The evidence behind the calls.

Across 24,993 next-day trades on US and EU movers, blindly chasing a big move is a coin flip that loses to costs. The edge only appears when the move carries confirmation — volume expansion, an existing multi-day trend, and a strong close.

About 63% of qualifying EURUSD breakouts extend and 37% fake out — but that headline hides everything. The size of the breakout candle is the real discriminator: weak breaks fail six times in ten, decisive ones succeed nine in ten.

We hand-inspected all 195 A-grade EUR/USD level breaks over five months. Direction relative to the live multi-day thrust decided almost everything: breaks taken with the thrust worked about 85% of the time, breaks taken against it about 13%. A rich cluster inside a live trend behaved as a launch pad, not a wall.

Across 1,462 graded EUR/USD level breaks over five months, breaks from a genuine multi-signal cluster followed through ~62-66% of the time — roughly double the rate of breaks from a single isolated line. Popular 'confirmations' like session timing and volume barely moved the needle; the confluence itself did the work.

The stop-hunt narrative says a wick beyond a swing high is a reversal signal. Across 4,986 EURUSD sweeps, only ~30% reversed — nearly 68% continued in the sweep direction. Most 'stop hunts' are just the level genuinely breaking.

Buying the dip in a EURUSD trend is a near coin flip: of pullbacks that resolve within a day, 48.7% continue and 51.3% reverse. But three measurable traits — how deep, how fast, and what time — split the winners from the losers cleanly, and the most-taught rule (wait for the EMA20) is worse than a coin flip.

The instinct at a range boundary is to fade it back toward the middle. Across 3,111 EURUSD range-boundary touches over 16 years, only 14.6% of resolved touches reverted to the opposite side — 85.4% broke out instead.

Where gold's Asian session closes inside its own overnight range foretells which side the London open breaks first. Across 64 breaking sessions from January to August 2026, a close in the top third of the range preceded a London break of the Asian high 97% of the time; a close in the bottom third preceded a break of the low 100% of the time.

Across 79 sessions where London broke gold's overnight range, 61% were back inside the range two hours later. The depth of the first poke past the level tells you nothing about which outcome you're in — how far the break subsequently extends tells you almost everything.

Trading gold's first-hour breakout looks obvious and mostly loses. Across ~7,900 sessions the London open produced a clean breakout only 17% of the time and a whipsaw ('Judas') 47% of the time — while the New York open was structurally cleaner.

Gold ranges roughly three-quarters of the time and trends only ~17% — and its trends are short while its ranges are sticky. Add a 97% Asian-range sweep rate and volatility that has multiplied since 2024, and gold's 'personality' comes into focus.

Small S&P 500 gaps fill almost every time; large gaps mostly don't. Across 418 gap events, sub-5-point gaps filled 96% of the session while gaps over 50 points filled just 20% — the size of the gap decides whether it's noise or signal.

The first 60 minutes after the 14:30 UTC NYSE cash open aren't just noise before the 'real' session starts. Across 606 cash-open events, the opening hour's direction matched the full session's direction 70.5% of the time — and the wider that opening hour, the more reliable the read.

We already found gold's opening-range breakout is a trap in London and cleaner in New York. The same session asymmetry shows up in the S&P 500 — and here it's even sharper: London ORB breaks round-trip (Judas) 53.4% of the time versus just 16.2% in New York.

The instinct in a volatility spike is to de-risk. The data says the opposite pays: S&P 500 forward-20-day returns are about 6× larger when volatility is extreme than when markets are calm — +2.06% vs +0.32%, and the win rate goes up, not down.