🎯 In 30 seconds: Trend followers buy what is already rising (or short what is falling) and hold until the trend clearly ends, often for weeks or months. They lose on most trades, in small amounts, and make their money on a few large trends. Research covering many markets and more than a century finds trend following has worked on average. The hard part is living through long flat or losing stretches without abandoning the rules.
What position trading means
Position trading is holding a trade for weeks to months. Trend following is the most common way to do it: a rules-based approach that enters in the direction of an established trend and exits only when the trend breaks.
A trend follower does not try to predict tops or bottoms. They accept buying after the low and selling after the high. The bet is only that, on average, a market moving strongly in one direction tends to keep moving that way for a while.
This sits at the slow end of trading styles. Compare swing trading (days to weeks) and momentum trading, which ranks stocks against each other rather than against their own past.
Two classic systems
1. Moving-average system. A moving average is the average closing price over the last N days (see Moving averages).
- Buy when the close is above the 50-day moving average. Sell when it closes below.
- A slower version uses the 200-day average. A two-line version buys when the 50-day crosses above the 200-day and sells on the reverse cross.
2. Breakout (channel) system.
- Buy when the close is the highest close of the last 50 days.
- Sell when the close is the lowest close of the last 20 days.
Notice the asymmetry in the breakout version. It takes a lot of strength to get in and only a modest weakness to get out. That keeps losers small while letting winners run as long as the trend holds.
Most trend systems add a stop based on volatility, such as 3 times the Average True Range (ATR) below the entry, so a sudden reversal does not wait for the slow exit signal.
Low win rate, big winners: the math
Trend-following systems often win on fewer than half of their trades. That sounds bad until you look at the size of the wins. Here is a hypothetical run of 10 trades, measured in R (1R = the amount risked on each trade):
The expectancy formula makes it exact:
Expectancy = (win rate x average win) minus (loss rate x average loss)
| Win rate | 30% |
| Average win: (4.5 + 3 + 4) / 3 | 3.83R |
| Loss rate | 70% |
| Average loss: (6 x 1 + 0.5) / 7 | 0.93R |
| Expectancy: 0.30 x 3.83 minus 0.70 x 0.93 | +0.50R per trade |
If you risk $100 per trade, that is about +$50 per trade on average over many trades. The same win rate with winners capped at 1.5R would lose money. That is why trend followers never take quick profits just because a trade "feels" good. More on this in Risk/reward and expectancy.
What the evidence says
Trend following has some of the strongest long-run evidence of any trading style, with important caveats.
- Moskowitz, Ooi and Pedersen (Journal of Financial Economics, 2012) studied 58 futures and forward markets: stock indexes, currencies, commodities and government bonds. A market's past 12-month return predicted its future return. The effect lasted about a year and then partly reversed.
- Hurst, Ooi and Pedersen simulated a trend-following strategy back to 1880 across many markets and found it was consistently profitable over the long run, not only in the recent decades that made it famous.
The caveats matter for you:
- These studies use a diversified basket of dozens of futures markets. A trend system on one stock or a handful of stocks is far noisier.
- They are simulations with assumed costs. Real returns include slippage, fees and taxes.
- Average profitability says nothing about the order of returns. Multi-year flat periods happened along the way.
The hard parts nobody advertises
Long drawdowns. Choppy, trendless markets produce one whipsaw after another. Over 100 trades, a system that wins 35% of the time should expect a losing streak of roughly 10 in a row somewhere along the way (the longest streak is about log(100) / log(1 / 0.65), close to 11). See Drawdowns and losing streaks for how likely those streaks are.
Giving back open profits. The exit only triggers after the trend turns. In the figure, price peaked at 150 and the system sold at 139. Giving back that part of the move is the price of staying in for the whole trend.
Boredom and doubt. Most days nothing happens. The temptation to "improve" the rules after a few losses is constant, and changing rules mid-stream usually means you exit right before the next big trend.
Taxes. Trades held for a year or less are short-term gains, taxed as ordinary income. Some trend trades last long enough to cross into long-term treatment, which is taxed at lower rates. See Taxes for traders.
Down trends, cash and the short side
The futures studies above trade both directions: long in up trends, short in down trends.
For a beginner with a stock account, the simpler choice is long or cash. When the exit signal fires, you sell and wait in cash until the next entry signal. You give up the profits from shorting, but you also avoid the extra risks of short selling, such as borrow fees and losses with no upper limit.
Long-or-cash still does something useful in a bear market: it gets you out. A 200-day average rule would typically exit early in a long decline, after the first part of the drop, and stay out until prices recover above the average. It will not sell at the top, and it may buy back above the bottom, but it can avoid the middle of a deep decline. Remember why that matters: after a 50% loss you need a 100% gain to get back to even (see Risk comes first).
The cost shows up in sideways years, when the price crosses the average again and again. Each cross is a small loss plus trading costs. Whether the protection is worth the whipsaws depends on the market and the period, which is exactly what a backtest can show you.
Building a simple trend plan
- Universe: 5 to 10 liquid ETFs from different asset classes.
- Entry: close at a 50-day high.
- Exit: close at a 20-day low, or a stop at 3 x ATR below entry, whichever comes first.
- Size: risk 0.5% to 1% of the account per position from entry to stop.
- Review: once a week after the close. Change nothing for at least 30 trades.
Try it
Open the backtest tool and test a moving-average rule on a broad index ETF: buy above the 200-day average, sell below. Compare the result with simply holding the ETF. Look at three numbers for each: total return, the largest drawdown, and the number of trades. Then try the 50-day average and see how the number of whipsaws changes. Write down which you could actually stick with, not just which made more.
- What are the entry and exit rules of the breakout (channel) system in the chapter? Answer: Buy when the close is the highest close of the last 50 days. Sell when the close is the lowest close of the last 20 days.
- A system wins 30% of the time, with an average win of 3.83R and an average loss of 0.93R. What is its expectancy? Answer: About +0.50R per trade (0.30 x 3.83 minus 0.70 x 0.93), or about +$50 per trade if you risk $100.
- The long-run studies of trend following used dozens of futures markets. Why should a trader of one or a few stocks be careful with those results? Answer: A trend system on one stock or a handful is far noisier than a diversified basket, and the studies are simulations with assumed costs.
- You tested 40-day, 41-day and 42-day averages and found one that looks perfect on the past. What does the chapter say? Answer: That "best" setting is usually luck. Pick round, sensible settings such as 50 and 200 days, check that nearby settings also work, and leave them alone.