Hull MA System

Hull MA has less lag than SMA or EMA, which means faster signals. However, faster isn't always better - it can mean more false signals in choppy markets. HMA is 'better' when you need quick trend identification and can accept occasional whipsaws. It's not better if you prefer smoother, more conservative signals. The best MA depends on your trading style and timeframe.

Start with 16 - Alan Hull's original default. This works well for swing trading on daily charts. If you're getting too many false signals, increase to 20-30. If signals are too slow, decrease to 9-14. For intraday trading, try 9-16 on 5-15 minute charts. Adjust based on your experience with each instrument.

Most charting platforms color-code HMA: green when rising, red when falling. The color change is the direction change signal. If your platform doesn't color-code, compare current HMA value to previous bar - if current is higher, it's rising; if lower, it's falling. A turn occurs when it changes from rising to falling or vice versa.

Yes, Hull MA works on any timeframe - from 1-minute to monthly charts. Shorter timeframes will give more signals and more noise. Longer timeframes give fewer, more significant signals. Match your period to your timeframe: shorter periods on longer timeframes, or vice versa. A 16-period HMA on a weekly chart is very different from a 16-period HMA on a 5-minute chart.

The simplest approach is yes - exit when HMA reverses. This keeps you on the right side of trends. However, in choppy markets, this might cause frequent entries/exits. Alternatives: use a trailing stop instead of waiting for full reversal, or require HMA to reverse AND price to cross HMA. Start with the simple approach and refine based on your results.

Single HMA gives signals on direction change - more frequent but more whipsaws. Dual HMA (crossover) gives signals when fast crosses slow - fewer signals but often more reliable because both MAs must agree. Single HMA is better for catching quick moves; dual HMA is better for filtering noise. Consider your preference for trade frequency vs. reliability.

Check the higher TF (weekly if trading daily, daily if trading intraday) HMA direction. Only take signals on your trading TF in the same direction as the higher TF. If weekly HMA is rising, only take daily HMA buy signals. This filters out counter-trend trades. Accept that you'll miss some counter-trend moves but gain reliability.

HMA's reduced lag comes at a cost - it's more sensitive to price movements, including noise. In ranging or choppy markets, this sensitivity causes the HMA to oscillate and give false turn signals. Solutions: use longer periods, add filters (ADX > 20, higher TF confirmation), or simply accept that some losses are the price of faster entries on the winners.

Yes, and it often improves results. Best combinations: HMA + RSI (confirm with momentum), HMA + ADX (only trade when ADX confirms trend), HMA + MACD (double momentum confirmation), HMA + price action (candlestick patterns at HMA turns). Don't over-complicate - one or two confirmation indicators is enough.

Several approaches: (1) Use longer period to filter noise, (2) Add ADX filter - only trade when ADX > 20-25, (3) Require price to be on the 'right' side of HMA (above for longs, below for shorts), (4) Use higher TF confirmation, (5) Accept whipsaws as cost of fast entries and manage with proper position sizing. No method eliminates whipsaws entirely.

Two main approaches: (1) ATR-based: Period = Base × (AvgATR / CurrentATR). High volatility = longer period. (2) Efficiency Ratio: ER = Abs(Direction) / Sum(Abs(Changes)). High ER (trending) = shorter period; low ER (choppy) = longer period. Implement by recalculating period each bar or at fixed intervals. Backtest to find optimal adaptation formula for your instruments.

Key considerations: (1) Test across multiple instruments and market conditions, (2) Use walk-forward optimization, not static backtest, (3) Track performance by regime (trending vs. ranging), (4) Test multiple periods, not just one optimized value, (5) Account for slippage and costs, (6) Compare to simple benchmark (buy-and-hold). A robust system works across instruments and conditions, not just one optimized scenario.

HMA divergence compares price highs/lows to HMA highs/lows directly, while RSI divergence compares price to a separate momentum oscillator. HMA divergence is often subtler because HMA already follows price closely. RSI divergence can appear before HMA divergence. Both are early warnings, not immediate signals. Use them together: RSI divergence + HMA divergence + HMA turn = high probability reversal.

Essential metrics: Win rate, average R-multiple (winner/loser), profit factor, max drawdown, Sharpe ratio. Also track by regime: win rate when ADX>25 vs <25, win rate by market (SPY, QQQ, individual stocks), performance by year/quarter. This reveals when your system works best and worst, allowing for adaptive position sizing or regime-based trading.

Yes. Approaches: (1) Train classifier to predict which HMA signals will be profitable - features include HMA slope, higher TF HMA, RSI, ADX, volume, recent volatility, (2) Use ML to determine optimal period based on market state, (3) Ensemble HMA with other signals via ML weighting. Caution: ML can easily overfit. Always use walk-forward validation and test on out-of-sample data.