal fires when bandwidth contracts below your threshold and then price breaks out of the compressed range.

Determining Breakout Direction

The squeeze tells you that a big move is coming. It does not tell you the direction. You need a directional filter:

MACD direction: Enter in the direction of the MACD histogram slope during the squeeze. See our MACD backtesting guide for MACD configuration.
RSI position: If RSI is above 50, bias is bullish. Below 50, bearish.
Volume analysis: Monitor which side (buy or sell) shows increasing volume during the squeeze.
Price position relative to middle band: If price is above the SMA when the squeeze fires, bias is bullish.

Combine the squeeze detection with a directional filter using Sentinel's AND composite logic.

Expected Squeeze Backtest Results

Squeeze strategies behave differently from bounce strategies:

Win rate: 45-55%. Not every squeeze leads to a tradable breakout. Some produce false starts.
Average win/loss ratio: 2:1 to 4:1. When squeezes work, they produce outsized moves.
Profit factor: 1.5-3.0. High reward-to-risk when executed correctly.
Trade frequency: Low. Squeezes are relatively rare events, especially with tight bandwidth thresholds.
Max drawdown: -10% to -20%. Individual false breakouts are contained by stop-losses.

The main challenge is generating enough trades for statistical significance. On a daily timeframe, you might only get 10-15 squeeze signals per year per pair. Consider testing across multiple pairs to accumulate more data points.

Period and Deviation Optimization

Period Optimization

The default 20-period Bollinger Bands are widely used but not necessarily optimal for crypto:

| Period | Behavior | Best For |

|---|---|---|

| 10-14 | Tight bands, more band touches | Short-term trading, frequent signals |

| 15-20 | Standard behavior | General purpose |

| 21-30 | Wider bands, fewer touches | Swing trading, higher conviction signals |

| 30-50 | Very wide bands, rare touches | Position trading, major reversals only |

Use Sentinel's grid search to test periods from 10 to 40 in steps of 2.

Deviation Multiplier Optimization

The standard deviation multiplier controls how far the bands sit from the middle:

|---|---|---|---|

| 2.5 | Wide | Low | High |

For crypto's higher volatility, multipliers of 2.0 to 2.5 typically perform best. Multipliers below 1.5 generate too many false signals. Multipliers above 3.0 rarely trigger.

Combined Grid Search

Sweep both parameters simultaneously:

Period: 10 to 35, step 5 (6 values)
Deviation: 1.5 to 3.0, step 0.25 (7 values)

This produces 42 combinations. Examine the results for stable performance clusters where changing parameters slightly does not dramatically alter results.

Combining Bollinger Bands with Other Indicators

BB + RSI: Double Mean-Reversion Confirmation

The most natural Bollinger Bands combination:

Long: Price below lower band AND RSI < 30
Short: Price above upper band AND RSI > 70

Both indicators independently signal an extreme. When they agree, the probability of mean-reversion is significantly higher. This combination typically improves win rate by 5-15% compared to standalone BB signals. See our RSI backtesting guide for optimal RSI thresholds.

BB + Volume: Conviction Filter

Require above-average volume on band touches:

Band touches with high volume indicate genuine selling/buying pressure that has exhausted itself (capitulation)
Band touches with low volume may be insignificant drift rather than actionable extremes

BB Squeeze + MACD: Direction Confirmation

As discussed in the squeeze section, combine bandwidth contraction with MACD direction for high-conviction breakout trades. This is one of the strongest multi-indicator combinations in crypto strategy building.

BB + EMA Trend Filter

Add a 200 EMA trend filter:

Only take lower-band bounces (longs) when price is above the 200 EMA
Only take upper-band fades (shorts) when price is below the 200 EMA

This prevents the most dangerous BB trades: buying bounces in a bear market downtrend.

Real-World Considerations

Bollinger Bands and Crypto Volatility Regimes

Crypto alternates between periods of compressed volatility (accumulation/distribution) and explosive volatility (trends). Bollinger Bands naturally adapt to this because the standard deviation calculation expands and contracts the bands. However, the lookback period determines how quickly the bands adapt.

Shorter periods (10-14): Bands adapt quickly to volatility changes. Better for capturing short-term regime shifts but more whipsaw.
Longer periods (25-40): Bands change slowly. Better for identifying major regime transitions but may lag during rapid volatility shifts.

Backtest both and compare how quickly each adapts to the volatility regimes present in your data. The equity curve will reveal which parameter set handles regime transitions better.

Limitations of BB Backtesting

Squeezes are rare events. Even on 4H data over 2 years, you may get only 20-40 squeeze signals. This limits statistical confidence.
Band walk is unpredictable. No backtesting can fully capture the psychological pressure of watching price ride the lower band for days without bouncing.
Standard deviation assumes normal distribution. Crypto returns are not normally distributed. Extreme moves happen more often than the bands predict.

Frequently Asked Questions

What Bollinger Bands period works best for crypto?

Backtesting consistently shows that periods between 15 and 25 work well for crypto on 4-hour and daily timeframes. The default period of 20 is a solid starting point. Grid search results typically show a stable performance zone in this range rather than a single optimal value.

Should I use 2.0 or 2.5 standard deviations for crypto?

Crypto is more volatile than traditional markets, so 2.0 to 2.5 is the recommended range. A 2.0 multiplier generates more signals but more false ones. A 2.5 multiplier is more selective, with higher win rate but fewer opportunities. Backtest both and compare Sharpe ratios.

How do I detect a Bollinger Band squeeze in Sentinel?

Use the Bollinger Bands block with squeeze detection mode. Set a bandwidth threshold (e.g., bandwidth below its 20th percentile over the last 100 candles). Combine with a directional indicator like MACD or RSI to determine entry direction.

Can I combine bounce and squeeze strategies?

Yes, but run them as separate strategies (or separate bots), not in a single combined strategy. They have opposite philosophies: bounces bet on mean-reversion while squeezes bet on breakouts. Combining them in one backtest would produce confusing, unreliable results. Test and optimize each independently, then allocate capital between them.

Do Bollinger Bands work on altcoins?

Bollinger Bands work on any asset with sufficient liquidity and price history. However, low-cap altcoins often have thin order books and erratic price action that can pierce bands without meaningful reversal. Stick to pairs with at least moderate daily volume for reliable BB signals.

Ready to discover whether Bollinger Band bounces or squeezes work best for your trading? Sign up for Sentinel Bot and backtest both strategies with grid search optimization. Visual block builder, professional metrics, and no coding required.

Disclaimer: This article is for educational purposes only and does not constitute financial advice. Backtesting results do not guarantee future performance. Cryptocurrency trading involves significant risk of loss. Always do your own research and never trade with money you cannot afford to lose.