Its appeal is that it needs no forecast: it waits for the market to show a new high or low and follows it. Its difficulty is that it loses on most trades, and depends on a few long trends to pay for them. This guide explains how the channel works, the classic rules and their history, worked examples on a stock, Nifty futures and a commodity, why breakout systems have low win rates and what that means for sizing, the common filters, costs, and how to build, test and run a Donchian strategy in QuantPower.
At a glance
| Donchian channel breakout | |
|---|---|
| Style | Long-term trend following |
| Instruments | Index futures, stocks, commodities |
| Timeframe | Usually daily candles |
| Indicator | Highest high and lowest low of the last N periods |
| Entry | A break of the N-period high or low |
| Exit | A break of a shorter channel in the opposite direction |
| Stop | Often 2 times ATR from entry |
| Win rate | Typically well below half |
A short history of the Donchian channel
Richard Donchian, born in 1905, is often called the father of trend following. In 1949 he started Futures Inc., one of the first publicly held commodity funds, and traded it on mechanical rules built from moving averages and price channels.1 He published his methods in a weekly newsletter for decades. His best-known rule, the "four-week rule", was as plain as a rule can be: buy when the price exceeds the highs of the previous four calendar weeks, and sell when it falls below the lows of the previous four weeks.
The channel became famous through the Turtle experiment. In 1983 the commodity traders Richard Dennis and William Eckhardt recruited a group of new traders, taught them a written system in two weeks and gave them money to trade. The system's entries were Donchian breakouts: a 20-day breakout for System 1 and a 55-day breakout for System 2, with exits on a shorter channel in the opposite direction, stops at 2 times the 20-day average true range and positions sized by that range.2 Several of the Turtles went on to run trend-following funds of their own, and the rules, once secret, were later published in full by members of the group.
Through the 1980s and 1990s, systematic trend following grew into a large industry of managed futures funds, many of them using channel breakouts among their rules.3 In India, the channel is part of the standard indicator set on charting platforms, and breakout rules on 20-day and 55-day highs are common on Nifty futures, liquid stocks and MCX commodities. Arth.live charts, inside QuantPower, include the Donchian channel alongside ATR, Keltner Channels and Supertrend.
How the Donchian channel works
- Upper line: the highest high of the last N periods, such as 20 days.
- Lower line: the lowest low of the last N periods.
- Middle line: the average of the two, used by some as a reference.
When the price breaks above the upper line, it has made a new 20-day high, a simple sign of upward momentum. When it breaks below the lower line, it has made a new 20-day low.
The channel moves in steps, not smoothly. The upper line stays flat until a new high is made or an old high drops out of the window, and then jumps. That makes it easy to see on a chart where the next breakout level is, and where the exit level of an open trade sits.
The classic rules
The best-known version comes from the Turtle Traders, trained by Richard Dennis and William Eckhardt in 1983:
- Entry: buy when the price exceeds the 20-day high; sell short when it falls below the 20-day low.
- Exit: exit longs when the price falls below the 10-day low; exit shorts when it rises above the 10-day high.
- Stop: a stop at 2 times the 20-day ATR from entry.
- Size: set the position so that a one-ATR move equals a fixed small share of capital.
- Longer system: a slower version used a 55-day breakout with a 20-day exit.
The exit channel is shorter than the entry channel so that the strategy exits faster than it enters, protecting part of the gains. System 1 also had a filter: skip a 20-day breakout if the previous 20-day breakout in the same market would have been a winner, on the idea that winners tend to be followed by failed signals. To avoid missing a large trend because of the filter, a 55-day breakout was always taken.
A worked example on a stock
This is a hypothetical example, not a result.
- A stock's 20-day high is ₹1,240. The 20-day ATR is ₹28.
- The stock trades at ₹1,242, a new 20-day high. The order buys 50 shares at ₹1,242, with a stop at 2 × ATR below, ₹1,186. The risk is 50 × ₹56 = ₹2,800.
- The stock trends to ₹1,390 over two months. The 10-day low rises with it, to ₹1,352.
- The price falls to ₹1,351, below the 10-day low. The position is sold at ₹1,351, for a gain of 50 × ₹109 = ₹5,450 before costs.
The trade made about twice its risk, though the stock had been ₹148 above the entry at its high. The 10-day exit gave back ₹39 per share from the top, as it is built to do: it stays in until the trend has clearly stalled.
A worked example on Nifty futures
Again hypothetical, on daily candles. The 20-day high of Nifty futures is 25,180 and the 20-day ATR is 210 points.
- Entry. The futures close at 25,200, above the 20-day high. One lot of 65 is bought at the next open, 25,210.4
- Stop. 2 × 210 = 420 points below, at 24,790: a risk of ₹27,300 per lot.
- Sizing. At 1% risk, one lot needs an account of about ₹27.3 lakh. A smaller account must use a smaller position, which futures do not allow, or a defined-risk option trade.
- The trend. Over seven weeks the futures rise to 26,300. The 10-day low rises to 25,980.
- Exit. A close at 25,960, below the 10-day low, closes the trade at the next open, 25,950: a gain of 740 points, ₹48,100, before costs of about ₹977 for the round trip and a rollover into the next month's contract on the way.5
The trade made about 1.8 times its initial risk. Before this trade, the same rules may well have produced three or four small losses on false breakouts. That is how the system is meant to work.
A worked example on a commodity
Commodities are where the channel was born. MCX crude oil trades from 9 am into the late evening, and one lot is 100 barrels, so a ₹1 move is ₹100 per lot. Hypothetically:
- Entry. Crude oil futures make a new 20-day high at ₹6,450. The 20-day ATR is ₹130. One lot is bought at ₹6,455.
- Stop. 2 × ₹130 = ₹260 below, at ₹6,195: a risk of ₹26,000 per lot.
- Exit. Crude trends to ₹6,980 over four weeks and the 10-day low rises to ₹6,760. A break below it closes the trade at ₹6,750: a gain of ₹295, ₹29,500 per lot, before costs.
Commodity trends are often driven by supply news and by events abroad, during the evening session. A breakout system does not need to know why; it needs the stop and size set before the news arrives.
Why breakout systems have low win rates
Many breakouts fail. The price makes a new high, then falls back. A Donchian system typically wins far less than half the time, sometimes a third or less, and depends on a few large trends to pay for many small losses. That makes it psychologically hard to follow by hand, and well suited to automation, which follows the rules through the losing streaks.
The arithmetic
Suppose a rule wins 35% of the time, with an average win of 3 times the initial risk, and loses 65% of the time, with an average loss of 1 times the risk. The expected result per trade is 0.35 × 3 − 0.65 × 1 = 0.40 times the risk, before costs. That is a healthy edge, earned in a lumpy way.
The same numbers mean long losing runs. With a 35% win rate, a run of eight or more losses in a row is likely somewhere in a hundred trades. At 1% risk per trade, that is a drawdown of 8% or more before the next winner. At 3% risk, it is a quarter of the account. The rule only works for a trader who sized it to survive the run.
What a losing streak looks like
Picture the first two months of a Donchian rule on Nifty futures in a choppy market. A 20-day breakout above 25,200 reverses within a week and hits the stop. A breakdown below 24,600 bounces back three days later. Another upside breakout fails at the old high. By the end of the second month there have been five trades and five losses, each close to 1% of the account. Nothing is wrong with the rule; the market has been in a range, and ranges produce false breakouts. The sixth breakout starts a trend that runs for four months and pays for all five. Traders who stop after the fifth loss keep the losses and miss the trend. That is the most common way a sound breakout rule fails in practice.
Adding to winners
The Turtle system did not take a full position at once. It entered one unit at the breakout and added another unit each time the price moved half an ATR further in its favour, up to four units, with the stop on all units raised to 2 ATR below the latest entry.2 Adding only to trades that are working puts the most money into the trends that pay, and keeps the risk on failed breakouts to one unit. It also concentrates risk late in a trend, and a sharp reversal after the fourth unit can take back a large part of the gain. For index futures, where one lot is already a large position for most accounts, adding is only practical with a larger account or with options.
Settings and variations
| Settings | Behaviour |
|---|---|
| 20-day entry, 10-day exit | The classic short-term Turtle system |
| 55-day entry, 20-day exit | Slower, fewer and longer trades |
| Intraday channels, such as the last 30 candles | Faster signals for intraday trading, with more noise |
| Four-week rule | Donchian's own: the highs and lows of the previous four calendar weeks |
Variations include a trend filter, such as trading only in the direction of the 200-day moving average, or skipping a breakout if the previous breakout was a winner, as one Turtle rule did.
Choosing the channel length
Shorter channels, such as 10 or 20 days, catch trends earlier and give more false breakouts. Longer channels, such as 55 or 100 days, enter later, trade less often and hold the trades that survive for longer. A longer entry channel also usually means a longer exit channel and a wider gap between the two, so each winning trade gives back more from the top. Many traders run two lengths side by side, as the Turtles did with their 20-day and 55-day systems, so that one catches the early part of a trend and the other the main body. Whatever the lengths, use the same ones across instruments, and judge them on several years rather than the last few months.
Filters, tested one at a time
- Trend filter. Long breakouts only above the 200-day average, short breakouts only below. Fewer trades against the main trend; later entries at turns.
- Volatility filter. Breakouts only when ATR is above its own longer average, so the market is active. See the atr volatility strategy for this and other ATR rules.
- Close confirmation. Enter on a close above the channel, not an intraday touch. Fewer false breaks; slightly worse entry prices.
- Volume. For stocks, a breakout on volume above its average. Less useful on index futures, where volume rises around expiry for reasons of its own.
Each filter removes some false breakouts and some real ones. Keep a filter only if the backtest improves with it across several years, not just one.
When Donchian breakouts work, and when they fail
Works best when: markets make long, sustained trends, such as strong bull runs or prolonged declines.
Struggles when: markets move sideways for months, producing breakout after breakout that reverses.
Indian indices have had both: long trends that a 55-day system would have ridden for months, and ranges where every 20-day breakout reversed within a week. No setting does well in both, and no indicator reliably says in advance which kind of market is coming. The rule's answer is to take every breakout at a small, fixed risk and let the market decide which ones become trends. A portfolio of several instruments, each traded by the same rules, smooths the results, because ranges and trends rarely come everywhere at once.
Trading several instruments
Trend followers rarely trade one market. A rule applied to Nifty futures, Bank Nifty futures, a handful of liquid stocks and two or three MCX commodities, each sized by its own ATR, has a better chance of catching at least one trend while the others range. The cost is more trades, more costs and more attention, and a rule for total open risk: for example, no more than 6% of the account at risk across all open positions at once, so that a day when every stop is hit stays survivable.
Donchian breakouts compared
| Donchian breakout | Bollinger breakout | Moving average crossover | |
|---|---|---|---|
| Signal | A new N-period high or low | A close outside a band after a squeeze | One average crossing another |
| Reacts to | Price levels | Volatility and price | Smoothed price |
| Exit | A shorter opposite channel | The middle band or a trail | The opposite cross |
| Strength | Simple, visible levels | Waits for a quiet market | Smooth, few whipsaws in trends |
| Weakness | Many false breaks in ranges | Fewer setups | Late at turns |
The Bollinger breakout waits for a quiet spell before acting; the Donchian channel acts on any new high or low. Many trend followers use the channel for entries and an ATR trail for exits.
Position sizing with ATR
Sizing by ATR keeps risk consistent across instruments and volatility levels. If you are willing to lose 1% of a ₹10 lakh account, or ₹10,000, on a one-ATR move, and the ATR is ₹28, you can hold about 357 shares. The position size calculator explains the method and does the arithmetic.
For futures, round down to whole lots. If the size comes to less than one lot, the trade is too large for the account at that risk. The choices are to skip it, to trade a smaller instrument, or to use options with a fixed maximum loss.
Intraday Donchian channels
On 15-minute candles, a 20-candle channel covers about five hours. Breakouts are frequent and small, and the opening gap can create a false new high in the first candle. Common adjustments: skip the first 15 minutes after 9:15 am, close all trades by 3:15 pm, and require the breakout candle to close beyond the channel. The opening range breakout is a special case of an intraday channel: the high and low of the first N minutes, traded once. See the opening range breakout strategy.
Trading breakouts with options
For accounts too small for a futures lot at a 2-ATR stop, options carry the same signal with a fixed maximum loss. After a 20-day breakout on Nifty, a monthly call, or a bull call spread with its short strike near the next resistance, risks only the premium paid. Trends that follow Donchian breakouts take weeks, so monthly options suit them better than weeklies, whose decay works against a slow move. Option legs in QuantPower can carry a stop in underlying points, so the option trade can still be closed when the index breaks its 10-day low.
What it costs to trade
A round trip in one Nifty futures lot costs about ₹977 in brokerage and statutory charges at the rates from 1 April 2026, most of it the STT on the sale.5 Trend trades held through expiry also need a rollover into the next contract, which is a second round trip. Stocks held for delivery pay STT on both the buy and the sell. Slippage on breakouts is real: a breakout fills in a rising market, often a few points above the trigger. Against a 740-point winner, costs are small; against the many small losers, they add up.
Building a Donchian strategy in QuantPower
QuantPower runs rules on your own broker account, through the two static IPs included with every account.
- Set up the chart. Add a 20-period and a 10-period Donchian channel, and ATR, to the Arth chart in the app.
- Test the rule. QuantPower's rule backtest reads up to 2,000 past candles of one instrument from your broker, about eight years of daily candles. Write the entry as "close crosses above the upper 20-period Donchian line", set the stop in per cent close to 2 × ATR, and add the exit rule "close crosses below the lower 10-period Donchian line". Slippage and charges are deducted from every trade. See backtesting.
- Create the signals. In Algos → Signal algos, create the entry signal and an exit signal on closed daily candles. A signal can combine up to 20 conditions, such as the breakout and a trend filter.
- One-touch or auto. In one-touch mode you get a notice with the order ready to place; in auto mode, once armed, the signal places the order itself, with every risk check. A signal that places its own order costs no tokens.
- Set the stop. Place it as a GTT order, which can wait up to 365 days, or a two-leg OCO with a target. On the chart, the GTT line can be dragged to a new trigger as the trade moves.
- Forward test. Point the signals at the Forward test account and let them trade on live prices with test money. See forward test.
- Roll before expiry. Close a futures position that is still open before the expiry week and re-enter in the next month's contract.
Account safety above the rules
Smart Exit sets a profit target, a stop-loss and a trailing step for the whole account; on a hit it stops the algos and squares off. The daily loss limit refuses new orders once today's loss reaches it. The kill switch stops a segment until its next session. They cover every connected broker together.
Backtesting tips
- Use a long history. Trend systems need several years, including trending and sideways periods.
- Include costs and slippage. Breakouts often fill at worse prices than the signal level.
- Look at drawdowns and losing streaks, not just total profit.
- Test on several instruments. A rule that works on one stock may be luck.
- Do not tune the channel length to one market's history; 20 and 55 days are widely used because they are round numbers, not because they are optimal.
- Hold back recent years. Design the rule on older data and judge it on the latest years it has not seen.
Reviewing the trades
A Donchian rule should be judged over dozens of trades, not a few. Keep a record of each: the instrument, the breakout level, the ATR at entry, the result in multiples of the initial risk, the number of days held and the exit reason. The record answers the questions that matter: is the average winner still about three times the average loser, are the losses staying close to their planned size, and are the winners coming from the instruments and periods the rule was built for? If the average winner shrinks, the market may have changed; if the losses grow, the stops are being moved or slipping.
Who the strategy suits
The Donchian breakout suits a patient, rules-first trader who can accept being wrong on most trades and hold the few right ones for weeks or months. It suits automation well, because the rules are exact and the hard part, following them through losses, is easier for software than for people. It does not suit a trader who needs frequent wins, or one with an account too small to survive a run of losses at a sensible risk per trade.
Common Donchian mistakes
- Abandoning the rules after a losing streak, just before the trend that pays for it.
- Taking profits early on the trades that are supposed to run.
- Sizing by margin instead of by the loss at the stop.
- Trading only one instrument, so a long range means a long drawdown.
- Ignoring rollover on futures held through expiry.
- Adding filters after every loss. Each new filter fits the last few trades; test it across years before keeping it.
- Entering on an intraday touch without a rule for it. Decide in advance whether the breakout needs a close beyond the channel, and keep to it.
- Holding a stock breakout through results at full size. A result gap can jump the stop by several ATRs.
Where Donchian trading is going
The rules are seventy years old and still in use, which says something about their simplicity. What has changed is the cost of following them: data, testing and execution that once needed a trading desk now fit in a browser. Scanners can list the instruments making new 20-day and 55-day highs each evening, signal algos can act on the close, and GTT orders can hold a stop for weeks. Under the retail algo framework in force since 1 April 2026, such rules run through brokers as registered algos.6 The hard part remains the one the Turtles found: following the rules through the losses, and sizing every trade so that the losses can be followed through at all.
Start with 500 free tokens
Connect your broker, build a strategy and test it on paper before a single live order goes out.
Frequently asked questions
What is a Donchian channel?
A band formed by the highest high and the lowest low over a set number of periods, such as 20 days. A break above the band signals a new high.
What did the Turtle Traders use?
A 20-day breakout entry with a 10-day exit, and a slower 55-day entry with a 20-day exit, both with ATR-based stops and position sizing.
Do Donchian breakouts work on Indian stocks?
They can, during strong trends, but many stocks spend long periods in ranges. Test across a broad list of stocks and several market phases.
Why is the exit channel shorter than the entry channel?
So that the strategy exits a fading trend before giving back too much of the gain, while still requiring a strong move to enter.
What was Donchian's four-week rule?
Buy when the price exceeds the highs of the previous four calendar weeks, and sell when it falls below the lows of the previous four weeks. It is the same channel idea, measured in weeks rather than trading days.
Related guides
- Momentum breakout strategy: trading strength after a range
- Volume breakout strategy: breakouts confirmed by volume
- Bollinger band breakout strategy: breakouts after a volatility squeeze
References
- Michael W. Covel, *Trend Following* (FT Press, 2004), on Richard Donchian, Futures Inc. and the four-week rule
- Curtis M. Faith, *Way of the Turtle* (McGraw-Hill, 2007), on the Turtle systems' entries, exits, stops and sizing
- Jack D. Schwager, *Market Wizards* (1989), interviews with Richard Dennis and other trend followers
- NSE, Nifty futures contract specifications (nseindia.com)
- Union Budget 2026-27: securities transaction tax on futures and options raised from 1 April 2026 (indiabudget.gov.in)
- SEBI, circular on safer participation of retail investors in algorithmic trading, 4 February 2025, in force for every broker from 1 April 2026 (sebi.gov.in)