The setup was clean: price pressing a well-tested level, momentum aligned, invalidation parked just beyond the noise. Then 8:30 a.m. in New York arrived, an inflation print landed hotter than forecast, and price knifed through three levels in under a minute. Nothing was manipulated and nothing "broke". A scheduled event simply did what scheduled events do.
Every week, a handful of releases carry enough weight to reprice entire asset classes in seconds. The dates and times are published far in advance on any economic calendar; only the numbers themselves are unknown. That combination — known timing, unknown content — concentrates an unusual amount of uncertainty into a single timestamp.
| Event | Cadence | What it tends to move |
|---|---|---|
| CPI (US inflation) | Monthly | Anything priced off interest-rate expectations: FX, gold, indices, crypto |
| Non-farm payrolls (NFP) | Monthly, usually the first Friday | USD pairs first, then risk assets broadly |
| Fed rate decision (FOMC) | 8 scheduled meetings per year | Broad repricing; the press conference often adds a second wave |
| Other central banks (ECB, BoE, BoJ) | Roughly every 6 weeks each | Their own currency and rate-sensitive assets |
| Company earnings | Quarterly per company | The single stock; index-movers when the company is large enough |
Support and resistance work because real orders cluster around remembered prices. That is the mechanism — and high-impact events switch it off. In the minutes before a top-tier release, market makers widen spreads and pull resting orders, because nobody wants to be the one offering liquidity a half-second after a surprise. The order book thins out exactly when the largest burst of market orders is about to arrive.
When the number hits, the first move is not a chart pattern; it is information being repriced. A thin book means even routine repricing travels far, slicing through levels that would comfortably hold on a normal afternoon. Triggered stops become market orders that fuel the next leg, and the initial spike frequently reverses within minutes as the details — core versus headline, revisions, forward guidance — get digested. None of this means technical analysis is wrong. It means the conditions technical analysis depends on are temporarily suspended.
An illustrative example. A trader goes long EUR/USD at 1.0830 off a well-defined support zone, with an ATR-based stop 40 pips away at 1.0790 — sensible sizing, defined risk, planned loss of 1R if wrong. On a quiet Tuesday, that plan means what it says: either the level holds, or the stop cuts the loss at approximately −1R.
Now rerun the same trade into a CPI morning. The print lands hot, the pair gaps through the zone, and the stop — a market order once touched — fills at 1.0760 instead of 1.0790. The planned 1R loss became 1.75R, not because the analysis was worse, but because the risk math quietly assumed continuous prices and normal liquidity. Events break that assumption. Position sizing was correct, the level was real, and the loss still exceeded plan — the only variable fully under the trader's control was exposure at event time.
A few habits remove most of the damage. Check the calendar before every entry — the question is not "is there news today?" but "is there a top-tier release inside my expected holding period?". Treat flat as a position. Skipping a binary event costs nothing but variance; there is no prize for being on the field when a coin gets flipped. If holding through, size for the gap, not the chart — assume the stop can fill well beyond its level and shrink exposure accordingly. Distrust the first spike; initial moves often reverse once details are read. And expect the aftermath to linger: volatility stays elevated after big prints, stops sized on pre-event ATR go stale, and the market's whole mode can shift — the high-volatility regime described in the market regimes guide.
This caution is built into the scoring rather than left to discretion. In the published methodology, a high-impact calendar event inside the next 24 hours costs a setup 15 points — usually the difference between alerting and staying silent — and clusters of negative headlines subtract further. After the deterministic score, a separate AI review stage reads the actual headlines and calendar attached to the setup and can veto the alert outright, a central bank decision inside 24 hours being a standard example.
Many otherwise-clean charts die in that stage, and that is the intended behaviour: a clean chart twelve hours before a rate decision is not a clean trade. It is one reason alerts are rare by design — and every alert that does survive the calendar check is logged permanently on the track record, wins, losses and expiries alike.
SRA Quant reads the macro calendar before it alerts. Setups within 24 hours of a high-impact event are penalised or vetoed — silence included, by design.
SRA Quant provides market analysis and educational content only. Nothing on this page or the platform constitutes financial advice, and past or backtested performance does not guarantee future outcomes. Trading involves substantial risk of loss.