A good rule, borrowed by the wrong trades

The 7 percent rule for stop-loss means you sell any stock that drops 7 to 8% below your purchase price. No waiting, no averaging down, no second chances.

It comes from William O'Neil, the founder of Investor's Business Daily, and his CAN SLIM approach to growth stocks. The idea was to buy a strong stock as it broke out of a base. If the buy was right, the stock shouldn't fall that far. If it did, something was wrong, and you got out while the loss was still small.

The same approach takes profits on winners somewhere around 20 to 25%. So the 7% cap was never meant to stand alone. It sits inside a full plan for buying and holding shares over weeks or months.

You'll also see "the 7 percent rule" used for unrelated ideas, like retirement withdrawals or long-run market returns. This page is only about the stop-loss version.

Losses get harder to win back as they grow. A 7% loss needs a gain of about 7.5% to get back to even. That's because 1 divided by 0.93 is 1.075. A 25% loss needs 33%. A 50% loss needs 100%.

The cap also limits damage to the whole account. Say one position is 20% of your account. A 7% loss on it costs 1.4% of the account. Ten of those in a row cost about 13% once compounding is counted. That hurts, but it's survivable.

Then pair it with the profit side. Taking 21% on a winner while cutting losers at 7% gives a reward three times the risk. At three to one, you break even winning one trade in four. One win of 3 cancels three losses of 1.

That's why the rule works for its home use. The numbers only hold when the stock can realistically run 20% or more after you buy it.

An option's price swings much harder than the stock under it. A contract can lose 7% on a move of a few cents in the stock. Your stop fires on noise, long before the chart says you were wrong.

The spread alone can eat most of it. Say a contract is quoted 1.95 bid and 2.05 ask. Buy at 2.05 and sell straight back at 1.95, and you're down 4.9% before price moves at all. On a same-day contract, time decay keeps chipping away even when price sits still.

Intraday shares have the opposite problem. A large-cap stock seldom moves 7% in one session without big news. Buy a 50 stock and your 7% stop sits 3.50 below. If your target is the next level at 51.20, you're risking 3.50 to make 1.20.

In both cases the percentage is unrelated to the chart. The price that proves you wrong might be 30 cents away, or two dollars.

Start with the level that makes your idea wrong. For a long trade, that's often the low of a reversal pattern, such as a hammer. Draw it on the 1-hour chart or higher. Your exit is a candle close below that low, not a wick through it.

Then size the position from the distance. Say your account is 25,000 and you'll risk 1% on a trade, which is 250. If the stop sits 0.60 below your entry, you can buy about 416 shares. If it sits 1.25 below, you buy 200.

For options, measure the stop and target on the stock's price, then choose how many contracts to buy. A wider stop means fewer contracts. Don't set the exit as a percentage of premium.

Last, check the target. The next level in your direction should be at least twice as far as your stop. If it isn't, skip the trade and wait for a better entry.

Keep the rule as an outer limit for swing trades in shares. Say your chart stop for a multi-week trade sits more than 8% below your entry. Then the setup is too wide. Pass, or take a smaller position so the dollar loss stays the same.

It also helps on the days your discipline slips. A stock bought on a breakout that's down 8% a week later rarely needs more thought. Selling it is already the plan.

For day trades and options, drop the percentage and keep the habit behind it. Decide where you're wrong before you enter. Write that price down. Exit when a candle closes beyond it.

Check how it went once a week. Compare your average losing trade with your average winning trade. If losers run bigger than winners, your stops are drifting, whatever rule you think you're following.