Spread During News — Measured Minute by Minute
What the latest major release actually cost on FxPro’s Raw+ feed, tick by tick: EUR/USD all-in about $44 per lot at the peak vs $9 quiet — ×5 for a few seconds. The bigger hazard was the 1.7 s quote gap: news slippage lives in that silence, not in the spread. Window: Initial Jobless Claims, Average Hourly Earnings m/m, Nonfarm Payrolls, Unemployment Rate — 2 Jul 2026, 22:30 AEST.
Open FxPro Account →Initial Jobless Claims, Average Hourly Earnings m/m, Nonfarm Payrolls, Unemployment Rate — 2 Jul 2026, 22:30 AEST
| Instrument | Quiet spread (pips) | Peak at release | All-in at peak | Peak vs daily range | Peak hits after | Longest quote gap | Back to normal |
|---|---|---|---|---|---|---|---|
| EUR/USD | 0.2 | 3.7 (×18.5) | $44/lot | 6% | 2 s | 0.7 s | 94 s |
| GBP/USD | 0.6 | 5.3 (×8.8) | $60/lot | 7% | 3 s | 1.1 s | 9 s |
| AUD/USD | 0.4 | 4.9 (×12.2) | $56/lot | 11% | 1 s | 1.2 s | 15 s |
| USD/CAD | 0.4 | 6.2 (×15.5) | $51/lot | 11% | 2 s | 1.7 s | 11 s |
| USD/JPY | 0.3 | 8 (×26.7) | $56/lot | 11% | 1 s | 1.1 s | 98 s |
| XAU/USD (Gold) | 15 | 150 (×10.0) | $157/lot | 1% * | 3 s | 1.4 s | 3 s * |
Definitions: quiet spread = median of the 15 minutes before the release; peak = widest single quote after it. All-in = peak spread × pip value + $7 round-turn Raw+ commission ($3.50/side, 1 lot) — quiet EUR/USD all-in is $9 ($2 spread + $7 commission). Peak vs daily range = peak spread as a share of the instrument’s average daily range. Back to normal = first quote under 1.5× the quiet median. * Gold always runs wide relative to FX, so its share of daily range is small and it “recovers” almost instantly by this measure. Times in AEST.
What printed
| Release | Actual | Forecast |
|---|---|---|
| Initial Jobless Claims | 215 | 209 |
| Average Hourly Earnings m/m | 0.3 | 0.3 |
| Nonfarm Payrolls | 57 | 43 |
| Unemployment Rate | 4.2 | 4.2 |
Values as published by the platform calendar feed (its own scale — e.g. payrolls in thousands).
Execution reality: the gap, not the spread
In this window the feed went quiet for up to 1.7 s on one instrument — roughly 20× the measured order round-trip. A market order sent into that silence fills at the next real quote, wherever it prints: that gap — not the spread — is where news slippage lives.
| Instrument (1.00 lot) | Order round-trip | Avg slippage | Worse-side fills |
|---|---|---|---|
| EUR/USD | 89 ms | +0.0 pt | 0/3 fills worse |
| GBP/USD | 88 ms | +1.0 pt | 2/3 fills worse |
| XAU/USD (Gold) | 83 ms | +9.7 pt | 2/3 fills worse |
Order latency and slippage measured with real orders in a CALM market — a news-time fill inherits the quote gaps above on top of this baseline.
All six instruments, minute by minute
Each cell is the average spread of that minute as a multiple of the quiet median — the instant peaks in the table above are single quotes inside minute 0, so a ×4 minute average and a ×18 instant peak describe the same event.
| Minute vs release | -3 | -2 | -1 | +0 | +1 | +2 | +3 |
|---|---|---|---|---|---|---|---|
| EUR/USD | ×1.0 | ×2.9 | ×3.2 | ×4.2 | ×2.4 | ×1.0 | ×1.0 |
| GBP/USD | ×1.0 | ×1.8 | ×2.3 | ×3.0 | ×1.8 | ×1.0 | ×1.0 |
| AUD/USD | ×0.8 | ×1.8 | ×2.9 | ×3.3 | ×1.7 | ×0.8 | ×0.7 |
| USD/CAD | ×0.7 | ×1.7 | ×2.9 | ×3.9 | ×1.6 | ×0.7 | ×0.8 |
| USD/JPY | ×1.0 | ×2.5 | ×4.0 | ×5.8 | ×3.1 | ×1.0 | ×1.0 |
| XAU/USD (Gold) | ×1.0 | ×2.3 | ×2.4 | ×4.2 | ×1.9 | ×1.0 | ×1.0 |
Tick density across instruments ran ×1.3–×2.4 the usual rate in the release minute; everything is flat again by about +2 minutes.
Honest context: news vs the daily rollover
| Instrument | This release peak (pips) | Worst daily hour (rollover) | Which is worse? |
|---|---|---|---|
| EUR/USD | 3.7 | 8.3 | rollover is worse |
| GBP/USD | 5.3 | 15 | rollover is worse |
| AUD/USD | 4.9 | 25 | rollover is worse |
| USD/CAD | 6.2 | 25 | rollover is worse |
| USD/JPY | 8 | 16 | rollover is worse |
| XAU/USD (Gold) | 150 | 175 | rollover is worse |
On several majors the everyday rollover hour is wider than a top-tier release — one more reason the quote gap, not the spread, is the real news risk.
Stops that survive the spike
FxPro’s measured stops level is 0 points on all six instruments — stops and straddle orders can sit at any distance. The practical floor during a release is the spread itself: at this print an EUR/USD stop closer than about 5.6 pips could have been filled by the spread alone, with price effectively unmoved, and a gold stop inside 225 points ($2.25) was equally exposed. The 1.5× buffer over the measured peak covers the post-peak chop before quotes settle.
Every release we have measured (9 windows since Jul 2026)
| When (AEST) | Ccy | Release | Actual vs forecast | Biggest widening | Recovery |
|---|---|---|---|---|---|
| 7 Jul, 00:00 | USD | ISM Non-Manufacturing Prices Paid, ISM Non-Manufacturing PMI | — | USD/JPY ×6.3 | 43 s |
| 6 Jul, 23:45 | USD | S&P Global Services PMI | S&P Global Services PMI: 51.2 vs 50.9 f'cast | AUD/USD ×6.0 | — s |
| 2 Jul, 22:30 | USD | Initial Jobless Claims, Average Hourly Earnings m/m, Nonfarm Payrolls | Initial Jobless Claims: 215 vs 209 f'cast; Average Hourly Earnings m/m: 0.3 vs 0.3 f'cast | USD/JPY ×26.7 | 98 s |
| 5 scheduled speeches / quiet events (10-Year Note Auction, BoE Governor Bailey Speech, BoJ L Money Stock y/y, ECB President Lagarde Speech) | no measurable reaction (<×1.5) | — | |||
This history grows automatically: every high-impact release is captured once and kept.
How this was measured
- Release times come from the platform’s economic calendar (high-impact only).
- For every release, all ticks 15 minutes either side are pulled from FxPro’s own MT5 feed and reduced to a per-minute spread curve.
- Recovery = first quote after the peak back under 1.5× the quiet median.
- Figures refresh on a schedule and vary release to release.