News Trading Risk: How to Protect Your Book During NFP

It is 1:29 PM London time on the first Friday of the month. Non-Farm Payrolls (NFP) drops in sixty seconds. Your dealing desk knows what is coming. Spreads will widen. Volatility will spike. The price will move 30, 50, maybe 100 pips in the first seconds after the release. This is not a surprise. It […]

It is 1:29 PM London time on the first Friday of the month. Non-Farm Payrolls (NFP) drops in sixty seconds.

Your dealing desk knows what is coming. Spreads will widen. Volatility will spike. The price will move 30, 50, maybe 100 pips in the first seconds after the release. This is not a surprise. It happens every month.

What your dealing desk may not know is which accounts on your platform are positioned specifically to exploit the next thirty seconds. The news event is not the risk. The risk is the accounts that treat the news event as an execution window.

The 30-Second Window: Infrastructure vs. Market Risk

When a high-impact economic release hits—NFP, CPI, FOMC—the market does not move smoothly. It jumps. Liquidity evaporates. Price discovery becomes chaotic as different venues update at different speeds.

During this window, three things happen simultaneously on your platform:

1. Feed Lag: Under normal conditions, your pipeline adds 20-50 milliseconds of latency. During a news spike, this can expand to 200-500 milliseconds. For that half-second, your clients see a price that no longer exists.
2. Spread Stale-ness: LPs widen spreads to reflect uncertainty. If your bridge delay lags behind the LP, your quoted spread may not reflect the true cost of execution.
3. Processing Queues: The burst of orders creates a queue. The gap between order receipt and execution widens, and that gap is where the exploit lives.

Calendar-Aware Risk: Why Static Thresholds Fail

Most broker risk systems use static thresholds. They apply the same rules at 1:30 PM on NFP Friday as they do at 3:00 AM on a Tuesday. This leads to two critical failure modes:

• False Negatives: Every account’s PnL moves during a spike. Thresholds fire on dozens of accounts, and the genuinely toxic traders hide in the noise of “alert fatigue.”
• False Positives: High thresholds avoid fatigue but miss the accounts designed to operate within the event window—those whose activity is modest in size but surgically timed.

Calendar-aware risk adjusts the detection framework based on the event schedule. It knows that the 30 seconds around the release are a different operating environment with different signals.

Event Toxicity: The Consistent Exploiter

The key distinction is between accounts that trade during news events and accounts that trade because of news events.

• The Legitimate Trader: Positions based on a market thesis (e.g., “I think the number will beat expectations”). Their activity is spread across the minutes before and after the release.
• The Event Exploiter: Has no view on the data; they have a view on your infrastructure. They target the 5 seconds before/after to capture the lag.

The Metric that Matters: Measure the timing distribution across multiple events. An account that consistently trades within a 5-second window of releases across dozens of events is running a timing operation, not making economic forecasts.

Pre-Event vs. Post-Event Patterns

Event exploitation is not monolithic. To protect your book, you must identify which of these two patterns an account is utilizing:

The Lag Adjustment: Normalizing Slippage

During a news spike, the “physics” of execution change. A 50-millisecond delay that costs a fraction of a pip in a quiet market can cost 5 pips or more during NFP.

Without lag adjustment, your execution metrics are meaningless. You must normalize raw slippage numbers against market velocity. This allows you to distinguish between normal event-driven slippage (the market moved, everyone got slipped) and exploitative slippage (the account captured the favorable side of the move during the lag).

A Practical Framework for Protection

Protecting your book isn’t about widening spreads until no one can trade. It is about forensic transparency:

1. Map Activity: Align every trade against the high-impact economic calendar.
2. Measure Intent: Calculate markout curves for event trades versus non-event trades.
3. Audit the Gap: Identify accounts whose performance is consistently favorable only during high-volatility windows.

The calendar is not a surprise. If you know when the events are, you can measure who is exploiting them.

Trust the Math. Protect the House.

See event toxicity in your data →