The short answer
We tested a news straddle on gold around 44 real US NFP and CPI releases (November 2024 to October 2026): a buy stop and a sell stop placed an equal distance from the price before 8:30 am New York time, one-cancels-other, with the stop loss back at the starting price and a target of twice the risk.
- Whipsaws were the norm. With orders $5 away, both sides were touched within the release hour in 30 of 44 events. The median range of the release hour was $31.50, and the largest was $104.50 (4 September 2026 payrolls).
- With normal costs and no slippage, the $5 straddle lost $83 per ounce and won only 22.7% of trades. The $10 straddle was flat (+$4).
- With news-time slippage, every version lost. At a $1.00 spread and $1.50 slippage, the $5 straddle lost $231 per ounce, and losses were up to 1.8 times the planned risk.
- Account impact: risking a planned 5% per release, a $10,000 account fell to $835 in that scenario. At 2% it fell to $3,864.
What a news straddle is
The idea is simple:
- A minute or two before the release, note the price, P.
- Place a buy stop at P + X and a sell stop at P − X, linked as one-cancels-other.
- The release blasts price through one order; the other is cancelled.
- Protect the trade with a stop (we used P, so the risk is X) and a target (we used 2X beyond the entry).
It feels direction-neutral, and on a clean one-way release it works beautifully. Our news trading lesson has a section on straddle risks; this page puts numbers on them.
The releases
We used the Bureau of Labor Statistics release archive for the Employment Situation (NFP) and CPI reports, all published at 8:30 am New York time. Between November 2024 and October 2026 that is 23 NFP and 23 CPI releases, including the delayed releases after the autumn 2025 US government shutdown (for example CPI on 24 October 2025 and NFP on 20 November and 16 December 2025). Two releases had no usable hourly bar in our data (3 July 2025, a gap in the data feed, and 3 April 2026, Good Friday), leaving 44 events.
Exact rules
| Rule | Hourly test (44 events) | 5-minute and 1-minute checks |
|---|---|---|
| Reference price P | Open of the 8:00 am New York hourly bar (a proxy: 30 minutes before the release) | Close of the 8:25 bar (M5) or 8:28 bar (M1) |
| Orders | Buy stop at P + X, sell stop at P − X, one-cancels-other | Same, X = $3 |
| Distance X | $5 and $10 | $3 |
| Order life | The release hour | 30 minutes |
| Stop loss | P (risk = X before slippage) | P |
| Target | 2X beyond the entry (2R) | 2X beyond the entry |
| Exit | Close of the 9:00 am bar (10:00 New York time) | 9:25 (M5) or 9:29 (M1) |
Cost scenarios
| Scenario | Round-trip spread and commission | Extra slippage on stop entries and stop exits |
|---|---|---|
| A: normal (unrealistic at news) | $0.30 | $0 |
| B: moderate news conditions | $1.00 | $1.50 each |
| C: fast news conditions | $1.50 | $3.00 each |
We cannot know your broker’s fills, so these are stated assumptions, not measurements. The point is to show how sensitive the straddle is to them.
Prices. COMEX gold futures (GC=F) from Yahoo Finance: 1-hour bars October 2024 to October 2026, 5-minute bars (4 releases, September to October 2026) and 1-minute bars (2 releases, October 2026). Fill logic. Inside an hourly bar we assume the usual path (up bar: open, low, high, close; down bar: open, high, low, close); the order on that path’s first extreme fills first, and only moves after the fill can hit the stop or target, with the stop counted first if both are possible. Hourly bars hide the exact sequence within the hour, which is why we also show the 5 and 1-minute checks. Results in US dollars per ounce.
The results
How violent were the release hours?
| Measure (44 events) | Value |
|---|---|
| Median high-to-low range of the release hour | $31.50 |
| Smallest / largest | $11.30 / $104.50 |
| Both $5 levels touched in the release hour | 30 of 44 |
| Both $10 levels touched | 11 of 44 |
Release-hour ranges grew sharply in 2026: every release from July to October 2026 had a release-hour range of $54 or more.
Hourly straddle results (44 events)
| Version | Scenario | Win rate | Avg per trade | Net per ounce | Profit factor | Worst loss vs planned risk |
|---|---|---|---|---|---|---|
| $5 straddle | A: normal | 22.7% | −$1.89 | −$83.20 | 0.54 | 1.06x |
| $5 straddle | B: moderate | 22.7% | −$5.25 | −$231.00 | 0.25 | 1.8x |
| $5 straddle | C: fast | 22.7% | −$8.41 | −$370.00 | 0.13 | 2.5x |
| $10 straddle | A: normal | 38.5% | +$0.11 | +$4.20 | 1.02 | 1.03x |
| $10 straddle | B: moderate | 35.9% | −$3.02 | −$117.60 | 0.65 | 1.4x |
| $10 straddle | C: fast | 33.3% | −$5.94 | −$231.60 | 0.45 | 1.75x |
The account: $10,000, sized from the planned risk
Each trade was sized so that a normal stop-out would lose 2% (or 5%) of the account. Slippage made real losses larger.
| Version | Scenario | Ending balance at 2% planned risk | Max drawdown | Ending balance at 5% planned risk | Max drawdown |
|---|---|---|---|---|---|
| $5 straddle | A | $7,062 | 29.4% | $3,962 | 60.4% |
| $5 straddle | B | $3,864 | 61.4% | $835 | 91.7% |
| $5 straddle | C | $2,173 | 78.3% | $182 | 98.2% |
| $10 straddle | A | $9,939 | 11.7% | $9,341 | 27.5% |
| $10 straddle | B | $7,772 | 22.7% | $4,996 | 50.6% |
| $10 straddle | C | $6,163 | 38.4% | $2,751 | 72.5% |
Exact checks on 5-minute and 1-minute data
| Data | Releases | Both $3 levels touched (first 15 min on M5, first 10 min on M1) | Trades won | Net, scenario B |
|---|---|---|---|---|
| M5 (4 Sep, 11 Sep, 2 Oct, 14 Oct 2026) | 4 | 3 of 4 | 1 of 4 | −$19.00 per ounce |
| M1 (2 Oct, 14 Oct 2026) | 2 | 1 of 2 | 1 of 2 | −$3.50 per ounce |
These are tiny samples, but they show the same pattern at a finer resolution: the 4 September payrolls release moved $104.50 in 15 minutes, and our long was stopped out on the way.
What the results say, in plain English
- The release usually picks both directions. With $5 orders, both sides were touched in two-thirds of release hours. A straddle with a stop at the starting price is stopped out every time price crosses back through it.
- A wider straddle only helps with perfect fills. At $10 the whipsaw rate dropped to 11 of 44 and the result was flat with normal costs. Add news-time slippage and it lost.
- Slippage breaks your position sizing. A trade sized to lose 2% can lose 3.5% or more when the stop entry and stop exit both slip. That is why the 5% rows collapse.
- Normal costs are the wrong assumption. Scenario A, the only one where anything broke even, is the one least like real conditions at 8:30 am on payrolls day.
Other risks a backtest cannot show
- Both orders filling. In a fast market, one-cancels-other can fail if both sides trigger within milliseconds, leaving you long and short.
- Requotes and rejected orders during the release second.
- Spreads of several dollars for the first seconds after the print, which can trigger a stop order without the market truly trading there.
- Platform freezes when many traders act at once.
What traders do instead
Most experienced gold traders either stay flat into the release or cut size sharply, then trade after the first move:
- Mark the high and low of the first 5 to 15 minutes after the release.
- Wait for a retest or a clear failure of one side.
- Size the trade from a stop beyond that post-release level.
Our NFP day guide, CPI day playbook and event-day playbooks set out step-by-step routines, and the economic calendar lists upcoming release times. Size every news-day trade with the position size calculator, and assume your stop may fill worse than its price.
Check it yourself
Open a 1-minute XAUUSD chart for the last three payrolls and CPI releases on your platform. For each, note the price at 8:28 New York time, then mark P + $5 and P − $5. Count how many times both levels were hit in the first 15 minutes. Then look at your platform’s spread at 8:30 if you can see it. Those two observations are the whole case against the straddle.
Learn it with other traders
News days are where traders most often break their own rules. In the XAUUSD Global community, members plan release days together, share their post-release levels and review what happened. The first month is free, then $19.99 a month.
Educational content only, not financial advice. Trading gold with leverage carries a high risk of loss.
Frequently asked questions
What is a news straddle in gold trading?
Just before a major release, you place a buy stop a few dollars above the current price and a sell stop the same distance below. Whichever side the release pushes price through gets filled, and the other order is cancelled (one-cancels-other). The idea is to catch the big move without predicting its direction.
Does straddling NFP or CPI work on gold?
Not in our test. Around 44 real releases, a $5 straddle saw both sides touched in the first hour 30 times. It lost $83 per ounce even with normal costs, and $231 per ounce with a $1.00 spread and $1.50 slippage. A wider $10 straddle broke even with normal costs and lost $118 with realistic news slippage.
Why do straddles lose money on news?
Three reasons: whipsaws (price spikes through one order, then reverses through the stop and sometimes the other order), slippage (stop orders fill worse than their price in a fast market), and wider spreads at the release. In our tests the worst loss per trade was 1.8 to 3.9 times the planned risk, depending on the slippage assumption.
What slippage should I expect on gold at NFP or CPI?
It varies by broker, liquidity and how big the surprise is, and we cannot measure your broker’s fills. We therefore show three scenarios: normal costs with no slippage, a $1.00 spread with $1.50 slippage, and a $1.50 spread with $3.00 slippage, so you can see how sensitive the result is.
Which releases did you test?
US Employment Situation (NFP) and CPI releases published by the Bureau of Labor Statistics from November 2024 to October 2026, all at 8:30 am New York time: 46 releases, of which 44 had usable hourly data. The schedule included delayed releases after the 2025 government shutdown.
What is a safer way to trade gold around NFP and CPI?
Many traders stay flat into the release and trade only after the first move has settled, using the post-release high and low as levels. Our news trading lesson and the NFP and CPI day playbooks describe those approaches.
Sources and further reading
- CME Group — Gold futures overview — Contract specifications and trading hours for COMEX gold
- CFTC — Commitments of Traders — Weekly futures positioning, including gold
- Investor.gov — Stop order — How stop orders work, and their limits in fast markets
External sites are independent of XAUUSD Global. Check dates and data yourself.