Zero names in this universe meet the strict screen (RSI < 30 AND price below its 20-day average) — the deepest RSI print in the whole 68-ticker set is ORCL at 31.3. That's the headline: most of the prior scan's oversold names have bounced out of the bin entirely (META, WMT, ADBE, PYPL, LULU, NKE all posted meaningful RSI recoveries). The one name that got worse, not better, is ORCL — its RSI barely moved (30.5→31.3) but its 30-day change deepened to -23.3%, which flips it from last scan's "earliest possible bounce" read into this scan's "still collapsing" bucket. NFLX is the other name still actively deteriorating (RSI 43.7→37.5, -42.6% from its 52-week high). Per desk doctrine, this remains a falling-knife screen, not a buy list.

Data as-of 2026-07-10 close. Source: summaries/bargain-bin.json, filtered for RSI < 30 AND vsSma20 < 0, then widened to RSI < 45 (the closest practical cluster, since the strict filter returned nothing) — ranked by thesis quality × discount depth. Prior scan: 2026-07-07-bargain-bin (2026-07-06 close).

The strict bargain-bin filter — RSI below 30 and price below its 20-day average — returns nothing this cycle across all 68 tracked names. The prior scan's two closest names to that threshold, ORCL (RSI 30.5) and WMT (RSI 34.0), have diverged hard: WMT recovered to RSI 44.2 as its price stabilized (-15.7% from high, up from -18.1%), while ORCL barely moved on RSI (30.5→31.3) even though its 30-day change deepened from a flat read to -23.3% — a real fresh leg down, not a plateau. That reclassifies ORCL from last scan's "earliest possible bounce" candidate into this scan's "still collapsing" bucket. NFLX is the other name actively worsening: RSI cooled from 43.7 to 37.5 and its distance from its 52-week high widened to -42.6%, with no basing signal in the data.

Everywhere else in the universe, the story is recovery, not further damage. META rallied hard (RSI 54.5→66.4, 7D +14.0%, 30D +17.8%) and has effectively bounced out of "bargain" territory altogether. ADBE, PYPL, LULU, and NKE all posted genuine RSI improvements alongside better 30-day reads — the basing pattern the desk wants to see before any name-level follow-up. INTU carries an explicit thesis: broken flag in the data (RSI 45.2, -66.4% from its 52-week high) — per desk doctrine this stays an avoid regardless of how cheap it screens; a bounce in price (vsSma20 +2.3%) does not override a broken thesis.

This bucket screens for cheap, not for buyable. regime labels (collapse/downtrend/pullback/basing) describe the chart, not a recommendation — a name in collapse with a flat or improving RSI is not the same setup as a name in collapse whose RSI and 30-day change are both deteriorating (ORCL, NFLX this cycle). thesis: broken is a hard override: no price level makes a broken-thesis name attractive, and it should be avoided regardless of discount. Per desk doctrine, this bucket exists to separate names that are basing (RSI turning, price stabilizing) from names still actively collapsing — the RSI number alone doesn't answer that question; the 30-day trend direction does.

(40 remaining tracked names — AAPL, AMZN, NVDA, TSLA, NOW, SNOW, DDOG, NET, TMO, LLY, ABBV, UNH, JNJ, SBUX, AVGO, AMD, ANET, KLAC, LRCX, ASML, V, MA, XYZ, PANW, CRWD, OKTA, HD, GE, KO, JPM, GS, BRK-B, SPGI, CAT, UPS, RTX, BKNG, ABNB, UBER, TTD — all sit at RSI 49-65.5, healthy and not oversold; omitted from a bargain-bin screen by design.)

Still collapsing (deteriorating, not basing) — ORCL and NFLX. ORCL is the more important flag: it looked like a bounce candidate last scan (RSI 30.5, +2.5% same-day) but the 30-day trend has since turned sharply negative (-23.3%) with RSI barely improved — this is the pattern of a knife still falling, not one that's found a floor.

Broken thesis — hard avoid — INTU. The data carries an explicit thesis: broken tag; no bounce in price changes that.

Approaching / unresolved — the RSI-45-to-50 cluster (COST, DIS, PEP, LOW, WMT, CME, ISRG, CEG, MRVL, DPZ, QCOM, HON, DE, MCD, GOOGL, CRM, MSFT, PLTR) is a mix of genuine early stabilization (WMT, CEG, DE, HON — all goldenCross-intact or improving 30D) and names still net negative on both RSI and price trend. None of these confirm a base yet.

Recovered — no longer bargain-bin material — META, PYPL, ADBE, LULU, NKE, WDAY, ZS all posted real RSI and/or 30-day improvements since the last scan. This is the most important structural read this cycle: the bin thinned because names bounced, not because the universe changed.