The First Seconds After Big News: Spread, Slippage and Straddles

High-impact US releases such as NFP (nonfarm payrolls), CPI (inflation) and the FOMC decision (the Federal Reserve’s policy committee) can move a chart by dozens of pips in a few seconds. Most people see a long candle and think “opportunity”. Inside that candle, though, prices vanish, the spread widens several times over, and orders fill away from where they were set. This article explains the mechanism, works through how much a two-sided pending-order setup can lose, and shows how to measure your own slippage.

How to read the calendar itself (Actual, Forecast, Previous) is covered in How to read the economic calendar. This article doesn’t repeat it and focuses only on what happens to your orders.

What time US news comes out in Thai time

Release	New York time	During US daylight saving	Rest of the year
NFP, CPI	08:30	19:30	20:30
FOMC statement	14:00	01:00 next day	02:00 next day
Fed chair press conference	14:30	01:30 next day	02:30 next day

US daylight saving runs from the second Sunday of March to the first Sunday of November. During that period New York is 4 hours behind UTC, otherwise 5 hours. Thailand is UTC+7 all year, so 08:30 + 4 + 7 = 19:30 Thai time (UTC+7). In 2026 the US goes back to standard time on 1 November, so from then on an 08:30 release lands at 20:30 Thai time.

FOMC comes in two steps. The statement goes out first, then half an hour later the Fed chair starts taking questions. Price often jolts a second time during the press conference.

The mechanism: why prices disappear at the release

The prices you see in MT5 come from liquidity providers (LPs), the banks and firms that quote buy and sell prices to the market. The structure is explained in FX market structure and liquidity. Think about it from the LP’s side.

1. Before the release the LP doesn’t know the number. Any quote it leaves out could be wrong a split second after the data appears, and whoever reads the number fastest will hit that stale quote first. So LPs pull their quotes, cut their size, or widen the spread to protect themselves.
2. The second the number comes out automated systems read it and send orders within milliseconds, but there is little left on the other side. Price jumps over several levels instead of moving one step at a time.
3. Seconds to minutes later the market digests the details. NFP has payrolls, the unemployment rate, wages and revisions to earlier months. CPI has a headline and a core figure that can point different ways. The order book is still thin, so modest orders push price a long way. Price whips back and forth (a whipsaw).
4. As LPs return the spread narrows step by step and price moves more continuously again.

Andersen, Bollerslev, Diebold and Vega (2003) studied exchange rates around US macro announcements and found that prices adjust to the surprise (the gap between the number and expectations) very quickly, with most of the reaction concentrated at the very start. In practice a retail trader has almost no chance of trading before the old quotes are gone. What is left for us is the most expensive part.

Slippage is not symmetric

Slippage is the difference between the price you asked for and the price you got. In normal conditions it comes out roughly as often in your favour as against you. Around news, some order types are built to get a worse price.

• A Buy Stop in MT5 triggers when the Ask rises to its level, and a Sell Stop when the Bid falls to its level. It then becomes an order at the next available price. If price jumps past the level, the next price is further away in the same direction price is moving, so it is always worse.
• A Stop loss works the same way. A Buy’s Stop loss watches the Bid and a Sell’s watches the Ask. When price runs hard against you, it closes beyond your level.
• A limit order (Buy Limit, Sell Limit) only fills at your price or better, so it doesn’t get negative slippage. But it may not fill at all.

Each order type is covered in Order types explained. The key point here: whoever places stop orders around news is asking to buy while price surges and to sell while it drops, exactly when the other side has disappeared.

Also note that some execution setups pass on only negative slippage and keep the positive part. You can check by measuring your own fills as described at the end. If your normal-hours fills almost never show positive slippage, that is worth asking your broker about.

Worked example: a straddle that loses on both legs

A straddle here means placing a Buy Stop above price and a Sell Stop below it before the release, hoping to catch whichever way it runs. Let’s compute it on an assumed price path, EURUSD at 0.10 lot (1 pip = 1 dollar).

• Mid price before the release 1.08500, normal spread 0.6 pip
• Buy Stop 1.08650, Stop loss 1.08450, Take profit 1.08850
• Sell Stop 1.08350, Stop loss 1.08550, Take profit 1.08150
• The plan: the triggered leg runs 40 pips to Take profit and the other gets cancelled. If wrong, each leg loses 20 pips

On the assumed path the mid falls to 1.08260, shoots up to 1.08780, then drifts back to about 1.08480 within a minute. The spread peaks at 9 pips and then narrows. Orders are matched by MT5’s rules: an order triggers on the first tick where the Bid or Ask crosses its level and fills at that tick’s price.

Step by step (positive pips in bold are slippage against you):

1. At 0.4 seconds the mid jumps from 1.08502 down to 1.08310 with a 9-pip spread. Bid = 1.08310 − 0.00045 = 1.08265, already below the Sell Stop. The Sell fills at 1.08265, worse than 1.08350 by 1.08350 − 1.08265 = 8.5 pips
2. At 1.9 seconds price bounces. Ask = 1.08560 + 0.00030 = 1.08590, past the 1.08550 Stop loss. The Sell closes at 1.08590, 4 pips past its SL. The Sell loses 1.08590 − 1.08265 = 32.5 pips
3. At 2.3 seconds price keeps rising. Ask = 1.08700 + 0.00030 = 1.08730, past the Buy Stop. The Buy fills at 1.08730, 8 pips worse than its level
4. Price peaks at a Bid of 1.08755, short of the 1.08850 Take profit, then falls. At 10 seconds Bid = 1.08420 − 0.00010 = 1.08410, past the 1.08450 Stop loss. It closes at 1.08410, 4 pips past the SL. The Buy loses 1.08730 − 1.08410 = 32 pips

Both legs together lose 32.5 + 32 = 64.5 pips = 64.50 dollars. The mid covered 1.08780 − 1.08260 = 52 pips, more than the 30 pips between the Buy Stop and the Sell Stop, yet neither leg reached Take profit. A minute later the mid sat at 1.08480, just 2 pips from where it started.

The cost versus the “move” on the chart

The 1-minute candle for this period would show a 52-pip move. Someone looking back sees a long candle and assumes a straddle must have caught it. The numbers that matter:

• The 9-pip peak spread is 9 ÷ 30 = 30% of the gap between the two orders. The spread alone uses up almost a third of it without the mid moving at all
• The four slippages add up to 8.5 + 4 + 8 + 4 = 24.5 pips, more than the Stop loss distance
• A tester that fills at the exact order prices would show a loss of 20 per leg, 40 pips in total. In this example the real loss was 64.5, which is (64.5 − 40) ÷ 40 = 61% worse

A long move is not the same as a move you can capture. If most of it happens as jumps over prices nobody was quoting, the profit you see on the chart never existed for anyone. When you plan, count costs in money per trade as in Spread cost per trade, and use news-time numbers, not normal ones.

A long candle after news shows where price went. It doesn’t show how much anyone actually traded at those prices.

Why MT5 backtests hide this cost

• Fixed spread. If the test uses one spread value throughout, news minutes get the same spread as quiet ones
• 1 minute OHLC uses four prices per minute and the spread recorded in the bar, which is the lowest of that minute. Spread spikes lasting a few seconds disappear completely
• Every tick generates the path inside each 1-minute bar itself, so the order of the swings in the news minute isn’t real, and that order decides which straddle leg gets hit first
• Even real ticks only show the quotes that were recorded. They don’t show how long your order takes to travel, or how much size was available at that price

Each mode is covered in MT5 backtest tick data quality, and the other reasons live results differ are in Backtest vs live results.

Our own data shows how much an average can hide. We measured the gold (XAUUSD) spread from 61.3 million real ticks between January and August 2026, on the one account type we test with. The average was 0.272 dollars, the 95th percentile (the value 95% of ticks were at or below) was 0.40 dollars, and the maximum was 2.00 dollars. The maximum is 2.00 ÷ 0.272 ≈ 7.4 times the average. At 0.10 lot (10 ounces) that means 2.72 dollars per round trip normally versus 20 dollars at the widest. This measurement covers every hour; we didn’t isolate news minutes, so it can’t say the maximum always happens at news. It does show that hourly or yearly averages can’t see short bursts like these. These figures come from one broker and one period, and past data doesn’t tell you the future will look the same.

Measure your own slippage

The numbers in this article are examples. The numbers you can make decisions with are your own account’s. Here’s how:

1. Pending orders and Stop losses. In MT5’s History tab, show Orders and Deals. The level you set is on the Order line and the price you actually got is on the Deal line. For trades closed by Stop loss, compare the close price with the SL level you set
2. Market orders. MT5 doesn’t record the price you saw when you clicked. If you trade by hand, note the on-screen price before you click. If you use an EA, log the price you sent (MqlTradeRequest.price) against the price you got (MqlTradeResult.price), plus the Bid and Ask at the moment you sent it
3. Record it in signed pips. Use one rule for the whole log, for example positive = worse than requested, negative = better
4. Split the groups. News trades (say 5 minutes before to 15 minutes after a high-impact release) versus normal trades. Look at the average, the worst, and how many went for you versus against you

An assumed log of 15 EURUSD trades at 0.10 lot:

Group	Slippage per trade (pips, positive = worse)	Average	Worst	Against / for / exact
Normal hours, 10 trades	0, 0.1, −0.1, 0, 0.2, 0, −0.1, 0.1, 0, 0.1	0.03	0.2	4 / 2 / 4
News, 5 trades	1.5, 3.2, 0.8, 6.0, 2.4	2.78	6.0	5 / 0 / 0

Read it like this: in normal hours the cost is almost zero and goes both ways. Around news it averages 2.78 pips, or 2.78 dollars per trade, and not one trade slipped in your favour. If your system’s average profit per trade is smaller than that, trading the news is turning it into a loser. Keep this for every trade in your trading journal. The more trades you log, the more you can trust the number.

Practical rules

• Be flat before the release. Unless your system is designed for news, close positions or cancel pending orders before the release time. Pending orders sit on the broker’s server and still trigger after you close MT5
• Wait after the release. Set a rule that you won’t open new trades until a certain number of minutes has passed and the spread is back within a certain multiple of normal. Base both numbers on your own logs. For FOMC, remember the press conference half an hour later
• Reduce size if you must hold through news. In the example, the real loss was about 1.6 times the plan. If you want the real loss to stay at 1% of your account even then, trade 1 ÷ 1.6 = 62.5% of the size you would get from your normal Stop loss. Work out the size with the lot calculator and Position size from Stop loss
• Use a limit instead of a stop when it fits the plan. If the plan is to wait for a pullback to a level, a limit order can’t slip against you on entry. If you need a breakout entry, MT5 has Buy Stop Limit and Sell Stop Limit orders that cap the worst entry price, at the cost of sometimes not filling
• Give EAs filters. Set a maximum spread for entries and a news-time filter, then test that it really stops both new trades and pending orders
• Read your broker’s news conditions. Some brokers lower leverage for new positions around high-impact news, which raises the margin you need. Check the trading conditions and contract specification

Common mistakes

• Measuring a news candle’s High-to-Low after the fact and treating it as profit you “would have made”, without subtracting the spread and slippage of that moment
• Placing a straddle and assuming only one leg can lose, when a whipsaw can hit both and each can lose more than its Stop loss distance
• Trusting a fixed-spread or OHLC backtest for a system that trades through news
• Measuring slippage as one average over all trades, so a handful of news trades get buried under many normal ones
• Forgetting the US clock change, or the press conference after the FOMC statement, and being in a trade when price jolts a second time
• Thinking a Stop loss guarantees a price. Around news it only tells you when closing starts, not at what price it ends