Best Moving Averages for Trend Following Trading
Trend following operates on a singular premise: identify a market moving in a sustained direction and hold a position until the trend exhausts itself. Moving averages convert chaotic price data into a smooth, directional line, making them the backbone of nearly every trend-following system ever built. The choice of which moving average to use—and how to configure it—profoundly affects entries, exits, drawdown tolerance, and profitability. This article examines the specific moving averages that consistently deliver value to trend followers, the mathematics behind their behavior, and the practical rules that separate effective trend systems from curve-fitted noise.
Simple Moving Average (SMA): The Baseline Standard
The SMA calculates the arithmetic mean of closing prices over a fixed window. A 50-day SMA equally weights fifty sessions; a 200-day SMA equally weights two hundred. Its appeal lies in transparency—every trader sees the same line, creating self-reinforcing support and resistance. Institutional algorithms, pension fund rebalancing models, and risk-parity strategies frequently reference the 200-day SMA, which is why price reactions around it are unusually reliable.
For trend following, the classic configurations are the 50-day, 100-day, and 200-day SMAs. The 50-day tracks intermediate trends with acceptable lag; the 200-day defines the primary trend and acts as a regime filter. When price trades above a rising 200-day SMA, long-only trend systems should be active. When price falls below a declining 200-day SMA, long exposure should be reduced or eliminated. This single rule avoided the worst of the 2000–2002 and 2008 bear markets.
The SMA’s weakness is lag. A 200-day SMA responds slowly to fresh price information because old data carries equal weight. In fast-moving markets, an SMA-based exit can surrender 15–25% of a trend’s gains before generating a signal. Shorter SMAs (10, 20) reduce lag but produce whipsaws in consolidations.
Exponential Moving Average (EMA): Responsiveness Without Excess Noise
The EMA applies a weighting multiplier that gives recent prices greater influence. The multiplier formula is 2 ÷ (period + 1). A 20-period EMA weights the most recent close at approximately 9.5% versus 5% for a 20-period SMA. This makes the EMA turn faster at tops and bottoms, entering trends earlier and exiting earlier.
Trend followers often use the 20-period EMA for short-term signals and the 50-period EMA for intermediate confirmation. The 8-EMA, 21-EMA combination, popularized by swing traders, produces frequent but tradeable signals when filtered by a higher-timeframe trend indicator. The 13-EMA and 34-EMA pair, derived from Fibonacci sequences, appears in numerous institutional trend models.
The EMA’s faster response creates a vulnerability: in choppy, sideways markets, an EMA crossover strategy generates repeated small losses. This is why EMA-based trend systems require a volatility filter—such as an ATR threshold or an ADX reading above 20—to avoid trading during non-trending regimes.
Smoothed Moving Average (SMMA) and Hull Moving Average (HMA)
The SMMA, also called the Wilder Moving Average, assigns a weight of 1/period to the newest price and carries forward the previous SMMA value. This creates an exceptionally smooth line with very low noise. J. Welles Wilder used the SMMA (often mislabeled as an EMA variant) in his Parabolic SAR and RSI calculations. For trend following, a 100-period SMMA on a weekly chart captures multi-year secular trends with minimal false signals, though its lag is substantial.
The Hull Moving Average (HMA), developed by Alan Hull, attempts to solve the lag-versus-smoothness tradeoff. It calculates a weighted moving average of a weighted moving average, then applies a square root of the period as a final smoothing length. The result is a line that hugs price closely while remaining smooth. A 55-period HMA on a daily chart often signals trend reversals 3–5 bars before a comparable EMA. The tradeoff is reduced reliability; HMA flips direction more frequently, requiring disciplined stop placement.
The Triple Moving Average System
Many professional trend followers layer three moving averages of different speeds. A canonical setup: 10-period EMA (fast), 30-period EMA (medium), 60-period EMA (slow). A long signal occurs when the fast crosses above the medium, and the medium is above the slow. The slow average acts as a trend filter; the fast provides timing; the medium confirms. Exits trigger when the fast crosses back below the medium. This structure reduces whipsaws compared to a dual-cross system while entering earlier than a single 200-day SMA rule.
Backtests across futures, currencies, and large-cap equities show the triple-EMA system delivers positive expectancy in trending markets but suffers in range-bound conditions. Position sizing based on the distance between the fast and slow averages (wider spread = stronger trend = larger position) improves risk-adjusted returns.
Moving Average Envelopes and Keltner Channels
Trend followers do not rely solely on line crossovers. Envelopes place bands at fixed percentages above and below a moving average (e.g., 2% around a 20-day EMA). Keltner Channels use ATR (Average True Range) to set bands around an EMA. When price closes outside the upper band and the EMA is rising, the trend is strong. Pullbacks to the EMA that hold the band provide low-risk entries.
The 20-period EMA with 2.5 ATR bands is a robust trend-following tool on daily and 4-hour charts. A close beyond the band signals momentum; a retest of the EMA that fails to break the opposite band signals trend continuation.
The 200-Day SMA as a Regime Filter
No moving average discussion for trend following is complete without emphasizing the 200-day SMA. Research from Meb Faber and others demonstrates that a simple rule—hold equities when the 200-day SMA is rising, move to cash when it is falling—matches or beats buy-and-hold with substantially lower drawdown. The 200-day SMA does not generate frequent signals; it defines the battlefield. Trend followers should never take long swing trades when price is below a declining 200-day SMA, regardless of shorter-term crossovers.
Adaptive Moving Averages: KAMA and VIDYA
The Kaufman Adaptive Moving Average (KAMA) adjusts its smoothing constant based on the efficiency ratio—the net change divided by the sum of absolute changes. In trending markets, KAMA behaves like a fast EMA; in choppy markets, it slows to a crawl. The Variable Index Dynamic Average (VIDYA) uses a volatility index (like Chande Momentum Oscillator) to modulate responsiveness. Both reduce whipsaws, but their complexity makes them harder to interpret in real time. For systematic traders, KAMA with a 10-period efficiency ratio and 30-period slow length offers a balanced compromise.
Choosing the Right Moving Average for Your Timeframe
Scalpers on 1-minute charts benefit from a 9-EMA and 21-EMA crossover. Swing traders on daily charts should anchor to the 20-EMA, 50-SMA, and 200-SMA. Position traders on weekly charts use the 10-week EMA and 40-week SMA (equivalent to 50-day and 200-day on daily). The principle: match the moving average period to the expected duration of the trend you intend to capture. A 200-period moving average on a 5-minute chart tracks a trend lasting roughly 16 hours—useless for a position trader.
Backtesting and Parameter Sensitivity
Every moving average strategy must be tested for parameter sensitivity. A 50-200 crossover that works on the S&P 500 may fail on crude oil. Robust systems show profitability across a range of nearby parameters (e.g., 45-55 and 190-210) rather than a single optimized pair. Walk-forward analysis and out-of-sample testing prevent curve fitting. The best moving average is not the one with the highest historical return; it is the one that survives across markets, timeframes, and regimes.
Combining Moving Averages with Non-Correlation Filters
Moving averages alone cannot distinguish a trending market from a ranging one. Add the Average Directional Index (ADX): only take crossover signals when ADX is above 20 and rising. Add volume: a moving average crossover accompanied by above-average volume has higher follow-through. Add intermarket confirmation: a gold trend signal is stronger when the dollar index is breaking down. These filters reduce trade count but increase win rate and average win/loss ratio.
Final Operational Rules for Trend Followers
Use a 200-period SMA or EMA to define the primary trend direction. Use a 50-period moving average to confirm intermediate momentum. Use a 20-period EMA or HMA for entry timing. Never take a counter-trend crossover against the 200-period slope. Place stops at 1.5 to 2.5 ATR from entry, not at the moving average itself, to avoid noise stop-outs. Trail stops using a Chandelier Exit (3 ATR from highest high) or a faster moving average (e.g., 10-EMA) once the trade is in profit. Accept that moving averages lag—they confirm trends, they do not predict them. The profit comes from capturing the middle 60–70% of a trend, not the exact top or bottom.
