Market vs Limit vs Stop Orders: The One That Feels Safest Isn't
You set a stop at $90 to cap the loss at $1,000. It opens at $71 and you are filled at $71. Nothing malfunctioned.
By Pavel Penev, MScFounder, TradeWize · 10+ years trading the marketsYou own 100 shares. You paid $100 for each of them, and you weren't casual about it. You set a stop at $90, so if the thing goes wrong you lose $1,000 and then you're out. That's a sensible plan, and it's the plan every beginner's guide tells you to make.
Then bad news lands after the close. The next morning it opens at $71, your stop fires, and you're sold at $71. Your loss is $2,900.
Nothing broke. Your broker didn't fail you and nobody cheated you. The stop did exactly what a stop does. It's just that what a stop does isn't what most people think it does, and the gap between those two things cost you $1,900 you never agreed to lose.
The short answer
A market order guarantees you get filled and says nothing about the price. A limit order guarantees the price and says nothing about whether you get filled. A stop order guarantees neither, because a stop isn't a price. It's a trigger. When the price you named is reached, your stop becomes a market order, and a market order takes whatever is there. On a quiet day that's the price you expected. On the morning above, it's $71 when you asked for $90.
Two guarantees, and you only get one
Every order type ever invented answers one question. When your order and the market disagree, which of them gives way? There are only two answers. Either the order gives way and takes whatever price is going, or the order holds out for its price and might not trade at all.
That's the whole trade-off, and there's no way around it. No broker sells an order that promises you both. Promising both would mean promising somebody will be standing there to take the other side at a price you picked, and nobody can promise you that. The order types are just different ways of choosing which half you'd rather have.
Whether you're guaranteed a fill runs down the side. Whether you're guaranteed a price runs across the top. A market order takes the first, a limit order takes the second, and the top-right box is empty because nothing fills it. No broker anywhere sells an order that promises you both. The stop sits in the opposite corner, promising you neither.
You can watch it in the simulator further down. A plain market sell fills on all 4 of the days in there, including the ugly ones, and it never argues about the price. A limit order set away from the market fills on some of those days and not others. Its price is never in doubt.
A market order fills now, at whatever price is there
A market order says: I want in, or I want out, and I'm not arguing about the price. It's the default button on almost every app, it fills in a fraction of a second, and most of the time nothing interesting happens.
It isn't free, though, and no statement will ever show you the bill. At any moment there are two prices, not one. The bid is what buyers will pay you. The ask is what sellers want from you. Say the bid is $49.90 and the ask is $50.10. The quote page calls that stock $50. Nobody trades at $50. It's just the midpoint of the two real prices.
A market order always crosses to the far side. Buy, and you pay the ask. Sell, and you take the bid. The distance between them is $0.20, which is 0.40% of the price, and you pay it going in and again coming out. On 100 shares a round trip costs you $20 before a single commission or fee is charged. Your broker will call the trade free, and by their definition it was.
On a calm morning in a big stock that's pennies, and you can stop worrying about it. The catch is that the spread isn't a fixed cost. It's the market's own measure of how nervous it is, so it widens exactly when things go wrong. On the crash day below, the spread starts at $0.10 and blows out to $0.70 at the low. That's 7 times wider, and a market order placed at that moment pays it.
Where a market order goes properly wrong
Thin books and odd hours. A huge index ETF at eleven in the morning has buyers and sellers stacked a penny apart, and your order barely moves it. A small, rarely traded fund in the first minute of the day might have almost nobody quoting it. A market order there takes the best of whatever handful of prices exist. Same button, wildly different bill.
A limit order gets your price, or nothing
A limit order names a price and refuses to do worse. Buying, it's the most you'll pay. Selling, it's the least you'll take. If the market never comes to you, nothing happens at all, and that's not a malfunction. Not trading is one of the two outcomes you signed up for.
The half people forget is that a limit fills at your price or better. Take a limit buy at $89 on the crash day. It doesn't fill at $89. It fills at $88.50, because that's the first price on offer once the market came down to meet it, and it was better than what you asked for. That's $50 across 100 shares that a market order would have handed over without mentioning it.
Now run the same order on the other days. On the quiet day it never fills. On the day that dips 4% and climbs all the way back, it never fills either. The price just never came down that far. You kept your money and you kept your discipline. You also sat on the sidelines while the thing you wanted went up without you, and that's a real cost.
It works the same in reverse, and that's worth seeing, because it's the trap the stop-limit sets later. A limit sell at $100 on the gap morning never fills at all. You said you wouldn't sell below what you paid. The market went nowhere near it, so you still own every share. Your price was protected perfectly. You're just not out.
A limit order also lets you leave the screen. Name a price you'd be happy with, place it, and go and live your life. That beats watching a chart and making the decision at speed with money on the line.
Learn it by doing
Reading about it is one thing — it clicks when you do it. Practise this hands-on in a free, interactive lesson (Stage 7: Brokers, Accounts & Getting Started).
Try the free lesson →A stop order is a trigger, not a price
Here's the one that gets people, and it gets them because it sounds like safety. You set a stop and you feel protected. The number you typed feels like a floor under your position. It's not a floor. It's a doorbell.
A stop order isn't in the order book at all. It sits dormant at your broker, watching, doing nothing. Nobody can see it. Then price touches the number you named, and the order wakes up. The SEC's investor bulletin on stop orders puts what happens next in one sentence: "When the stop price is reached, a stop order becomes a market order."
Read that again, because everything else follows from it. Your stop price isn't the price you sell at. It's the price that starts you selling at whatever the market feels like paying. Every stop order has two prices in it, and most people only ever think about the first one.
On a normal day those two prices sit close together and nobody notices. Set a stop at $90, price drifts down through $90, you get filled around $90. Fine. The difference only shows up when it costs you, which is the worst way to learn it.
100 shares bought at $100, a stop at $90, and news that lands after the close. The shaded band is every price between the stop and the fill, and not one share changed hands anywhere in it. The stop was obeyed to the letter. It fired at the first price actually available, and the first price actually available was $71.
That's the gap in the opening story. It opens 29% below last night's close. Nothing traded at $90, because between the close and the open nothing trades at all. When the market reopens, the queue starts at $71. Your stop didn't fail you. It got you the first price that existed. You planned to lose $1,000 and you lost $2,900, and the extra $1,900 is the distance between the price you named and the price that was real.
The overnight gap is the dramatic version. The everyday version is smaller and far more common, and it's in the simulator below too. On the fast-crash day the same $90 stop fires and fills at $88.65, which is $135 worse than the stop price on 100 shares. Then the price half recovers and ends the day back at $93.94. You sold into the hole. Doing nothing would have left you $529 better off by the close.
This is a documented failure mode, not a thought experiment. The SEC's Division of Trading and Markets wrote about it in January 2016, and the day it points at is 6 May 2010, the flash crash. Falling prices triggered a wave of retail stop orders. Each one turned into a market order and tried to sell straight away, into a market where the buyers had gone. They got poor prices, and their own selling drove the price down further, which triggered more stops. Some exchanges later removed the stop order type from their rule books over exactly this.
How far from the value of a thing can a price at the open be?
Further than most people imagine. On 24 August 2015, US markets opened badly enough to trigger 1,278 volatility halts across 471 securities, 1,058 of them in exchange-traded products. QQQ, the second most actively traded exchange-traded product of the day, hit a low more than 17% below the previous close while the basket of shares it holds was down about 10%. IVV fell more than 20% at its low while the net asset value of SPY, which tracks the same index, was down 8%. The SEC's research note on that morning doesn't mention stop orders anywhere at all, and this isn't a story about stops. It's a story about prices. The price printed at an open can be a long way from what the thing is actually worth, and a market order can't tell the difference.
One more wrinkle, and it's the sort you only find out afterwards. Brokers don't agree on what counts as "reaching" your stop. Some watch the last traded price and some watch the quotes. The same stop at the same number can fire at different moments at two different firms. It's in your broker's order documentation, and it's worth five minutes.
None of this is an argument for trading without a stop. A position with no stop has no defined loss, and a loss you can't define is a loss you can't size. It's an argument for knowing that the number you typed is a trigger and not a promise. Where to put that trigger is a separate problem, and your first instinct puts it in the same place as everyone else's. Our piece on why price spikes through your stop and reverses covers it.
A stop-limit protects your price by not selling
So the fix looks obvious. If the problem is that a triggered stop takes any price going, put a floor under it. That's a stop-limit: two numbers instead of one. The stop still triggers, but now it becomes a limit order rather than a market order, and that limit won't go below your second number. Set the stop at $90 and the floor at $89, and nothing can sell your shares under $89.
That is exactly what it does, and it's also the trap. Read the sentence again. Nothing can sell your shares under that price. Including on a day when every price is under that price.
The same overnight gap, the same $90 stop, now with a selling floor at $89 beneath it. The stop triggers on the open exactly as designed. Then nothing trades at $89 or better for the rest of the day, so the order sits there unfilled and the shares stay yours.
Run it on the gap and the stop-limit doesn't fill. The stop is touched on the first quote of the day, so your protection fires exactly as promised. Then the order waits for $89, and $89 never arrives. The best bid all day is $75, $14 a share short of your floor. At the close you still hold 100 shares worth $7,500 against the $10,000 you paid. You're fully exposed to whatever tomorrow brings, and your safety net has already been used up.
A plain stop gets you out at an unknown price. A stop-limit names the price and might not get you out. That's the first chart again, wearing different clothes. It's the same fork all the way down.
That doesn't make the stop-limit the worse order, and the crash day is where it earns its keep. Same two numbers, different market. The price falls fast, the stop triggers, and the floor refuses the panic price underneath it. A buyer appears 2 quotes later and the order fills at $90.30. The plain stop on that same day took $88.65. Waiting was worth $165 on 100 shares, and you got out either way.
That's the honest summary of the pair. When the bad price is a brief air pocket, the floor steps over it and you come out ahead. When the bad price is the only price all day, the floor leaves you holding the position. Neither order knows in advance which kind of day it's about to meet, and nor do you when you place it.
Don't take any of that on trust. Below is the same simulator every number in this article comes from. Set the ticket, pick a day, and run it.
Write a ticket, pick the day the market’s about to have, and press Run. The quotes arrive one at a time, the same way they would on a screen. Nothing here is a guess — every fill comes from the same simulator the charts in this article are drawn from.
If it trades here, send a market order.
Ten quotes stand in for a whole session, and real spreads move around more than these do. The mechanics are the real part. Set the stop at $90, pick the overnight gap, and watch which number you end up with — then try the same stop on the quiet day and see it never fire at all. The order type doesn’t decide what happens. The day does.
Two runs worth doing before you scroll on
Leave the ticket as it arrives, a $90 stop to sell 100 shares, and run it against the overnight gap. Watch the fill print at $71. Then switch the order type to stop-limit, leave the $89 floor where it is, and run the identical day again. This time nothing sells. Same news, same two numbers, opposite outcome, and neither of them is a bug.
How long your order stays alive
Every order carries a lifespan as well as a price, and the field is usually a small dropdown you scrolled past. It's called time in force, and there are two settings that matter.
- Day. If it hasn't filled by the closing bell, it's cancelled. Tomorrow it isn't there, and if you still want the trade you place it again.
- Good-till-cancelled, or GTC. It stays out there across days until it fills or you kill it. Most brokers expire them after a few months anyway, and the exact rule is theirs, not the market's.
The difference sounds administrative. It isn't. A day order that expires unfilled is a decision that quietly undid itself. The classic version is a limit buy set slightly too low, dying at the close while you carry on assuming you're in the queue.
A GTC stop has the opposite problem. It's still working weeks after you forgot about it, which is the whole point of setting one and also how it ends up firing at an open you slept through. The $2,900 morning at the top of this article is a GTC stop doing its job perfectly at the opening bell while you're asleep. Set a stop and forget it, and the market will remember it for you.
There's also a moving version of the trigger. A trailing stop follows the price up at a fixed distance or percentage behind the high, so it ratchets up as you make money and never slides back down. It solves a real problem, which is having to move the stop by hand and then talking yourself out of it. It doesn't solve this article's problem. A trailing stop still becomes a market order when it's hit, so over an overnight gap it's filled at the same first-available price as a fixed one. The trailing part changes where the trigger sits. It changes nothing about what happens once the trigger is touched.
What each order does, situation by situation
None of the rows below tell you what to do with your money. They say what each order type does when it meets that situation, which is the part you need before you can decide anything.
| The situation | What each order type does there |
|---|---|
| Your monthly buy of a huge index ETF, mid-morning | A market order fills instantly, and the spread on a fund like that is usually a penny or two, so the cost is small change. A limit order at or just above the ask fills within seconds too, and caps what you pay. Set it too low and the month's contribution sits there unspent until you notice. |
| A thinly traded or exotic ETF, or a small company | A market order takes the best price on offer, and on a thin book that can be a long way from what the thing is worth. A limit order caps the damage and accepts that some days nothing happens. The 24 August 2015 dislocations above were in ETFs, some of the largest in the world. |
| The first and last few minutes of the trading day | Spreads are widest and quotes move fastest at the open and into the close, so a market order at those moments pays the widest spread of the day. A limit order puts a boundary on it. A stop triggered in that window becomes a market order in that same window, which is where the two effects compound. |
| Holding through earnings, with a stop in place | The stop is a trigger, so the price it gets depends on where trading restarts. If the stock opens 29% lower, the fill lands near the open and nowhere near the stop. A stop-limit prevents that fill and can leave the position open instead, exposed to the next day. |
| You want out of this position today, whatever it costs | A market order does that, and it's the one job it guarantees. A limit order and a stop-limit both put a price condition in the way, and a price condition is exactly what leaves you still holding the position at the close. |
Read down the second column and the same pattern repeats in every row. The order that removes the price risk adds the risk of not trading. The order that removes the risk of not trading adds the price risk. Picking an order type is picking which of the two you'd rather carry today.
Is a limit order always better than a market order?
No. A limit order removes price risk and adds the risk of not trading, and which of those matters more depends on the day. On a large, heavily traded stock or ETF in normal hours, the spread a market order pays is often a cent or two. Set the limit slightly wrong there and the trade simply doesn't happen. On a thin instrument, at the open, or in a fast market, the price a market order takes can be a long way from the quote you saw. That's where a limit earns its keep.
Can a stop loss fail?
It depends what you mean by fail. A stop triggers reliably when the price you set is reached, and in that sense it almost never fails. What it doesn't do is guarantee the price you get afterwards. Once triggered it becomes a market order and takes the first price available, so if the market gaps past your stop overnight you're filled at the open rather than at your stop. In this article's worked example a $90 stop was filled at $71, turning a planned $1,000 loss into $2,900.
Does a stop-limit order guarantee a price?
It guarantees you won't sell below your limit, which isn't the same as guaranteeing you sell. If the price falls straight past the limit and stays there, the order triggers, waits for a price that never comes, and you keep the position. On the gap day in this article the stop-limit triggered on the first quote and never filled, because the best bid all day was $75 against a limit of $89. On a fast crash that recovers, the same order fills at $90.30 while a plain stop takes $88.65, so the limit is worth $165 on 100 shares. Same order, opposite results, and the market picks which.
What happens to my order overnight?
That's set by the time in force you chose. A day order is cancelled at the close if it hasn't filled, so it's gone by morning. A good-till-cancelled order stays live across days until it fills or you cancel it, though most brokers expire GTC orders after a few months. A resting GTC stop is the one that fires at an open you weren't watching. That's both the reason to use one and the reason to know it's there.
Why did my stop fill lower than my stop price?
Because a stop price is a trigger, not a fill price. When the stop is reached the order becomes a market order and takes the best price available right then. In a fast or gapping market that can be well below the trigger. It isn't a broker error, and it usually isn't anybody hunting you. The two common causes are an overnight gap, where nothing traded at your price at all, and a sharp intraday drop where the spread widened at the same time.
Should a beginner use market orders?
We can't tell you what to place, but the mechanics are worth knowing. A market order's cost is the spread. That's usually tiny on large, liquid stocks and funds in regular hours, and it can be substantial on thin instruments, at the open and close, and in volatile conditions. Plenty of people use market orders for routine buys of big index funds and switch to limits for anything small, unusual or fast-moving. What causes trouble isn't the order type. It's using one without knowing which guarantee it gives up.
What's the difference between a stop order and a stop-limit order?
A stop order has one price. When it's reached, the order becomes a market order and fills at whatever the market offers, so you're guaranteed to be out and not guaranteed a price. A stop-limit has two prices, a trigger and a floor. When the trigger is reached it becomes a limit order that won't go past the floor, so you're guaranteed a price and not guaranteed to be out. The plain stop's risk is a bad fill. The stop-limit's risk is no fill, and still owning the position.
Placing the order is the part nobody teaches
Choosing what to buy gets all the attention, and then the actual buying happens through a screen full of dropdowns nobody explained. Which order type, which price, how long it lives, what your broker does with it after you tap the button. That's the mechanical half of investing, and getting it wrong quietly costs money that never shows up as a fee.