What Trend Following Actually Is

Trend following is a momentum-based methodology that seeks to capture sustained directional price moves across markets. Instead of predicting reversals or forecasting economic data, a trend follower reacts to price itself. The core premise is simple: markets that have moved tend to keep moving in the same direction until they don’t. Your job is not to be right about why a market moves, but to be positioned when it does and to exit when the move exhausts.

This distinction matters because most beginners fail by trying to outsmart the market. They buy oversold conditions, short overbought conditions, and interpret every pullback as a reversal. Trend following inverts that instinct. You buy strength, sell weakness, and accept that you will never catch the exact top or bottom. The edge comes from asymmetric payoff distribution: many small losses offset by a few large winners.

The Statistical Foundation

Trend following rests on two observed market properties. First, prices exhibit positive autocorrelation over medium-to-long horizons. Academic research on time-series momentum, notably Moskowitz, Ooi, and Pedersen’s 2012 paper, documented persistent return continuation across futures markets going back decades. Second, trends cluster. Volatility and directional persistence arrive in regimes, not uniformly.

These properties do not guarantee profits on any single trade. They only suggest that a systematic, rules-based approach to following trends has historically produced positive expectancy when applied across a diversified basket with disciplined risk control. That expectancy is thin per trade, which is why execution discipline and cost control determine whether the edge survives.

Step 1: Choose Your Market Universe

You cannot trend-follow a single market and expect smooth results. Trends are episodic. A diversified universe increases the probability that something is trending at any given time.

For beginners, practical universes include:

• Futures: equity indices (ES, NQ), rates (ZN, ZB), currencies (6E, 6J), metals (GC, SI), energy (CL, NG), agriculture (ZC, ZS)
• ETFs: SPY, QQQ, TLT, GLD, USO, EFA, EEM
• Crypto: BTC, ETH (higher volatility, 24/7 execution)
• Stocks: liquid large caps with sufficient average daily volume

Start with 15–30 instruments. Fewer than 10 creates too much concentration risk; more than 50 demands infrastructure you likely don’t have yet. Avoid illiquid markets — slippage destroys trend-following edges faster than almost anything else.

Step 2: Select a Timeframe

Trend following works across daily, weekly, and intraday timeframes, but the daily chart is the beginner’s sweet spot. It offers enough signal frequency to build a track record without the noise and screen time of intraday trading. Weekly charts produce fewer, longer trades and are even more forgiving of execution errors.

Avoid minute charts initially. Intraday trend following requires low-latency execution, tight spreads, and significant screen time. The signal-to-noise ratio is poorer, and transaction costs consume a larger fraction of gross profits.

Step 3: Define Your Trend Filter

You need an objective, mechanical rule that answers: is this market in an uptrend, downtrend, or no trend?

Common filters:

• Moving average slope: 50-day SMA rising = uptrend, falling = downtrend
• Price vs. moving average: close above 200-day SMA = long bias, below = short bias
• Donchian channel breakout: new 20-day high = uptrend signal, new 20-day low = downtrend
• ADX: above 25 suggests trending conditions, below 20 suggests range

A robust beginner setup combines two: price above the 200-day SMA and a 50-day breakout. This dual confirmation reduces whipsaw frequency while preserving participation in major moves.

Step 4: Establish Entry Rules

Entries must be unambiguous and testable. Three workable templates:

1. Breakout entry: Enter long when price closes above the highest high of the prior N days (commonly 20, 50, or 55). Enter short on the mirror condition.
2. Moving average cross: Enter long when the fast MA crosses above the slow MA (e.g., 20/50 or 50/200). Slower signals, fewer trades.
3. Pullback entry: In an established uptrend, enter when price retraces to a moving average and resumes higher. More entries, more complexity.

Breakout entries are the cleanest starting point. They are objective, easy to backtest, and align with the underlying premise that new highs often precede further highs.

Step 5: Set Position Sizing

Position sizing determines survival. Most beginners focus on entries and neglect sizing, which is why a handful of losses wipes them out.

Use volatility-based sizing. Calculate the Average True Range (ATR) over 14 or 20 periods. Then size each position so that a 1 ATR adverse move equals a fixed percentage of account equity — typically 0.5% to 1%.

Formula: Position size = (Account equity × Risk per trade) ÷ (ATR × Point value)

Example: $50,000 account, 1% risk = $500. If ATR on an ETF is $2.00, you buy 250 shares. This normalizes risk across instruments with different volatilities, so a position in crude oil and a position in gold contribute equally to portfolio risk.

Step 6: Place Stops

Every trade needs a predefined stop. Trend followers use one of three:

• ATR stop: 2–3 ATR below entry
• Channel stop: exit long on a close below the N-day low (e.g., 10-day)
• Chandelier stop: highest high since entry minus 3 ATR

The channel stop is the most elegant because it tightens as the trend ages and never widens. Combine it with a hard catastrophic stop at 3 ATR for gap protection.

Never move a stop further away. Moving stops to avoid being stopped out is the single most common cause of catastrophic losses among discretionary trend traders.

Step 7: Define Exit Rules

Exits come in two forms:

Profit exits: Trail a stop behind price as the trend develops. The 10-day or 20-day low (for longs) is a classic trailing method. Alternatively, exit on a close below the 50-day MA.

Time or regime exits: If a trade hasn’t moved in X bars, exit and reallocate. If the trend filter flips (price crosses back below the 200-day MA), exit regardless of stop level.

Winners in trend following often run for months. Do not cap upside with arbitrary profit targets. The entire edge depends on letting a few trades produce outsized returns.

Step 8: Build a Portfolio-Level Risk Framework

Individual trade risk is not enough. You need aggregate limits:

• Max risk per trade: 0.5%–1% of equity
• Max correlated exposure: cap total risk in highly correlated instruments (e.g., all equity indices combined) at 2%–3%
• Max portfolio heat: total open risk across all positions should not exceed 5%–6%
• Max positions: 10–20 concurrent positions depending on account size

Correlation analysis prevents the illusion of diversification. Long ES, NQ, YM, and RTY is one trade, not four.

Step 9: Track and Journal Every Trade

A trading journal must capture:

• Date, instrument, direction
• Entry price, stop price, target logic
• Position size and dollar risk
• Exit price, exit reason
• MAE (maximum adverse excursion) and MFE (maximum favorable excursion)
• Screenshot of the chart at entry and exit

Review weekly. After 50–100 trades, calculate win rate, average win, average loss, expectancy, and maximum drawdown. Expectancy formula: (Win% × Avg Win) − (Loss% × Avg Loss). If expectancy is negative after 100 trades and rules were followed, the system needs revision.

Step 10: Backtest Before You Trade Live

Manual backtesting on 5–10 years of data across your universe is the minimum. Record every signal, entry, exit, and result in a spreadsheet. Then compute:

• Total return and CAGR
• Maximum drawdown (peak-to-trough)
• Sharpe or MAR ratio (CAGR ÷ Max DD)
• Win rate and average win/loss ratio
• Longest losing streak

A viable trend-following system typically shows a win rate of 35%–45%, average win/loss ratio above 2.0, and maximum drawdown under 25%. If your backtest shows a 70% win rate, you are either curve-fitting or misreading the data.

Then paper trade for 1–3 months. Live execution reveals slippage, emotional friction, and operational issues that backtests hide.

Step 11: Commit to a Sample Size

The hardest part of trend following is sitting through losing streaks. A system with a 40% win rate will produce runs of 8–12 consecutive losses. This is statistically normal, not a signal that the system is broken.

Define in advance how many trades constitute your evaluation window — typically 100. Do not abandon the system after 20 losses, and do not abandon it after 20 wins either. Consistency over a statistically meaningful sample is what produces the edge.

Step 12: Manage Yourself

Rules are easy to write and hard to follow. The psychological demands of trend following are counterintuitive:

• You must buy after price has already risen
• You must accept being wrong more often than right
• You must hold winners through violent pullbacks
• You must cut losers quickly without hesitation

Build a pre-trade checklist. Verify trend filter, entry signal, position size, stop placement, and portfolio heat before every order. Automate where possible. The less discretion in execution, the more reliable the results.

Common Beginner Mistakes

• Over-optimizing parameters: A 47-day breakout is not better than a 50-day breakout. Simplicity survives out-of-sample.
• Too few markets: Concentrated universes produce long dry spells.
• Ignoring costs: Commissions, spreads, and slippage must be modeled in backtests.
• Revenge trading: Doubling size after a loss violates every risk rule.
• Cutting winners early: Taking 1R profits when the average winner is 4R destroys expectancy.
• Adding to losers: Averaging down is not trend following; it is hope.
• Switching systems mid-drawdown: The worst time to abandon a system is usually right before it recovers.

Tools and Platforms

For backtesting and execution: TradingView, Thinkorswim, Interactive Brokers, NinjaTrader, or Python with pandas and backtrader. For futures data: CQG, Barchart, or CSI. For portfolio-level risk: a simple spreadsheet is sufficient until you exceed 20 positions.

Data quality matters more than platform features. Adjusted continuous contracts, survivorship-bias-free equity data, and clean dividend handling all affect backtest validity.

A Sample Rule Set for Beginners

• Universe: 20 liquid ETFs across equities, bonds, commodities, and currencies
• Timeframe: Daily
• Trend filter: Close above 200-day SMA
• Entry: Close above 50-day Donchian high
• Stop: 3 ATR(20) below entry, trailed at 10-day low
• Sizing: 1% equity risk per trade, ATR-normalized
• Portfolio heat cap: 6%
• Max positions: 15
• Exit: Trailing stop hit, or close below 200-day SMA

This ruleset is intentionally plain. Its value is not in cleverness but in the discipline it enforces. Run it on paper for 90 days. Then run it live with the smallest size your broker allows. Scale only after 100 trades of consistent execution.

Measuring Progress

Track these metrics monthly: number of trades taken vs. signals generated (execution fidelity), average R-multiple per trade, portfolio heat, drawdown from peak, and adherence to rules. The last metric matters most early on. A profitable month from broken rules is a warning, not a win. A losing month from perfect execution is a data point, not a failure. Over hundreds of trades, disciplined execution of a positive-expectancy system is the only variable you control.