Common Mistakes in Trend Following and How to Fix Them
Trend following is one of the most robust and time-tested strategies in financial markets, with roots stretching back to the turtle traders of the 1980s and beyond. Its core premise—buying assets that are already rising and selling those that are falling—is deceptively simple. Yet, execution is notoriously difficult. The psychological and technical pitfalls are numerous, and the vast majority of retail traders abandon the strategy within months, often while blaming the market rather than their own errors. This article dissects the eleven most common mistakes in systematic trend following, backed by behavioral finance research and empirical data, and provides specific, actionable fixes for each.
1. Ignoring Position Sizing and Risk Per Trade
The most catastrophic mistake in trend following is treating it as a binary “in or out” game without a structured risk management layer. Many traders allocate a fixed dollar amount or a fixed number of shares to every signal. This ignores the volatility of the underlying instrument. A stock with a historical average true range (ATR) of $0.50 versus one with an ATR of $5.00 requires vastly different capital exposure to produce the same risk of loss.
The Behavioral Trap: Overconfidence in the signal leads to uniform sizing. A sudden spike in volatility can blow a hole in the account from a single trade.
The Fix: Implement volatility-based position sizing. Use the 20-day Average True Range (ATR) to calculate position size. The formula is: Position Size = (Account Risk per Trade) / (ATR * Multiplier). For example, with a 1% account risk ($1,000 on a $100,000 account) and a stop-loss set at 2x ATR, the position size for an asset with an ATR of $10 is $1,000 / ($10 * 2) = 50 shares. This ensures each trade carries the same mathematical risk, regardless of the asset’s recent price swings.
2. Over-Optimizing the Trend Detection System
Searching for the perfect moving average crossover or the ideal channel breakout period is a form of data mining. Retail traders often run hundreds of backtests on historical data, tweaking parameters until the equity curve looks flawless. This creates a system that is exquisitely fitted to past noise but fails catastrophically in forward markets.
The Behavioral Trap: The “narrative fallacy” – creating a compelling story around a set of perfect parameters that ignores the reality of random walk variation.
The Fix: Adopt a regime-based or multi-timeframe approach instead of a single parameter. Use a slow filter (e.g., 200-day simple moving average) to determine the market “tide” and a faster trigger (e.g., 20-day high breakout) for entries. Do not optimize the exact length of the moving average. Instead, test a range (e.g., 100 to 250 days) and confirm the strategy is robust across that entire band. If performance collapses when you move from 200 to 210 days, your system is overfitted.
3. Premature Profit-Taking
Trend followers have a well-documented weakness: they take profits too early. After a winning streak, the psychological pressure to “lock in gains” becomes overwhelming. Traders exit a position that is still accelerating because it feels “too good to be true” or because they fear a sudden reversal.
The Behavioral Trap: Loss aversion magnifies the pain of giving back unrealized gains. Studies by Kahneman and Tversky show that the pain of a loss is roughly twice as powerful as the pleasure of an equivalent gain. Watching a $5,000 profit shrink to $4,000 feels like a $1,000 loss.
The Fix: Use a trailing stop that respects volatility, not a fixed percentage. The “Chandelier Exit” (based on ATR) is a proven method: set the stop at 3 * ATR below the highest high since entry. Do not tighten this stop as the trend accelerates. Allow the market to prove you wrong by a defined distance. Alternatively, implement a “scale-out” plan: exit 50% at a predetermined risk/reward ratio and let the rest run with the ATR-based stop. This satisfies the psychological need for partial profit while keeping you in the trend.
4. Averaging Down into a Losing Trend
This is the direct opposite of trend following logic, yet it remains a common error. A trader buys a stock at $100. It drops to $90. Instead of cutting the loss, they buy more to “lower the average cost,” hoping for a rebound. In a downtrend, this is equivalent to catching a falling knife.
The Behavioral Trap: The “endowment effect” and “sunk cost fallacy.” The trader treats the asset as an extension of their ego and believes buying more will validate their initial thesis.
The Fix: Strictly separate trend following from value investing. Trend followings do not lower average cost; they cut losses and re-enter on the next signal. Use a hard rule: never add to a losing position. If a new signal appears for the same asset but in the opposite direction (short), that is separate. For long-only trend following, a losing position is a sign of timing error. Add only to winning positions on pullbacks within the established trend, using a retracement entry (e.g., buying at the 50% Fibonacci level of a pullback within a rising 50-day moving average).
5. Trading a Trend That Is Too Short or Too Weak
Many traders define a “trend” too loosely. They attempt to follow a 5-minute trend in a quiet stock or a 3-day move in a sideways market. The result is a high number of small, losing trades with infrequent winners that are too small to cover the losses. This creates a negative expectancy system due to transaction costs and slippage.
The Behavioral Trap: Action bias – the need to feel engaged. Holding cash during a directionless market feels unproductive, so traders force trades in low-quality setups.
The Fix: Define a minimum trend strength using the Average Directional Index (ADX). Only take signals when the ADX is above 25 (indicating a strong trend). Additionally, filter by a minimum lookback period. For daily charts, avoid trends shorter than 20 periods. The longer the trend, the greater the potential for continuation. Use a weekly chart to verify the primary secular trend before trading the daily signal. If the weekly chart is flat, skip the trade.
6. Inconsistent Application of the System
The hallmark of a failed trend follower is “discretionary override.” A trader has a backtested system but, in real-time, chooses to ignore a signal because the news is scary (“Fed meeting tomorrow”), or they add extra filters (“This breakout looks different because the volume is low”). This cherry-picking destroys the statistical edge.
The Behavioral Trap: The “illusion of control.” Traders believe their subjective judgement at the moment of decision is superior to the objective algorithm they designed during calm analysis.
The Fix: Automate execution wherever possible. If you cannot use a fully automated trading bot, use a “trading journal with accountability.” Before the market opens, print a list of all potential signals based on your exact rules. During the session, execute them without deviation. After the trade, log whether you followed the rule exactly. Track your “rule-following ratio.” A ratio below 90% is a red flag. Fix this by reducing position size until you can follow the rules robotically.
7. Failing to Account for Major Regime Changes
Trend following systems are typically designed for trending environments. They perform poorly in sideways, range-bound markets (known as “whipsaw” markets). A common mistake is to keep trading the same system through a regime change without adapting volatility or frequency.
The Behavioral Trap: Anchoring – the trader is anchored to the belief that the last year’s trend will continue indefinitely.
The Fix: Implement a market regime filter. Use the 200-day moving average of the SPY or a major index to determine the “metatrend.” If the market is below its 200-day moving average and sloping downward, reduce position sizes for long-only trend systems by 50% or switch to a short-bias system. Alternatively, use the VIX (volatility index) as a regime filter. When VIX is rising (high fear), tighten trailing stops. When VIX is low and falling, widen stops. This prevents the system from being destroyed during a sideways consolidation period that follows a major trend.
8. Ignoring Portfolio Correlation and Drawdown Spikes
Many trend followers run a basket of assets. A critical mistake is assuming diversification when positions are highly correlated. During a market crash, long positions in stocks, commodities, and high-yield bonds all fall together. This exposes the portfolio to a simultaneous drawdown that far exceeds the model’s expectation.
The Behavioral Trap: Naive diversification – simply owning many assets without checking their correlation matrix during stress periods.
The Fix: Constrain correlation at the portfolio level. Use a correlation matrix based on 60-day rolling correlations. Do not allow any two positions to have a correlation coefficient above +0.70 unless they are hedged. For example, if you are long crude oil and long energy stocks, you are effectively betting on the same factor three times. Adjust the sizing so that the net exposure to a single macro factor (e.g., energy, technology, bonds) does not exceed 30% of the total portfolio risk. Alternatively, use a “volatility budgeting” approach where each position contributes equally to total portfolio volatility.
9. Chasing Past Performance of the Trend
“Performance chasing” is buying an asset after it has already made a spectacular move because the trader fears missing out (FOMO). This is often confused with trend following. A true trend follower buys at the breakout of a range, not after a parabolic run. Buying an asset that is 50% above its 200-day moving average with no pullback is not trend following; it is emotional gambling.
The Behavioral Trap: Recency bias – the asset’s recent strong performance feels like it will continue indefinitely.
The Fix: Use a “relative strength filter” combined with a “pullback entry.” Only consider assets that are in the top quartile of a relative strength ranking (e.g., using a 12-month rate of change). Then, wait for a pullback to a moving average (e.g., 20-day EMA) before entering. This prevents chasing extensions. Keep a distance limit: do not enter if the price is more than 10% above the 50-day moving average. This forces you to miss the first explosive leg, but it dramatically reduces the risk of buying the top.
10. Underestimating Slippage and Transaction Costs
Trend following generates frequent trades, especially if using short timeframes. A strategy that looks profitable in a backtest often becomes unprofitable after realistic slippage and commissions. Many traders use close-to-close data in backtests, ignoring that real fills happen at the bid/ask spread.
The Behavioral Trap: Optimization bias – backtesting software defaults to perfect execution.
The Fix: Backtest with conservative slippage assumptions. Add at least 0.1% to each trade for a liquid asset (ETFs, large caps) and 0.3% for illiquid assets. For each trade, subtract the bid-ask spread and a fixed commission of $5 per side. If the net profit factor (wins/losses) drops below 1.5 after costs, the system is not viable. Switch to longer timeframes (e.g., weekly from daily) to reduce turnover, or trade highly liquid futures or ETFs that offer tight spreads and low commission.
11. Expecting Continuous Profits and Abandoning the System After Drawdowns
The final and perhaps most damaging mistake is quitting the strategy during a drawdown. Trend following experiences long stretches of small losses (whipsaws) followed by infrequent, massive gains. The distribution of returns is highly kurtotic. A typical system might have a win rate of only 30-40%, but winners are 3-5 times larger than losers. Traders who cannot tolerate a 20-30% drawdown (peak-to-trough) will exit just before the next major trend.
The Behavioral Trap: “Experienced regret” – the pain of consecutive losses triggers a self-preservation response that overrides logic.
The Fix: Compute the maximum historical drawdown of your system and double it. That is your mental threshold. If the system has a historical max drawdown of 15%, be prepared for a 30% drawdown in the future. Size your account accordingly: ensure an initial capital level where a 30% drawdown would not cause margin calls or emotional collapse. Use a “drawdown stop” rule: if the system loses 25% of its peak value, take a 3-week break entirely. Do not trade. Reset mentally. This is distinct from abandoning the system. It is a tactical pause to prevent further damage from a potentially changing market regime.
Final Practical Framework for Implementation
To avoid these eleven mistakes, construct a simple, rule-based checklist that you run before every trade and every week:
- Pre-Trade Risk Check: Is this trade sized to risk no more than 0.5-1.0% of account?
- Trend Strength Check: Is ADX above 25 on the daily chart?
- Regime Check: Are we in a trending environment (VIX below 25, SPY above 200-day MA)?
- Correlation Check: Does this asset have a correlation >0.70 with any existing position?
- Entry Distance Check: Is the price within 10% of the 50-day MA?
- Execution Rule: Will I enter this exactly as backtested? Yes/No.
If the answer to any check is “No,” the trade is skipped. This systematic self-regulation is the only proven way to survive the inevitable streaks of losses that separate successful trend followers from those who fail.
