# The Framework Ate the Week — May 26–29, 2026
*Your weekly market recap, brought to you by [SignalDeck.live](https://signaldeck.live).*
This was the week the momentum framework fired on nearly every cylinder. A drone company delivered the single largest one-day hit in the running log. A software infrastructure cluster exploded into full mania Friday, capping a week where A6.4 mega-cap structural raisers went on a rampage. The one miss — a velocity-only pick with no fundamental backing — reminded us exactly why the quality floor exists. Nine rules were upgraded, confirmed, or born from new evidence. Not a bad five trading days.
## The Top Calls
**UMAC (Unusual Machines)** · Computer Hardware · SP Phase 2 ENTRY, picked 5/27 · RS 95.3 entry → RS 99.8 at close · Week +115%, Month +132% — The biggest single-day hit in the running log: +57% in one session. Pentagon Phase II drone program announcement landed directly on top of a hot Computer Hardware cluster already running hot. This is what the drone cohort thesis looks like when it fires: sector policy event + industry momentum + strong fundamentals = asymmetric single-day move. Day 4–6 second-entry window opens June 3–5 — the follow-through pattern from prior cohort trades suggests the real institutional sizing-in is still ahead.
**SNOW (Snowflake)** · Software Infrastructure · EP pick 5/28, A6.4 Day 2 · RS to 98.6, Month +67.7% — +36% in a single session on Day 2 of the AI data-infrastructure raise cycle. The structural re-rating thesis: when a mega-cap raises guidance and signals multi-year AI infrastructure demand, institutions don’t finish buying in one session. SNOW held its entire PM gap into close. A6.4 archetype confirmed for the 3rd time — every instance has delivered 3+ session continuation. Watch for Day 4–6 second-entry window June 3–4.
**DELL (Dell Technologies)** · Computer Hardware · EP pick 5/29, A6.4 mega-cap raise · RS 99.6, Month +105% — +33% in the session, CFO dead flat. A6.4 the fifth confirmed instance. The AI infrastructure earnings cycle is showing a pattern: DELL/SNOW/OKTA reporting back-to-back created a cluster feedback loop where each print repriced adjacent names upward. OKTA +30% on Day 2 wasn’t even in our top-5 — it was carried by the cohort. This is how mania builds. Day 4–6 window: June 3–5.
**RDW (Redwire)** · Aerospace & Defense · EP pick 5/26, R10.1 Day 5 · RS built through the week — The highest-magnitude R10.1 second-entry in the running log: opened -10% from prior close on a PM gap fakeout, then ran +40% intraday close-from-open. The stack: SpaceX IPO public filing + NATO Penguin Mk3 + Army Stalker $15M contract all hit simultaneously. R10.1 is now 8/8 since we started tracking. The “PM gap → deep open → institutional reverse rally” is the canonical entry pattern. Nobody catches this without watching the tape — that’s the edge.
**IBM (IBM)** · IT Services · SP Phase 2 ENTRY, picked 5/22 with 3.48x volume · RS 86.8 → RS 94.0 by Friday, Week +17.8%, Month +31.2% — The understated call of the week. Institutional bid was visible on May 22 in the volume data — 3.48x average volume is the institutional accumulation fingerprint. Held through mid-week relative strength chop and then exploded Friday in the software infrastructure cluster bid. Low-volatility, high-conviction compounding is the best kind of trade.
**HBM (Hudbay Minerals)** · Copper · SP Phase 1 STEALTH 5/27 · RS 79.4 → 93.8, Week +17.9%, Month +29.9% — Phase 1 STEALTH → Phase 2 conversion in two trading sessions. The copper thesis we flagged Tuesday is now fully in Phase 2 territory. Institutional accumulation was visible in the RS velocity before the price moved. Two-week compounder thesis intact — mining/metals cluster hasn’t gone vertical like tech, which means there’s still room to run.
**ZVRA (Zevra Therapeutics)** · Biotechnology · SP Phase 1 STEALTH 5/27 · RS 77.7 → 86.9, Week +10.7%, Month +20.1% — Second Phase 1 STEALTH → Phase 2 conversion of the week. The adjacent-cluster framework predicted this: when mania concentrates in AI/hardware, biotech with real fundamentals gets re-rated quietly. FQ 30, EPS QoQ +1,152%, EPS surprise +640%. This is the kind of name that doesn’t get noticed during a hardware mania week — and that’s exactly the opportunity.
> Phase 1 STEALTH → Phase 2 transition rate this week: 3/3 (HBM, ZVRA, VVX) when FQ ≥ 22. The framework caught all three before the price moved.
## The Quiet Beats
**CRDO (Credo Technology)** · Semiconductors · SP Phase 2 5/27 · RS 93.6 → 95.2, Week +22.2%, Month +34.4% — Steady institutional accumulation, no mania spikes. FQ 27 with EPS QoQ +408%. This is a multi-week compounder, not a momentum chaser. The fundamentals are real and the RS trend is intact.
**VVX (V2X)** · Aerospace & Defense · SP Phase 1 STEALTH 5/22 · RS 85.9 → 92.1, Week +16.1%, Month +27.0% — Called on May 22 as Phase 1 STEALTH with FQ 27. By Friday it’s Phase 2 territory. Defense-adjacent, not mania, no distribution pressure. The slow grind is working.
**KLAR (Klaviyo)** · Software Infrastructure · Watching since 5/18 · RS 90.3 → 93.7, Week +14.8%, Month +35.9% — Nine-for-nine on the momentum build since the May 18 Phase 1 STEALTH call. Software-Infra cluster mania Friday carried KLAR with it. FQ 30. Not in R14.3 extended territory. This name still has legs going into next week.
**DY (Dycom Industries)** · Engineering & Construction · SP Phase 2 ENTRY 5/27 · RS 94.6, Week +23.0%, Month +30.6% — The sleeper pick from the May 27 scan. Engineering & Construction is the cluster nobody talked about while Software-Infra exploded. FQ 28, massive RS expansion from 30→95 in a week, vol slope confirming institutional buying. Not in mania territory. Best risk-adjusted setup going into next week paired with HBM.
> The adjacent-cluster thesis is paying off: as AI hardware goes vertical, the infrastructure build-out (construction, copper, power equipment) gets a bid that nobody is writing about yet.
## Yesterday’s Lessons
**VAL (Valaris)** — The week’s clearest teaching moment. Picked May 22 as Phase 1 STEALTH on the basis of “fastest RS velocity on board” at +5.95/day. By May 27 RS had collapsed from 53.6 → 15.3 (down -38 points in 5 sessions). Price -17.2%. The lesson is not subtle: velocity without fundamental quality backing is a lottery ticket, not a thesis. FQ was 20. Oil & Gas Equipment & Services has no cluster tailwind. The FQ ≥ 22 floor on Phase 1 STEALTH picks was reinforced permanently from this miss — it’s now the first filter, not an afterthought.
**LASR (nLIGHT)** — RS collapsed from 88.8 → 56.4 in two sessions after a clean May 27 Phase 2 entry with FQ 30. Investigation needed: likely an earnings-driven R14.1 distribution event between sessions. The lesson is that FQ 30 does not protect against a bad print during the hold period — always check the earnings calendar for the holding window.
**CHA (Champion Power Equipment)** — The model said CAUTION (EP-β 35) on a stock it misread as buyback-speculation. Reality: +25.59% day, CFO +9.53%, RS jump from 27 → 90.5 in a single session. The $150M buyback announcement was 10% of a $1.45B float — that’s structural floor support, not corporate theater. R7.4 was born from this: when a small-cap (<$5B) announces a buyback ≥7% of float alongside an earnings miss, score it as a structural catalyst with full EP treatment.
**PONY (Pony AI)** — EP-β 83 STRONG on a post-print setup. Then a Macquarie mid-session PT cut triggered distribution all session from the +20% PM gap. New rule R14.4: analyst counter-action during the trading session = base case R14.1 distribution, regardless of how good the morning fundamentals looked. Check earnings calendar AND analyst coverage schedule.
**Hit-rate by thesis type this week:**
- **A6.4 (mega-cap structural raiser):** 3/3 · 100%
- **R10.1 (second-entry):** 1/1 · 100% · 8/8 all-time
- **R13.6 (drone/sector-policy cohort):** 4/4 · 100%
- **R8.1 (rubber-band snap):** 2/2 · 100%
- **SP Phase 2 ENTRY (FQ ≥ 24):** 10/15 · 67%
- **SP Phase 1 STEALTH (FQ ≥ 22):** 4/8 · 50%
- **Velocity-only Phase 1 (no FQ):** 0/1 · 0%
> **FLAG:** Phase 1 STEALTH hit rate is 50% even with FQ ≥ 22 floor applied. The pattern fires consistently — when it fires. The issue is the ones that fade for unrelated reasons (cluster reversal, earnings surprise during hold). Position sizing should reflect this: Phase 1 STEALTH = half the size of Phase 2 ENTRY.
## The Read
**Best call of the week:** UMAC on May 27 — +57% in a single session. Pentagon drone cohort catalyst + hot industry cluster + FQ 30 fundamentals. The framework caught it before the news hit. That’s what this is for.
**Worst call of the week:** VAL on May 22 — picked on raw RS velocity with FQ 20 and no cluster tailwind. RS collapsed from 54 to 7 by Friday. Velocity without fundamentals is not a thesis. It’s noise wearing a thesis hat. The FQ floor exists precisely to prevent this.
**Best two-week compounding setup going into next week:** DY (Dycom, Engineering & Construction) + HBM (Hudbay, Copper). Both are in adjacent clusters that haven’t gone vertical. Both have FQ 28+ with institutional RS accumulation confirming. Neither name is getting attention this weekend while everyone writes about DELL and SNOW. That’s the setup.
**Don’t chase into Monday:** The entire Software-Infrastructure mania cluster — DELL, SNOW, OKTA, NTAP, SAIL, RBRK. Fifty-three names are in the elevated-bar zone heading into the weekend. Every one of them will face distribution pressure on any Monday morning gap. The mania is real, the move is real, and the entry window for new capital closed Friday afternoon. The trade now is the adjacents, not the leaders.
**Lesson banked:** A6.4 mega-cap structural raisers are the highest-conviction thesis in the framework right now. Five confirmed instances, 100% hit rate, R14.3 elevated-bar fade formally disabled when it fires. When a mega-cap with institutional ownership raises guidance while signaling multi-year AI infrastructure demand, institutions don’t finish buying in one session. The first gap is not where you sell. DELL, SNOW, OKTA all confirmed this. The trade is: buy the first close, hold to Day 4–6, size appropriately, exit on the Day 4 re-acceleration.
---
**Not financial advice.** This is independent market analysis and momentum research, not a recommendation to buy or sell anything. Markets are volatile, individual securities can lose 100% of their value, and past performance does not predict future results. Do your own due diligence, size positions you can afford to lose, and consult a licensed financial advisor before making investment decisions. Your money, your risk, your call.