OptiNod Academy

Chandelier Exit — ATR Trailing Stop

Removes the need to time profit-taking and helps you stay in the trade through normal trend pullbacks.

The two most costly mistakes in trend following are failing to exit after the trend has ended, turning profit into loss, and exiting too early during a normal pullback while the trend is still intact, missing the larger move. Statistically, the second mistake happens far more often than the first.

The Chandelier Exit is a tool that takes the repeated judgment out of profit-taking. If you ask, "Should I get out now?" every time price wobbles within a trend, you create more room for bad decisions. A predefined trailing stop removes that question. You hold until price reaches the stop, then exit when it does. Fewer decisions means fewer bad decisions.

A trail set 3 ATR away from price is wide enough to absorb normal movement inside a trend, but close enough to get you out when the trend truly ends. That distance is what keeps you from being shaken out by ordinary trend noise.

The Market Behind Charles Le Beau’s 22 Bars and 3 ATR

The Chandelier Exit was formalized in the early 1990s by trading systems consultant Charles Le Beau in *Computer Analysis of the Futures Market*. Le Beau used 22 bars and 3 ATR as the standard because those settings fit the U.S. futures markets of the late 1980s, including grains, energy, and currencies. Twenty-two bars was roughly one trading month, and 3 ATR was just wide enough to cover a normal daily futures pullback at the time.

Those two values do not automatically fit every asset today. In crypto, which trades 24 hours a day, one month on the daily chart is closer to 30 bars, and normal pullbacks often reach 4 ATR. For U.S. stocks, 22 daily bars still maps well to one month, but in extremely volatile names such as AI and semiconductor stocks, 3 ATR can be too tight and may stop you out on every normal pullback. If you do not test the settings on your own asset, you are effectively using values calibrated to Le Beau's 1980s grain markets.

Reasonable starting points by asset class are:

• S&P 500 ETFs and large-cap quality stocks (MSFT, AAPL): 22 bars and 2.5-3 ATR are generally reasonable because normal pullbacks in this group often stop around 2.5 ATR.
• High-volatility U.S. stocks (META, AMD): 22 bars and 3-4 ATR are more appropriate. During earnings season, 4 ATR is safer.
• BTC and ETH daily charts: 30 bars and 4 ATR are recommended. Because these assets trade 24 hours a day, one month is about 30 bars, and normal pullbacks are deeper than in U.S. stocks.
• High-volatility alts (SOL, DOGE): 30 bars and 4-5 ATR are the starting point. Even at 5 ATR, normal pullbacks can still trigger exits from time to time.

A New-High Trail Lets You Stay With Major Trends

The Chandelier structure is simple. For a long position, the stop is set at the highest high of the previous 22 bars minus 3 ATR. When price makes a new high, the stop moves up with it. When price falls, the stop does not move down. As the trend progresses, the stop naturally rises above breakeven and into profit territory.

The result is statistically clear. In trend-following strategies, most profit usually comes from one or two large trends, while most other entries end near breakeven or with small losses. The Chandelier Exit is designed to keep you in those large trends until the end. It will stop out often in short trends, but one major trend can more than offset the series of small stops that came before it.

> Assume META is in an uptrend on the daily chart, trading above the 200 EMA with ADX above 25.

> A separate setup, such as a breakout or EMA pullback, identifies the long entry.

> The exit is assigned to the Chandelier Long Exit: the highest high of the previous 22 bars minus 3 ATR(22).

> Exit at the close of the candle that closes below the Chandelier line.

> Once the trade reaches 1 R of profit, where R is the initial stop distance, take partial profits on 50% to lock in breakeven.

> Leave the remaining half to pursue the larger trend with the Chandelier Exit.

> Twenty-four hours before earnings, close half of the remaining 50% to reduce gap risk.

The key is letting Chandelier handle the exit. You do not exit manually just because RSI moves above 70 or price reaches short-term resistance inside the trend. If discretionary exits keep interfering, you will not hold the one large trend to completion, and the whole system's expectancy can turn negative.

If you had held META through 2024 using a daily Chandelier system, the major rally from January to April would likely have more than covered all the smaller stops. During that move, RSI crossed 80 twice and price hit several short-term resistance areas, but with Chandelier responsible for the exit, all of those signals would have been ignored. A trader who exited manually in the middle would have missed the entire 30% move that followed.

Right After Entry, Chandelier Helps Avoid Obvious Stops

Using Chandelier as the stop immediately after entry has another benefit. Because the stop is calculated with ATR, it is less likely to sit at obvious levels such as the prior swing low or a round number, where clustered stops tend to attract stop hunts.

Placing a stop one tick below the prior swing low is common, but that level is visible to everyone. It is directly exposed to stop hunts, where large players briefly push price below the level to trigger clustered stops.

By contrast, the Chandelier stop is set at the 22-bar high minus 3 ATR, so it usually does not overlap with an obvious structural level on the chart. Since the stop is slightly away from where most participants are looking, it is less affected by intraday stop hunts.

This effect is especially useful during the first few days after entry. Early on, Chandelier can replace an obvious swing-low stop. Once the trend moves beyond breakeven, the Chandelier line trails upward and the two levels often converge.

On Intraday Timeframes, Chandelier Can React Too Quickly

Le Beau's 22-bar, 3 ATR setting is based on daily charts. If you apply the same values directly to a 1-hour or 15-minute chart, the stop often becomes too sensitive. Intraday noise distorts lower-timeframe ATR more heavily, in relative terms, than daily ATR.

For example, 15-minute ATR can become abnormally small during low-volume periods such as overnight trading or lunch hours. In that case, the 22-bar high minus 3 ATR sits very close to price, and one normal fluctuation can trigger an exit. During high-volume periods such as the open or just before the close, the same formula may return to a more appropriate distance. In other words, the same formula behaves with different intensity depending on the time of day.

There are two ways to address this. First, set a minimum stop distance for intraday timeframes. A floor such as 0.5% of price is one example. No matter how ATR is calculated, the stop cannot move closer than that. Second, exclude session open and close periods from the ATR calculation. For 24-hour assets, removing the lowest-volume hours from ATR calculation makes the distribution more stable.

The simplest rule is to use Chandelier first on daily and 4-hour charts. On 15-minute charts or lower, other trailing tools, such as a fixed ATR stop or the prior swing low, often behave more reliably.

Three Cases Where Automated Exits Break Down

• Running Chandelier in a sideways market: When ADX is below 20, Chandelier exits usually become repeated stop-outs. Price frequently crosses 3 ATR inside the range. Confirm that a trend exists using ADX and price structure first, and run the system only when the trend is clear.
• Earnings gaps and weekend gaps: In stocks and futures, a large gap can make the Chandelier exit fill far from the stop level. If a gap down causes an exit 5 ATR below the stop area, the loss on one trade can grow to 1.5-2 times normal. That is why many systems use a separate rule to reduce part of the remaining position before earnings announcements or weekends.
• Misunderstanding win rate: Chandelier systems often win only 30-45% of the time. The structure depends on one or two large trends offsetting many small losses, so the win rate is naturally low. If you add discretionary exits to raise the win rate, you can miss the large trends and push the system's expected value negative. This system should be evaluated by R-multiple and expectancy. Win rate is the wrong yardstick for it.

Two Conditions to Check Before Letting Chandelier Manage the Exit

For a Chandelier exit system to work properly, two things need to be in place.

• Entry tool: You need a tool that identifies the start or continuation of a trend, such as a Donchian breakout, EMA pullback, or price-structure breakout. Chandelier handles only the exit. Another tool defines the entry.
• Trend filter: The standard is ADX above 25 or a clearly directional 200 EMA. In sideways markets, turn Chandelier off and switch to a mean-reversion strategy.