When Spreads Widen: A Map of Predictable Danger Windows
Spreads are not constant. Knowing when they widen helps you avoid paying more than necessary and prevents stops being triggered by spread spikes.
Predictable windows exist. Around the daily rollover at 5 p.m. New York time, liquidity is thin as the trading day changes, and spreads can balloon to several pips for a few minutes. Just before and after high-impact news, spreads widen as liquidity providers pull quotes. At the weekly open on Sunday evening, quotes are sparse. And during market holidays, especially around Christmas and New Year, depth is poor.
A common problem is a stop loss triggered by a short spread spike. A long position's stop closes on the bid, so if the bid temporarily drops while the spread widens, the stop can be hit even though the chart looked fine.
Strategies to cope include avoiding trades that need tight stops during those windows, widening stops slightly during known periods, and closing positions before rollover if the strategy is sensitive.
Check your platform's spread display in real time, and compare typical spreads across the day. Some third-party tools plot historical spreads by hour.
Brokers differ in how widely spreads move, which is another test of quality.
Practical step: record your pair's spread at 4:55, 5:00, and 5:05 p.m. New York time over several days.
When Spreads Widen: A Map of Predictable Danger Windows
Re: When Spreads Widen: A Map of Predictable Danger Windows
The bid point catches people all the time. Your long stop gets hit at rollover, you look at the chart and price never touched it. The chart shows bid by default on most platforms, but the spike can be so short it barely draws.
It’s Fairman 
Re: When Spreads Widen: A Map of Predictable Danger Windows
You can turn on the ask line in MT4 and MT5 chart settings. Helps a lot when you're trying to understand why a short got stopped out.
It’s Fairman 
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: When Spreads Widen: A Map of Predictable Danger Windows
Fairman's suggestion to record the spread at those three New York times is worth doing, and it helps to note where the bid went as well. Converted for Lagos right now, the readings fall at 21:55, 22:00 and 22:05.
The chart is a made up EURUSD example: 0.8 pips at 21:55, 4.0 at 22:00, 1.6 at 22:05 and 1.0 by 22:10. The mid stays within a pip of 1.0850, but most of the widening is on the bid, which drops from 1.08496 to 1.08470.
Now picture a long with its stop at 1.08480, a comfortable 1.6 pips under the normal bid. It gets closed in that one minute, and by 22:15 the bid is back at 1.08496 as if nothing happened.
Shorts suffer less in this example, because the ask only moves from 1.08504 to 1.08510. Some brokers widen more on the bid than others, which is worth testing on demo before trusting tight stops overnight.
Does anyone here pull stops wider for those fifteen minutes, or simply close anything near its stop before 22:00 Lagos? The second option costs a little spread but avoids the argument with the broker afterwards.
The chart is a made up EURUSD example: 0.8 pips at 21:55, 4.0 at 22:00, 1.6 at 22:05 and 1.0 by 22:10. The mid stays within a pip of 1.0850, but most of the widening is on the bid, which drops from 1.08496 to 1.08470.
Now picture a long with its stop at 1.08480, a comfortable 1.6 pips under the normal bid. It gets closed in that one minute, and by 22:15 the bid is back at 1.08496 as if nothing happened.
Shorts suffer less in this example, because the ask only moves from 1.08504 to 1.08510. Some brokers widen more on the bid than others, which is worth testing on demo before trusting tight stops overnight.
Does anyone here pull stops wider for those fifteen minutes, or simply close anything near its stop before 22:00 Lagos? The second option costs a little spread but avoids the argument with the broker afterwards.
-
FXS Prop Desk
- Posts: 1689
- Joined: Wed Oct 07, 2026 9:52 am
Re: When Spreads Widen: A Map of Predictable Danger Windows
One extra idea is to record the actual spread at entry for every trade in the journal. After a few months, the data shows which hours have the widest spreads for the pairs you trade at your broker, which may differ from general guidelines. That information can then be used to adjust trading hours or add a maximum spread filter.
FXS Official Team | The Idea Lab & Strategy
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: When Spreads Widen: A Map of Predictable Danger Windows
FXS Prop Desk's idea of recording the spread at entry has one blind spot. It only samples the moments you chose to trade, so the windows you already avoid never show up, and the journal ends up confirming the habits you have rather than mapping the danger.
A small script or indicator that writes the spread every minute, all week, fixes that. A month of those readings covers every Sunday open, rollover and data release, whether you were at the desk or not.
The more useful number is spread as a share of the target. Say a 5 pip EURUSD scalp normally pays 0.6 pips, which is 12 percent of the target, while the same trade 2 minutes after a big US release pays 2.2 pips, or 44 percent.
The chart uses made up averages for five windows. At the Sunday open a 3.5 pip spread takes 70 percent of the target before price moves at all.
That suggests a filter written as a ratio rather than a fixed figure, such as no entry when the spread is above 20 percent of the target. On a 5 pip scalp that caps it at 1 pip, while a 15 pip target could tolerate 3.
A small script or indicator that writes the spread every minute, all week, fixes that. A month of those readings covers every Sunday open, rollover and data release, whether you were at the desk or not.
The more useful number is spread as a share of the target. Say a 5 pip EURUSD scalp normally pays 0.6 pips, which is 12 percent of the target, while the same trade 2 minutes after a big US release pays 2.2 pips, or 44 percent.
The chart uses made up averages for five windows. At the Sunday open a 3.5 pip spread takes 70 percent of the target before price moves at all.
That suggests a filter written as a ratio rather than a fixed figure, such as no entry when the spread is above 20 percent of the target. On a 5 pip scalp that caps it at 1 pip, while a 15 pip target could tolerate 3.
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: When Spreads Widen: A Map of Predictable Danger Windows
A surprise statement from a politician, a sudden geopolitical event, or an unexpected central bank comment can move markets without any warning. No calendar lists these.
What happens next is predictable in shape. Spreads widen, liquidity thins, and price jumps in whichever direction algorithms read the headline. Slippage increases, and stops can execute well away from their levels.
The best protection is a plan made in advance. That might include a maximum number of simultaneous trades, hard stop-losses on every position, and a rule about what to do when spreads exceed a threshold. In the moment, decisions tend to be poor.
Another useful habit is to keep a news feed visible. Not to trade every headline, but to know quickly why price just jumped and whether it is likely to continue.
Stepping aside for five to ten minutes after a shock is not weakness. It is often the cheapest way to avoid the worst fills of the month.
Preparation can also mean knowing the broker's behaviour in such moments. Some widen spreads sharply, some freeze quoting, and some reject orders. Reading past experience from a journal, or from a calm test during a smaller event, shows what to expect when it counts.
What is your rule when price suddenly spikes with no scheduled event, and has it been tested?
What happens next is predictable in shape. Spreads widen, liquidity thins, and price jumps in whichever direction algorithms read the headline. Slippage increases, and stops can execute well away from their levels.
The best protection is a plan made in advance. That might include a maximum number of simultaneous trades, hard stop-losses on every position, and a rule about what to do when spreads exceed a threshold. In the moment, decisions tend to be poor.
Another useful habit is to keep a news feed visible. Not to trade every headline, but to know quickly why price just jumped and whether it is likely to continue.
Stepping aside for five to ten minutes after a shock is not weakness. It is often the cheapest way to avoid the worst fills of the month.
Preparation can also mean knowing the broker's behaviour in such moments. Some widen spreads sharply, some freeze quoting, and some reject orders. Reading past experience from a journal, or from a calm test during a smaller event, shows what to expect when it counts.
What is your rule when price suddenly spikes with no scheduled event, and has it been tested?