SPY slips 0.23% as 18 of 107 setups break out; GDX is today’s pick

The UnxEdge breakout scanner flagged 107 wedge pattern setups today, but only 18 actually broke, which tells you the tape had motion without follow-through. In a market like this, wedge breakout quality matters more than raw count, and today the cleanest pressure sat on the bear side even as index damage stayed contained.

Macro drove the split tape. New U.S. action on aluminum imports kept the tariff and input-cost story alive, which matters directly for industrials and manufacturing-sensitive names, while headlines around fresh U.S. strikes on Iran and air defense activity near Bushehr kept geopolitical risk elevated into the close. That combination helped sustain defensive and commodity-linked relative strength while cyclicals and semis continued to lose sponsorship.

Sector rotation made the message clearer than the indices did. Retail is rotating in week over week while semis are rotating out, which is not what a healthy risk-on market looks like. If that divergence persists, traders should keep respecting weakness in chip names and stop assuming every dip in AMD or AMAT gets bought on schedule. For the bigger picture, revisit yesterday's debrief.

Market Snapshot

Flat headline tape, weak internals

The scanner leaned heavily bearish with 77 bear candidates against just 30 bulls, and that bias was directionally right even if the major indices did not crack much on the surface. IWM taking the biggest hit while QQQ held flat is another sign this was selective selling, not indiscriminate panic.

Grade quality was decent, but resolution quality was not. Eighteen breakouts against 28 failures is a poor conversion rate, and that usually means traders should shrink expectations, tighten filters, and avoid chasing second entries. For current grades and trigger levels, see live setups in the scanner.

Yesterday's Pick

BX bull failed cleanly

No sugarcoating it, yesterday's BX long was wrong. The setup never delivered enough upside acceptance to justify the risk, and the stop did its job before the tape got any sloppier.

Arxe Pick of the Day

GDX got the nod, but this was a tactical short, not a conviction swing

GDX scanner chart – July 20, 2026 | UnxEdge

The case for GDX was simple: among the bear candidates, it had the best statistical edge by a mile. That matters more on a day when the market is not offering broad upside participation and the scanner is skewed heavily toward downside structures.

The wrinkle is macro. Gold as a sector has been weak versus SPY over both the five-day and week-over-week lens, which supports the short, but geopolitical escalation around Iran can inject random bid support into precious metals at any point. That makes GDX tradeable, not comfortable. For live grades and breakout levels, see live setups in the scanner.

There was no Arxe swing pick generated today, which is the right call when conviction is statistical but the macro tape can still interfere with clean follow-through.

Bot Performance

Paused, and that is better than forcing bad trades

Wex and Xcel are in training mode while strategy is being refined. The scanner and Arxe intelligence remain fully active. When trading resumes, every trade will be published here. You can watch Wex and Xcel trade live in the Edge Lab.

Missed Trades

The only real miss was JD. The rest were scraps.

Total missed upside was +3.33R, but almost all of that came from one name. JD was the genuine blind spot because it delivered a full TP3 move without requiring abnormal volume. That is the kind of miss worth studying.

The rest were not painful misses. Low-RVOL drifts in ON, CRWD, BA, and NOC produced fractional returns that do not justify blowing up filters. If anything, those outcomes argue the system should stay selective instead of loosening standards just to capture noise.

Breakout Scorecard

Too many breaks, not enough commitment

A 39.1% breakout rate is weak. It means traders pressing every trigger got chopped, and only the names with real directional sponsorship had a chance to pay.

Notable names that did break included BA bear A-, C bear A-, CAT bear A-, CRM bull B, CRWD bear A-, and CSCO bear A-. Notice the imbalance: the better quality names still skewed bearish, which matched the scanner bias better than the headline index action did.

Watch Tomorrow

The best unresolved names are still on the bear side

These are the setups that still matter if the market opens with the same posture tomorrow. Levels below are framework levels built from the unresolved wedge structure, and traders should confirm live trigger precision before entry. For live updates on setup grades and exact breakout levels, see live setups in the scanner.