The 7% Stop Loss Rule: Where It Comes From, the Math, and When to Break It (2026 Guide)

The 7% stop loss rule says you sell any stock that falls 7% to 8% below the price you paid for it, every time, with no waiting for a bounce. It comes from William O'Neil, the founder of Investor's Business Daily, who built it into his CAN SLIM growth-stock method and treated it as a rule with no exceptions. Its job is to cap the damage from any single bad buy at a level you can recover from.

That is the whole rule. The useful part is everything around it: why the number is 7 and not 3 or 15, what it does to your position size, where it breaks, and how a stop order actually fills when the price gaps through it. This guide covers each, with a worked trade you can copy.

What the 7% rule for stop-loss actually says

Three details in O'Neil's wording get lost when the rule is passed around.

First, the 7% is measured from your purchase price. It is a loss limit on the money you put in, so a stock that runs from $50 to $60 and then falls back to $56 has not triggered it. Once a trade is in profit, the 7% rule has done its job and a different exit takes over, usually a trailing stop or a profit target.

Second, it is a maximum. O'Neil's books describe 7% to 8% as the ceiling on a loss, and his advice was that most losses should be cut sooner than that if the trade is already acting badly. Traders who treat 7% as the place to put every stop, regardless of the chart, are using the outer fence as the plan.

Third, it was written for a specific kind of trade: buying a growth stock as it breaks out of a base, at a defined buy point. In that setup, a properly timed entry should work fairly quickly. A stock that falls 7% to 8% below a breakout buy point has usually failed the breakout, and the base it came out of is being undone. The rule is a judgment about breakouts compressed into a number.

What the rule does not say matters as much. It says nothing about how many shares to buy, so it cannot by itself control how much of your account you lose. It says nothing about intraday swings, so it was never meant for day trades. And it assumes you can actually get out near 7%, which an overnight gap can make impossible. Each of those gets its own section below.

Where the 7% number comes from

O'Neil's reasoning had two parts. Only the second is something you can check yourself.

The first was observational. Looking back over his own trades and his studies of past market leaders, O'Neil reported that well-timed buys rarely fell more than 7% or 8% below a proper buy point before they moved up. That is his account, and it has not been published as a dataset you can test. A breakout that has dropped that far is more likely a failed breakout than a normal shakeout. So the 7% line sits roughly where "this is noise" turns into "this was wrong."

The second was arithmetic, and that part anyone can check. Losses compound against you, because every percentage you lose has to be earned back on a smaller amount. A 7% loss needs a 7.5% gain to get back to even. A 50% loss needs a 100% gain. Small losses are cheap to repair and large ones can take years, which is why O'Neil wanted a hard ceiling before the loss got large.

Be careful with anything beyond that. You will see claims online that the 7% rule "works" or that a certain share of winning trades never dip below it. Unless a claim names the dataset, the period, and the entry method, treat it as folklore. What the rule reliably does is put a ceiling on the damage from one mistake. Whether a ceiling at 7% beats one at 5% or 10% for your trades depends on how your entries behave, and that is something only your own trade log can answer.

The recovery math behind cutting losses early

The table every CAN SLIM book prints is worth seeing as a chart, because the curve is the argument. Each bar is the gain you need, on what is left, to get back to where you started.

Bar chart of the gain needed to recover from a loss. A 5% loss needs a 5.3% gain, 7% needs 7.5%, 8% needs 8.7%, 10% needs 11.1%, 15% needs 17.6%, 20% needs 25%, 30% needs 42.9%, and a 50% loss needs a 100% gain.

The formula is gain needed = loss / (1 - loss). At 7%, that is 0.07 / 0.93, or 7.53%. At 8%, 0.08 / 0.92, or 8.70%. Up to about 10%, recovering takes only slightly more than you lost. Past 20%, the gap widens quickly, and past 30% you need a trade most traders will not see more than a few times a year.

The second piece of math is the one that decides whether a 7% stop can make money at all: the break-even win rate. If your average loss is 7% and your average winner is W%, you need to win more than 7 / (7 + W) of your trades to come out ahead, before costs.

Bar chart of the break-even win rate with a 7% average loss. With a 7% average winner you must win 50% of trades, with a 14% winner 33.3%, with a 21% winner 25%, and with a 28% winner 20%. Costs and slippage are excluded.

This is why O'Neil's books pair the stop with a habit of letting winners run to around 20% to 25% before taking most profits. At a 3-to-1 payoff, you can be wrong on most trades and still break even at 25% winners. Use the same 7% stop with a habit of grabbing 5% gains, and you need well over half your trades to work, which is a much harder bar. The stop and the profit-taking habit are one system. Copying half of it gives you the losses without the payoff that was supposed to cover them.

The 7% rule is a price stop, not a risk budget

Here is the gap most explanations skip. A 7% stop tells you where to get out. It does not tell you how much money that costs, because the cost depends on how big the position is.

Put your whole $20,000 account into one stock with a 7% stop, and a stop-out costs $1,400, or 7% of the account. Put $2,000 in, and the same stop-out costs $140, or 0.7%. Same rule, ten times the damage. The thing that protects your account is the combination of the stop distance and the position size, which is the core idea in our risk management guide.

The usual way to join them is to pick an account risk per trade first, then size the position so a 7% stop-out costs exactly that. Fidelity's guide to managing positions states one common version: do not let a position lose more than 2% of the overall portfolio. Many active traders use 1% or less.

The sizing formula is position size = account risk / stop distance. With a 7% stop:

Two things fall out of that table. At 1% risk, a 7% stop means no single position should be much more than a seventh of the account, which forces some diversification on its own. And at 2% risk, positions get large enough that a gap through the stop (covered below) turns a planned 2% loss into something noticeably bigger. If you use an 8% stop, the positions shrink a little: 12.5% of the account at 1% risk instead of 14.3%.

The 3-5-7 rule is a different rule with a similar name. It caps risk per trade, total exposure, and the target profit ratio. You can run both together, with the 7% stop as the price exit and the 3-5-7 limits as the account-level caps.

A worked trade, step by step

The numbers below are a hypothetical example, built to show the arithmetic. They are not a real ticker or a real result.

A stock spends six weeks building a base between $47.60 and $50.00. You mark $50.00, the top of the base, as the buy point. It breaks out on rising volume and you get filled at $50.40. Your account is $20,000 and you risk 1% per trade, so $200.

1. Compute the 7% and 8% levels. 7% below $50.40 is $46.87 (50.40 x 0.93). 8% below is $46.37 (50.40 x 0.92).
2. Check them against the chart. The base low is $47.60. Both rule levels sit below it, so if price reaches them the base has fully failed. That is the agreement you want: the rule's number and the chart's invalidation point say the same thing.
3. Choose the stop. You could stop at $47.45, just under the base low, which is a 5.9% risk. Or you can use the rule's ceiling at $46.87. The structure stop is tighter and still makes sense on the chart, so it is the better choice here. The 7% level stays as the hard limit you will not move past.
4. Size the position. Using the $46.87 stop, the risk is $3.53 a share, so $200 / $3.53 = 56 shares, a $2,822 position (14.1% of the account). Using the $47.45 stop, the risk is $2.95 a share and you can take 67 shares for the same $200.
5. Place the stop order immediately as a good-til-canceled order. The next section covers which order type.

Line chart of a hypothetical stock over 14 sessions. It trades in a base between $47.60 and $50.00, breaks out to a $50.40 entry on session 8, then pulls back to $49.10 and recovers to $52.80. Horizontal lines mark the entry at $50.40, the base-low stop at $47.45, and the 7% rule stop at $46.87.

Now flip the chart. Say the base low were $44.80 instead. Then a stop that respects the structure sits about 11% below your entry, and the 7% level lands in the middle of the base, where ordinary chop can hit it. The rule and the chart now disagree, and you have three honest options: skip the trade, wait for a later entry closer to support so the structure stop fits inside 7%, or take a smaller position with the wider stop and accept that you are no longer running the 7% rule. What you should not do is keep the full position and use the 7% stop anyway, because the stop is then sitting at a level the chart gives no reason to respect.

How to place a 7% stop order, and why it may not fill at 7%

A stop you hold in your head is a stop you renegotiate every time the price nears it. Put the order in the broker as soon as the entry fills. Most brokers offer two types, and they fail in opposite ways.

A stop market order (a plain stop) turns into a market order once the stop price trades. It will get you out, but not necessarily at your price. FINRA's investor guide on stop orders in volatile markets is direct about it: "your stop order may be executed at a price that's significantly different from your stop price." The same page warns that "rapid price movement during a short period of time could trigger a stop order," which is how a brief spike can take you out of a position that closes the day above your level.

A stop-limit order adds a limit price, so you never sell below a floor you choose. The cost is that you may not sell at all. In FINRA's words, "there's no guarantee that your order will be executed." If the price drops through both your stop and your limit, you are still holding a falling stock, which defeats the purpose of a loss limit.

For a loss-cutting rule, the stop market order is usually the better fit, because the whole point is to be out. The buy stop limit order guide covers the order mechanics in more depth, including brackets and OCO orders.

The case neither order solves is the overnight gap. Stop orders on most US stocks trigger during regular hours. If a stock closes at $49.00 and opens at $44.90 after bad news, your $46.87 stop triggers at the open and fills somewhere near $44.90. In the worked example, that is a loss of about $5.50 a share instead of $3.53. On 56 shares, the planned $198 loss becomes roughly $308, or 1.5% of the account instead of 1%, and the position is down almost 11% instead of 7%.

You cannot stop that from happening with order types. You can only size for it. Earnings dates are the most predictable source of gaps, so many 7% rule users either exit before the report or cut the position size for any trade that will be held through one.

What the research on fixed-percentage stops says

There is real academic work on whether stop-loss rules help, and it gives a more careful answer than either "always use them" or "they only lock in losses."

Kathryn Kaminski and Andrew Lo's paper "When Do Stop-Loss Rules Stop Losses?", published in the Journal of Financial Markets and archived at MIT's DSpace repository, builds a framework for measuring what a stop-loss rule adds to or subtracts from a strategy. The core result: if prices follow a pure random walk, a simple stop rule lowers expected return, because you exit and re-enter at random. When returns show momentum, where losers tend to keep losing for a while, the stop can add value. Testing on index futures, they found that some stop-loss policies raised expected return while cutting volatility substantially at longer sampling frequencies.

A 2009 study by Bergsveinn Snorrason and Garib Yusupov at Lund University tested traditional and trailing stops on OMX Stockholm 30 stocks from January 1998 to April 2009, and concluded that "the stop-loss strategies can do better than the buy-and-hold" strategy, with the advantage clearer on a risk-adjusted basis. A summary of that paper by Quant Investing reports the level-by-level results: the 5% trailing stop was the only level that did worse than buy-and-hold, at -8.14% cumulative, while the 15% trailing stop had the highest cumulative return at 73.91%.

What that does and does not tell you:

• Neither study tested O'Neil's rule as he wrote it, a 7% to 8% stop from a breakout buy point on growth stocks. They tested other entry methods on other markets.
• The tight end of the range did worst in the Swedish sample. That fits the intuition that a stop inside normal daily noise exits too often. It does not prove that 7% is too tight for your trades.
• Momentum is the condition under which stops helped. O'Neil's method buys strong stocks breaking to new highs, which is a momentum strategy, so the rule is at least consistent with the conditions where stops have helped in the research.

None of this is a guarantee about future results. It is a reason to treat the rule as a reasonable default for momentum-style entries and to check it against your own fills.

When the 7% rule fits and when it does not

A flat percentage treats every stock as if it moves the same amount, and stocks do not. The test is to compare 7% with how far the stock typically moves in a day. The Average True Range (ATR) measures that, and dividing a 14-day ATR by the price gives it as a percentage.

If a stock's ATR is 1.5% of its price, a 7% stop is almost five average days of movement, which is plenty of room. If the ATR is 5%, the 7% stop is less than a day and a half, and a single normal swing can hit it. That is the core problem with the rule on volatile names: the same number is generous on a large, steady stock and a hair trigger on a small, fast one.

Here is where the rule tends to fit, and where it does not:

• Fits: swing and position trades in liquid growth stocks, bought at a defined breakout or pullback level, where the stock's daily range is a small fraction of 7%. This is the trade it was designed for.
• Strains: low-priced and small-cap stocks with wide daily ranges and wide spreads. The stop lands inside normal noise, and the spread adds slippage on every exit.
• Breaks: crypto. Major coins routinely move more than 7% in a day, so a flat 7% stop on a spot position gets hit by ordinary movement. If you trade crypto with a percentage cap, set it from the coin's own ATR and shrink the position to match.
• Breaks: day trading. On a 5-minute chart, a 7% stop is so far away that it means nothing, and the risk per share it implies would force tiny positions. Intraday trades need stops based on the intraday structure; our guide on where to place a stop loss covers those methods.
• Too tight: long-term investing. Someone holding an index fund or a stock for years will see 10% drops as routine. Fidelity's same managing-positions guide suggests using "no more than a 20% stop loss on each position" for that kind of holding, so that normal pullbacks do not whipsaw you out.

The common thread is that the rule works when 7% is a meaningful distance for that stock on that timeframe. When it is not, you either change the number or change the trade.

The 7% rule versus other stop methods

The 7% rule is one of five common ways to decide where a stop goes. Most experienced traders combine them, using one to set the level and another as a cap.

A combination that holds up well: set the stop at the structure level, check that it sits inside 7% of your entry, and size the position so the stop costs 1% or less of the account. Once the trade is up by more than your original risk, switch to a trailing stop to protect the gain. The 7% rule then does what it does best, which is to make sure no single entry can become a large loss.

How to adapt the rule without breaking it

You can tune one part of the rule. The other part is fixed.

The tunable part is the number. A trader whose entries are tight and well timed can use 5% or less as the ceiling. A trader holding through more volatility can use 10%, with smaller positions to keep the account risk the same. The adaptation that keeps the spirit intact is to set the ceiling from the stock's ATR and keep the account risk fixed.

The part you cannot tune is the direction. A stop only ever moves in the direction that reduces risk. The moment you lower a stop because the price is getting close, the rule is gone and you are holding a position with no exit. Averaging down by buying more as the price falls toward the stop is the same mistake with more money attached. Fidelity's guide calls adding to a losing position one of the fastest ways to lose money in the market.

A few other adjustments that traders use and that stay consistent with the rule:

• Cut sooner when the trade acts wrong. If a breakout immediately falls back into the base on heavy volume, many traders exit at 3% or 4% instead of waiting for 7%. The 7% is a ceiling.
• Tighten in a weak market. When most breakouts are failing, a stop closer to entry and smaller positions reduce how much you give back while the market sorts itself out.
• Use partial exits. Selling half at 4% down and the rest at 7% lowers the average loss, at the cost of more decisions.

Common mistakes with the 7% rule

• Measuring from the high. The rule measures from your entry. Measuring 7% from the peak after a run-up is a trailing stop, which is a different tool with different behavior.
• Using it without position sizing. A 7% stop on an oversized position can still cost 5% of the account. Size first, every time.
• Putting it inside the stock's normal range. If 7% is less than two average days of movement, you will get stopped out by noise. Check the ATR before you enter.
• Keeping it as a mental stop. Mental stops get moved. Place the order when the entry fills.
• Using a stop-limit and assuming you are protected. If the price gaps through your limit, the order does not fill and the loss keeps growing.
• Holding a full position through earnings. The gap risk is predictable and the stop cannot cap it. Exit or size down before the report.
• Copying the stop without the profit-taking. A 7% stop with tiny winners needs a win rate most strategies do not have. The math only works when winners are several times larger than the stop.

FAQ

What is the 7% rule for stop-loss? It is William O'Neil's rule to sell any stock that falls 7% to 8% below your purchase price, with no exceptions. It comes from his CAN SLIM method, where it caps the loss on any breakout that fails.

Is the rule 7% or 8%? O'Neil gave it as 7% to 8%, and he described that as a maximum. Many traders use 7% as the hard ceiling and cut sooner when a trade acts badly right after entry.

Is the 7% measured from my purchase price or from the stock's peak? From your purchase price. Once the stock has risen, a trailing stop or a profit-taking rule handles the exit. Measuring 7% from the peak turns the rule into a trailing stop.

Does the 7% rule work for crypto? Usually not as a flat number. Major coins often move more than 7% in a normal day, so the stop gets hit by ordinary volatility. Set the stop from the coin's own ATR or chart structure and shrink the position so the loss stays inside your account risk.

Should I use a 7% stop for day trading? No. On an intraday chart, 7% is far beyond any level that matters, and the risk per share would force very small positions. Day trades use stops based on intraday structure, often well under 1% from entry on liquid stocks.

What happens if the stock gaps below my 7% stop? A stop market order triggers at the open and fills near the opening price, which can be well below your stop. A stop-limit order may not fill at all. The only defense is position size, and avoiding full-size positions through known events like earnings.

Is a 7% stop better than a 10% stop? It depends on the stock's volatility and your payoff. A tighter stop means smaller losses but more stop-outs; a wider one means fewer stop-outs but larger losses and smaller positions for the same account risk. Compare each against the stock's ATR and track your own results over at least a few dozen trades.

How does the 7% rule relate to the 2% rule? They answer different questions. The 7% rule sets where you exit. The 2% rule, as Fidelity describes it, caps how much of the whole portfolio one position can lose. Use the stop distance and the account risk together to set the position size.

Can I use the 7% rule with a broker's app like Schwab or Robinhood? Most major US brokers support stop market and stop-limit orders on stocks, including good-til-canceled duration. Check your broker's order-type page for whether stops trigger in extended hours, since that changes how gaps are handled.

Reading the chart is still your job

The 7% rule is a ceiling on damage. It works best when the chart agrees with it: the structure stop sits inside 7%, the stock's daily range is small next to that distance, and the position is sized so a stop-out costs about 1% of the account. Each of those checks is a chart read.

Quant AI does that read from a screenshot. It marks the swing lows and support levels it finds, so you can see at a glance whether your 7% stop sits below the structure or in the middle of the noise. Deciding whether to take the trade, and how big to make it, stays with you.