Weekend Gaps: Why a Stop Loss Doesn't Guarantee Your Exit Price

A gap is when price jumps straight past a range of levels, with no prices in between to trade at. The most common one is on Monday morning. The market is shut all weekend, but the news keeps coming, so the first price of the week can be far from Friday’s close. If your Stop loss sits inside the range that price jumped over, the trade closes at the first available price, not at your level. This article covers the mechanics, a worked example on gold, the “gaps always fill” folklore, and how to plan trades you hold over the weekend.

When the market closes and opens in Thai time

Forex closes for the week at 5 p.m. Friday New York time and reopens on Monday morning in Australia. Most brokers therefore stop trading around 4 to 5 a.m. on Saturday and reopen around 4 to 5 a.m. on Monday, Thai time (UTC+7). It is 4 a.m. while the US is on daylight saving time and 5 a.m. in the US winter. The details are in Forex market hours in Thai time.

That is the common schedule, not a fixed rule.

• Each symbol is different. At some brokers gold, oil and indices open the week later than currency pairs, and they have a daily break of about an hour. Small gaps can appear after each break.
• Stock index CFDs often follow the real exchange’s hours, so they can gap every night, not just on Monday.
• Long holidays such as Christmas and New Year close the market for longer, so more news piles up.

To see the real hours in MT5, right-click the symbol in Market Watch, choose Specification and look at Sessions. It lists when prices are quoted and when trading is allowed, in server time. Brokers usually announce special holiday hours in advance by email or on their website.

What a gap does to a Stop loss: a gold example

On a normal account, a Stop loss means “when price reaches this level, close at the best price available”. When the market moves continuously, the best available price is right next to your level. During a gap, the first available price is already far away.

Example: a 5,000-dollar account, gold (XAUUSD), where 1 lot = 100 ounces. The prices are made up, not real prices.

1. On Friday you buy at 2,400.00 with a Stop loss at 2,395.00, 5 dollars away
2. Risk of 1% per trade = 50 dollars, so lot = 50 ÷ (5 × 100) = 0.10 lot (a 1-dollar move = 10 dollars)
3. The market closes on Friday at 2,397.50. The Stop loss hasn’t been hit
4. On Monday morning the first price is 2,385.00, 10 dollars below the Stop loss
5. The Stop loss fires straight away and the trade closes around 2,385.00. Loss = (2,400 − 2,385) × 100 × 0.10 = 150 dollars
6. 150 ÷ 5,000 = 3% of the account, or 3 times the plan

Nothing was done wrong. The Stop loss was in place and the lot size was correct by the formula. But sizing from the Stop loss distance only works when price actually trades through that level. On top of that, the spread is usually wider than normal at the Monday open, so the real fill can be even worse.

A Stop loss tells the broker when to start closing, not what price you’ll get. For any trade you hold over the weekend, ask whether the account can take a loss of 3 times the plan.

Pending orders get gapped too

Pending orders sit on the broker’s server and trigger as soon as Monday’s first price crosses their level (see Order types explained for a refresher).

• Buy Stop / Sell Stop turn into market orders and fill worse than the level you set. Example: a Buy Stop at 2,405.00 with a Stop loss at 2,400.00 has a planned risk of 5 × 100 × 0.10 = 50 dollars. If Monday opens with a jump to 2,415.00, the order fills there. The Stop loss is still at 2,400.00, now 15 dollars away, so the real risk becomes 15 × 100 × 0.10 = 150 dollars, even though price moved the way you wanted.
• Buy Limit / Sell Limit and Take profit: if price jumps past them in your favour, they usually fill at your level or better. Order-matching rules differ between brokers, so check your account terms.
• Stop out: if you are using nearly all your margin, a single gap can push the Margin level below the Stop out level the moment the market opens. The broker closes trades at whatever price is available, without waiting for you. The mechanics are in Leverage, margin and stop out. Some brokers offer negative balance protection and some don’t, depending on the broker and the country that regulates it.

If you don’t want new trades opening at the Monday open, cancel pending orders before Friday’s close or give them an expiry.

Do gaps always fill?

A gap fill is when price comes back to touch Friday’s close. Many traders believe gaps always fill and trade against the gap as soon as the market opens.

The catch is that price moving purely at random, with no pull back at all, already returns to nearby levels very often. For a random walk with no drift, the chance of touching a level a distance g away within time T is

Chance of touching = 2 × (1 − Φ(g ÷ σ√T))

Φ is the cumulative normal distribution and σ√T is the volatility over that period. Measuring the gap as a multiple of one week’s volatility gives:

Gap size (multiple of one week’s volatility)	Chance it fills within the week if price is random
0.1	92.0%
0.25	80.3%
0.5	61.7%
1	31.7%
2	4.6%

Worked for 0.25: Φ(0.25) = 0.5987 → 2 × (1 − 0.5987) = 80.3%. We checked it by simulating 20,000 random weeks and got 79.2% (slightly lower than the formula, because the simulation moves in steps rather than continuously).

Most gaps are small compared with a full week’s volatility, so they fill often with no pull back of any kind. “Fills 80% of the time” does not mean an edge. If you fade the gap and set your Stop loss as far away as your target, a random walk hits the target first exactly 50% of the time, which is a loss after costs. And the gaps that don’t fill tend to be the big ones, which are exactly the trades that lose the most.

Test the folklore on your own data

1. Export H1 bars for the symbol you trade from MT5, covering several years at least
2. For each week, record the close of Friday’s last bar, the open of Monday’s first bar, and gap = open − close
3. Count whether it filled, and how quickly (for example within 24 hours and within the week)
4. Divide each gap by normal volatility, such as weekly ATR, and group the gaps by size
5. Compare the fill rate in each group with the table above, or better, with random levels the same distance away on days without a gap
6. If you plan to trade it, include the spread at the open, define a Stop loss and split the results by year. “Fills often” and “makes money” are different questions

Gap sizes have fat tails

Most people picture the gaps they see often. The real risk is in the gaps that happen once in a long while. Market returns have fat tails: extreme moves happen far more often than a normal distribution predicts. The full idea is in Fat tails and black swans.

To make it concrete, we simulated 1,000,000 weekend gaps in two ways, scaled so the typical gap (the median gap size) is the same. The first uses a normal distribution. The second uses a Student-t distribution with 3 degrees of freedom, which has fat tails. This is an illustrative simulation, not real market data.

Gap bigger than the typical gap by	Normal (times per 1,000 weekends)	Fat-tailed (times per 1,000 weekends)
3x	42.8	105.1
5x	0.8	31.6
8x	almost 0	8.7
10x	0	4.6

1,000 weekends is about 19 years. In the normal version, the biggest gap in 19 years is usually about 5 times the typical gap. In the fat-tailed version it is usually about 19 times (the median across 1,000 runs of 1,000 weekends each).

The practical lesson: if you look back over a few years and say “the biggest gap I’ve seen is this big”, that number is the least of what is possible, not a ceiling. And volatility tends to cluster: a week full of heavy news is often followed by a big Monday gap.

Backtests rarely see this risk

In our own research, we tested a grid system on gold M1 that cuts the whole basket at −30 dollars, over about 2.5 months of 2026, using 1-minute bar data (not real ticks). The results looked good, but a system holding several positions at once can’t be judged on 1-minute bars and needs confirming on real ticks, and the conclusion we recorded was that a weekend gap can jump straight past the −30-dollar cut, and that this tail risk is still unmeasured. The test window was too short to contain a big gap. That is also why this article has no confirmed figure for the size of gold’s weekend gaps. Backtest results do not predict future results.

The lesson goes wider. Systems that lose a little at a time while holding several positions all the time, such as grid or Martingale, look safe in short backtests right up to the day a big gap arrives. Read more in Grid and Martingale.

Crypto trades 24/7, but crypto CFDs may not

Crypto on exchanges trades 24 hours a day, every day, with no market close, so there is no Forex-style weekend gap. That doesn’t make it safer.

• Weekend liquidity is thinner. Price can move hard while few people are trading, so a Stop loss can slip even without a gap.
• Crypto CFDs on Forex platforms: some brokers quote them at weekends, others pause at certain times, for example for system maintenance. After a pause, price can jump at the reopen just the same. Check Sessions in the Specification as before.
• Overnight fees on crypto CFDs: some providers charge them every day, weekends included. Check the values in the Specification.

Other differences between the two markets are in Crypto vs Forex.

Hold over the weekend, or close on Friday?

There is no single answer. It depends on where your system makes its money.

	Close before Friday’s close	Hold over the weekend
Gap risk	None	Yes, and losses can exceed the plan
Cost	Pay the spread to reopen, and Monday-open spreads are often wide	The usual Friday-night swap
Opportunity	You also miss gaps in your favour	You get gaps both for and against you
Effect on the system	It becomes a new system that needs testing again	Matches the tested system, if it was tested holding over weekends

On swap: most currency pairs charge triple swap on Wednesday night to cover the weekend, so a trade held over Friday night usually pays just one night. But for some symbols, such as certain CFDs, some brokers set the triple day to Friday. Check the triple-swap day in the Specification, and see how it works in Swap and overnight fees. In short, the main cost of holding over the weekend is usually gap risk, not swap.

For a trend-following system that holds trades for days, closing every Friday may cut out the very trades that make most of its profit. For a short-term system that finishes within the day, there’s no reason to hold over the weekend. Either way is fine, as long as it is a tested rule, not a mood on Friday evening.

Sizing a trade you’ll hold over the weekend

A normal trade is sized from the Stop loss distance. For a trade you plan to hold over the weekend, size it from the Stop loss distance plus an assumed gap, and decide how much you are willing to lose in that case.

Same example: a 5,000-dollar account and a Stop loss 5 dollars away. Assume a gap 15 dollars past the Stop loss (3 times the Stop loss distance), and cap the loss in that case at 2%.

1. Distance you might really lose = 5 + 15 = 20 dollars, which per 1 lot = 20 × 100 = 2,000 dollars
2. A 2% loss = 100 dollars
3. Lot = 100 ÷ 2,000 = 0.05 lot
4. Check: a normal Stop loss hit costs 5 × 100 × 0.05 = 25 dollars (0.5%). The assumed gap costs 100 dollars (2%)

Pick the assumed gap from your own symbol’s data, for example the biggest gap measured over several years, then add a margin, because fat tails mean the real one can be larger. Try your own numbers with the lot size calculator.

Other options:

• Cut size before Friday. Close part of the trade so the rest can survive a gap.
• Watch long holidays and weekends with big events, such as elections or major meetings over a Saturday or Sunday.
• Guaranteed stop-loss orders. Some providers offer them. They close at your level even if price jumps past it, in exchange for a fee. They often have to sit a minimum distance from the current price and aren’t available on every symbol. Read the terms before using one.

A wide Stop loss doesn’t solve this. A gap can jump a wide Stop loss too, and a wide Stop loss already means a smaller lot. On choosing the Stop loss distance, read Wide or tight Stop loss. For slippage during midweek news, which is the same mechanism, read News spikes and slippage.

Common mistakes

• Thinking a Stop loss locks in the exit price. It only fixes when closing starts. A gap can make the loss several times the plan
• Leaving pending orders over the weekend. A gapped Buy Stop or Sell Stop fills at a worse price, and the risk per trade grows instantly
• Believing gaps always fill. Small gaps fill often even at random, and the ones that don’t fill tend to be big. Fading a gap with no Stop loss is the most dangerous trade of all
• Using the biggest past gap as a ceiling. A few years of data don’t contain the rare events
• Holding full margin over the weekend. One gap can trigger a Stop out at the open, before you can do anything
• Thinking crypto is free of this risk. Exchanges never close, but some crypto CFDs pause, and weekend liquidity is thin