Crude Oil Bollinger Band

Standard setting (20 period, 2.0 standard deviation) works well for most conditions. Some traders prefer 2.5 SD for fewer but more reliable signals. Test on your timeframe - 15-minute charts are popular for intraday crude trading.

No! Band touch alone is not enough. You need: (1) Normal bandwidth state (not squeeze), (2) Reversal candle confirmation, (3) RSI oversold, (4) No strong downtrend (flat middle band). During strong downtrends, price can 'walk' along the lower band without bouncing.

A band walk occurs when price hugs one band during a strong trend instead of reverting. Trade WITH the walk: enter on pullbacks to middle band, not against the outer band. Exit when price closes on the opposite side of the middle band.

Calculate bandwidth: (Upper Band - Lower Band) / Middle Band × 100. If bandwidth is below 2% (or at multi-week lows), you're in a squeeze. Bands will look almost parallel and very close together. Expect a breakout soon.

No, Bollinger Bands show where price is relative to recent volatility - they don't predict direction. A squeeze tells you a big move is coming but not which way. Use other tools (momentum, trend, fundamentals) to determine direction.

Look for divergence at band extremes: At lower band, if price makes lower low but RSI makes higher low = bullish divergence (high probability long). At upper band, if price makes higher high but RSI makes lower high = bearish divergence (high probability short). This combination significantly increases reversal probability.

TTM Squeeze compares Bollinger Bands to Keltner Channels. When BB goes inside KC, it's a true squeeze (extreme compression). This is more reliable than BB squeeze alone. When BB expands outside KC (squeeze 'fires'), trade in the direction of momentum for high-probability breakout.

Squeeze (<2%): Don't mean revert, wait for breakout, smaller size. Normal (2-4%): Full mean reversion trading at extremes. Expanded (>4%): Band walk trades or exhaustion reversals only, wider stops. Always check bandwidth before deciding strategy.

Double BB (1 SD inner + 2 SD outer) creates zones. Beyond outer = reversal zone (highest probability). Between bands = trend zone. Inside inner = neutral (no trade). Use Double BB when you want clearer zone definition and better trade management levels.

1-hour BB determines bias. If 1-hour price is in upper half and trending up, favor 15-minute longs. Don't take 15-minute mean reversion shorts against 1-hour uptrend. Best trades align across timeframes. Skip conflicting signals.

Test 20+ combinations (periods 10-30, SD 1.5-3.0) on 2+ years of data. Use walk-forward optimization (optimize on 6 months, test on 3 months, repeat). Measure win rate, profit factor, max drawdown, Sharpe. Segment by bandwidth state and time of day. Choose parameters robust across periods.

Adaptive BB adjusts period based on current volatility regime. Use shorter period (15) in high volatility for faster response, longer period (25) in low volatility for stability. Calculate ATR percentile to determine regime. Adaptive BB catches turns faster in volatile conditions.

Components: (1) BB calculator (bands, bandwidth, %B). (2) State classifier (squeeze/normal/expanded). (3) Signal generator with rules. (4) Filters (RSI, volume, time, candle pattern). (5) Risk manager. (6) Execution interface. Backtest with slippage/commission. Walk-forward validate. Paper trade 6-8 weeks before live.

Monitor BB on both Brent (ICE) and WTI (NYMEX). Synchronised extremes (both at lower band) = high confidence. A breakout in one benchmark often leads the other. Divergence usually means the Brent-WTI spread is moving (or a USD/GBP currency driver). Synchronised squeeze = major move coming. Cross-market confirmation improves win rate 5-10%.

Options include: (1) Adaptive period (volatility-adjusted). (2) BB on bandwidth (precise squeeze detection). (3) Volume-weighted bands (emphasize significant levels). (4) Multi-SD bands (1, 2, 3 SD for zones). (5) Percentage bands (fixed % from middle). Test modifications on your specific market/timeframe before use.