Crypto Trend Following Strategy: Donchian Breakouts Explained

Trend following buys assets that are already rising and stays in until the trend breaks, accepting many small losses in exchange for occasional large wins. A classic version is the Donchian breakout: buy when price closes above its highest high of the last N days (often 20), and exit when price falls back below a trailing level such as the channel's midpoint. It suits crypto because large trends do happen, but it only works with strict position sizing and the patience to sit through long losing streaks.

Why trend following fits crypto

Crypto prices have fat tails: most days are ordinary, but now and then a coin rises several hundred percent over months. A strategy that cuts losers quickly and holds winners can capture those rare moves. Academic research on time-series momentum, the tendency of assets that have risen recently to keep rising for a while, has found the effect across many markets, which is part of why trend following is one of the oldest systematic approaches.

Trend following does not predict. It reacts. It is always late to the start of a trend and late to the end, and it makes its money in the middle.

The Donchian channel

The Donchian channel, named after the trader Richard Donchian, has three lines:

• Upper band: the highest high of the last N periods.
• Lower band: the lowest low of the last N periods.
• Midpoint: the average of the two.

A breakout is a close above the upper band: the price is higher than at any time in the lookback. In the 1980s the famous "Turtle" traders used 20-day and 55-day breakouts on futures markets, and variations are still widely used.

An example rule set

Here is a simple, fully specified system. It is for learning, not a recommendation, and it must be backtested before anyone trades it.

Rule	Setting
Universe	The most liquid coins, for example the 20 largest by market value, excluding stablecoins
Timeframe	Daily candles, decisions at the 00:00 UTC close
Entry	Daily close above the highest high of the previous 20 days
Initial stop	The 20-day channel midpoint at entry
Exit	A daily close below the trailing 20-day midpoint
Size	Risk 1% of the account per trade, from entry to stop
Filter	Only take new longs while Bitcoin closes above its 200-day moving average
Portfolio cap	No more than five open positions, no more than 5% total open risk

A worked trade

Notice that the exit gave back part of the gain: the coin peaked near $2.00. Trend following always gives back some profit at the end, because it waits for proof that the trend has ended.

What the results feel like

A typical stretch of ten trades might look like this (illustrative, not data):

Trades	Result each	Total
6 false breakouts	-1R	-6R
2 small trends	+0.5R	+1R
1 decent trend	+2R	+2R
1 big trend	+6R	+6R
10 trades		+3R

Six losers out of ten, and still a profit, but only because the one big trend was taken. Skip that trade out of frustration after six losses and the whole stretch loses money. That is why discipline and small size matter more than the entry rule.

Regime filters

Breakouts in altcoins tend to fail more during broad bear markets. A regime filter switches the system on or off based on the overall market:

• Only take new longs while Bitcoin is above its 200-day moving average.
• Some versions also exit all positions when Bitcoin closes below it.

The cost is that the filter keeps you out of the first part of a recovery, and it can flip back and forth around the average. The benefit is avoiding many of the losses of a long downtrend.

DanioX runs live versions of this family of strategies. Trend is a 20-day Donchian breakout with a trailing midpoint stop, and Regime Trend applies the same breakout to the 20 largest coins only while Bitcoin is above its 200-day average, exiting everything when it is not. Each page publishes its rules, its backtest and its live record, which change over time, so read the current numbers there rather than relying on any figure you saw earlier.

Practical issues

• Correlation: many altcoins often break out on the same day. Cap open positions and total risk, or rank signals and take only the strongest few.
• Slippage: breakouts happen when everyone is buying. Include realistic slippage in testing, especially for small coins.
• Survivorship: if you test only on coins that are big today, your results will look too good, because the coins that collapsed are missing. The next lesson shows how to test properly.
• Timing: decide on the daily close, 00:00 UTC (05:00 in Pakistan), and act consistently. Deciding on intraday prices changes the system.
• Parameter choice: 20 days is common and simple. If only one exact lookback works and 15 or 30 days fail, the result is probably luck.

Next, learn to test any system like this without fooling yourself in How to Backtest a Crypto Trading Strategy.