Stop-Loss Order

Reviewed 12 August 2026 · Sourced from SEC investor bulletins, FINRA Rule 5350, the SEC and CFTC report on 6 May 2010, and the SEC's own market-structure research

What a stop order actually is

You own 200 shares of something. You paid $52, and you work a shift that covers the entire trading day, so you cannot watch a screen. You have already made the only decision that matters: if this drops to $47, you are out. The problem is that at 10:40 in the morning your phone will be in a locker.

A stop order is the instruction you leave behind. You give your brokerage firm a price — the stop price — and the order sits dormant until the stock trades at or through it. Until then it is not a bid or an offer, and nobody is looking at it. It is a trigger, not a price you have offered anyone.

Now read the next sentence twice, because it is the whole page. When the stop price is reached, the stop order becomes a market order. That is the SEC's own wording, from its investor bulletin on exactly this subject. And a market order has no price condition at all. It transacts at whatever the market is currently showing.

So the stop price answers one question, when, and answers nothing whatsoever about at what price. The SEC says so in the same bulletin: “The stop price is not the guaranteed execution price for a stop order,” and the fill “can deviate significantly from the stop price in a fast-moving market.” FINRA, which writes the rules your brokerage firm operates under, puts the same warning under a heading that leaves nothing to interpret: “Stop Prices Aren't Guaranteed Execution Prices.”

One naming note, so the rest of the page reads cleanly. The regulatory term is stop order — that is the phrase in FINRA Rule 5350 and in the exchange rulebooks. “Stop-loss” is what traders call a sell stop placed below a position they already own. The SEC and CFTC did use “retail stop-loss orders” in their report on the 2010 flash crash, so both names are in the official record. They describe the same instrument.

Stop, stop-limit and trailing stop, side by side

Three order types share the stop trigger and differ entirely in what happens after it fires. Every definition in the middle column below is the SEC's, from the same 2017 bulletin.

The gap, worked in dollars

This is the failure that costs retail investors the most money, and it has nothing to do with placing the order wrong.

Price is not a continuous line. Between the 4:00 p.m. close and the 9:30 a.m. open there is no trading, but there is plenty of news — a guidance cut, a lawsuit, an earnings miss, a whole weekend of it. When the market reopens, the first print can be nowhere near the last one. That distance is a gap, and your stop does not skip over it. It is on the far side of it.

Nobody made a mistake in that example. The order did precisely what the SEC bulletin says it does. This is also why the same bulletin warns that a stop “may be triggered by a short-term, intraday price move” producing an execution price “substantially worse than the stock's closing price” — the trigger and the fill are two separate events, and only one of them is a number you chose.

6 May 2010, and the guardrails built afterward

Most of what you read about stop orders and the flash crash is secondhand. It does not need to be. The SEC and the CFTC published joint findings on 30 September 2010, and stop orders are named in them.

The report's own sentence: “sell orders placed for some individual securities and ETFs (including many retail stop-loss orders, triggered by declines in prices of those securities) found reduced buying interest, which led to further price declines in those securities.” That is the mechanism in one line. Stops fired, each became a market order, and market orders arrived in a market where the buyers had stepped back.

The prices that resulted are also on the record. Over 20,000 trades across more than 300 securities executed at 60 percent or more away from their 2:40 p.m. prices, some as low as one penny and some as high as $100,000. After the close, the exchanges and FINRA met and jointly agreed to cancel — to break — all such trades.

The regulatory response is equally citable. On 1 June 2012 the SEC approved two mechanisms that still run today:

• Limit up-limit down (LULD). Trades in an individual listed security cannot occur outside a price band around a rolling five-minute average price. The band is 5 percent for the more liquid names — S&P 500 and Russell 1000 members and certain exchange-traded products — and 10 percent for other listed securities, and the bands are doubled during the opening and closing periods. If trading cannot occur inside the band for more than 15 seconds, the stock pauses for five minutes.
• Market-wide circuit breakers. Rebuilt to trigger on declines of 7, 13 and 20 percent from the prior day's close, measured on the S&P 500 — replacing the old 10, 20 and 30 percent on the Dow Jones Industrial Average — with halts shortened to 15 minutes and the day split at 3:25 p.m. A 7 or 13 percent decline after 3:25 p.m. no longer halts the market; a 20 percent decline closes it for the rest of the session.

Both are real and both work. Neither is a stop-loss guarantee. A pause interrupts a cascade so that liquidity can reassemble; when the stock reopens, your triggered stop is still a market order and it meets the reopening price. The plumbing got better. The instrument did not change.

Where a stop order reaches, and where it doesn't

People assume a stop order is a feature of the market. It is closer to a feature of your brokerage firm, and the difference shows up in four places.

Your broker does not have to offer it

FINRA Rule 5350(a) is one clause long and settles the question: “A member may, but is not obligated to, accept a stop order or stop limit order.” The SEC's bulletin says the same thing from the investor's side — stop, stop-limit and trailing stop orders “may not be available through all brokerage firms.” What your firm will accept on a leveraged product, an option, a thinly traded over-the-counter stock, or during a pre-market session is a question with a different answer at every firm. It is on the order-entry screen and in the customer agreement, and it is worth reading before you need it rather than after.

Some exchanges stopped taking them

FINRA states it directly: “The inherent risks involved with stop orders have prompted certain stock exchanges to cease accepting stop orders.” The NYSE is the concrete case. A client notice dated 28 January 2016 removed stop orders and good-till-canceled orders from NYSE and NYSE MKT beginning 29 February 2016, and canceled the existing ones after the close on 26 February. The same notice adds the detail that matters most: this “does not prevent brokers from accepting stop orders on behalf of their clients.” So your stop is very likely held on your broker's system, and what arrives at an exchange when it triggers is a market order. Your firm's handling rules, not an exchange rule, decide the behavior in between.

Day or good-till-canceled, and neither covers the night

A stop carries a time in force like any other order. A day order expires at the close — which is the exact moment overnight gap risk begins. A good-till-canceled order persists, though FINRA notes firms typically cap how many days that is. On extended hours, be careful: the SEC's extended-hours bulletin says “many brokerage firms currently accept only limit orders” in those sessions, spreads are wider because there is less trading interest, and “certain existing mechanisms that address volatility in specific stocks or in the market as a whole may be limited or not available during extended-hours trading.” Read that last clause twice: the LULD band and the circuit breakers are regular-session machinery.

An ETF is not a stock, and the data says so

On 24 August 2015 the SEC's own research staff counted 1,278 LULD trading pauses across 471 securities in a single session. 83 percent of them — 1,058 pauses across 327 exchange-traded products — were in ETPs, and fewer than 2 percent involved S&P 500 or Nasdaq-100 members. On the same day, 19.2 percent of ETPs fell 20 percent or more, against 4.7 percent of corporate stocks. An ETF is a wrapper whose market price depends on the things inside it being priced; when they are not trading cleanly, the wrapper's price and the value of its holdings can separate. A stop resting under an ETF is exposed to that, and the SEC's numbers are the strongest available evidence for it.

Why stops pile up in the same places

A stop order is not displayed. It is not a bid, it is not an offer, and in most cases it does not leave your broker's system until it fires. That makes it invisible individually. It does not make it invisible in aggregate, and the reason is boring: people put them in the same places.

Just under a round number. Just under an obvious support level. Just under the line a breakout came from. Not because of coordination — because those are the only landmarks a chart offers, and everyone is reading the same chart.

The best evidence for this is genuinely good and comes from an unexpected place. Carol Osler, working from the complete conditional order book of a major foreign-exchange dealing bank — 9,667 orders between September 1999 and April 2000 — found that stop-loss orders cluster on the far side of round numbers. Stop-loss buy orders were roughly twice as common just above a round figure as just below it: 14.4 percent of them landed in the range immediately above, against 7.4 percent immediately below. She also found that trends gained momentum once support and resistance levels were crossed. The paper ran as a Federal Reserve Bank of New York staff report and then in the Journal of Finance in 2003.

Be honest about what that is and isn't. It is currency-dealer data from 1999 and 2000, not U.S. equities in 2026. The behavior is the same species of human behavior and the mechanism transfers plausibly, but no equity study establishes it, because no equity broker publishes its order book.

That is also one explanation offered for why a move through an obvious level so often reverses immediately: part of the push through was stops converting to market orders, which is mechanical selling with no opinion behind it, and it exhausts. Treat that as an interpretation many practitioners hold, not as an established finding. The false-breakout section of breakout covers the pattern; nobody has published a rate at which it happens.

The stop distance is your position size

The stop and the size of the position are not two decisions. They are one decision taken in two halves, and most people only notice the first half.

Now the part that formula hides, and it is the reason this section exists. The arithmetic assumes the stop fills at the stop price. The SEC has already told us it does not guarantee that. Put the gap from earlier through the tighter version: 120 shares, a stop at $49.50, and an open nine dollars below it is 120 × $9 = $1,080 of realized loss, not $300. The tighter the stop, the larger the position, and the larger the position, the more a gap costs. Tightening a stop does not reduce gap risk. It concentrates it.

The same caveat runs through the risk/reward ratio. The stop distance is the denominator. If the stop does not hold, the ratio you computed before you entered was a piece of fiction, and it was the flattering kind.

The psychology, said plainly

The real value of a stop is not precision, because there is no precision available. It is that it converts an open-ended fear into a decided number, at a moment when you are calm, and writes it down where your future self cannot negotiate with it. That is worth something on its own terms.

And then there is the way people undo it. Price approaches the stop, and the stop gets moved down to “give it room.” This is the single most common route from a small loss to a large one. Notice what the act actually is: canceling a decision made calmly, at the precise moment you are least calm, using an argument you would not have accepted an hour earlier. Loss aversion is why the small loss feels unbearable in that moment and the larger unrealized one somehow does not. The Trading Psychology course works this in Stage 3, and the behavior page has a five-question checklist meant to be run before the order exists, which is the only time the answers are honest.

What's documented here, and what isn't

This site labels its confidence rather than flattening everything into the same voice. Here is the accounting for this page.

What trips people up

• Reading the stop price as the price you will get. It is the price that starts the process. The SEC calls it out in one sentence and it is still the most common misunderstanding in retail investing.
• Not knowing which of the two you placed. A stop and a stop-limit look almost identical on the order screen and fail in opposite directions. Check the confirmation, not your memory.
• Treating a stop-limit as protection in a collapse. The collapse is the case where it does not fill and you keep the entire loss.
• Moving it down to give the trade room. The one habit most reliably responsible for turning a planned loss into an unplanned one, and it always arrives dressed as patience.
• Tightening the stop and then sizing up. Same nominal risk, more shares, more gap exposure. The formula does not warn you because the formula assumes the fill.
• Setting a trailing percentage smaller than the stock's ordinary weekly range. The SEC's own phrasing: short-term fluctuations can activate it. You will be exited by nothing happening.
• Assuming a day order covers the night. It expires at the close, which is when gap risk starts.
• Assuming every broker takes stops on everything. FINRA Rule 5350 says a member may but is not obliged to accept one, and the NYSE stopped taking them on its own book in 2016.
• Assuming a halt or a canceled trade is a safety net. A five-minute pause reassembles liquidity. It does not restore your position, and whether a bad print stands was decided by a committee after the close in 2010.
• Using money that has a job. No order type changes what a loss does to a household with no buffer under it. That is what Stage 2 is for, and it comes first.

The figures on this page are checked against the source that publishes them, and dated. Published rates move after the release named above — the linked source always carries the current number. This page explains a term; it does not recommend a product.