R14 IPP-leg deeper oversold — R13 frame confirmed and sharpening, posture unchanged

Update

R14 IPP-leg deeper oversold — R13 frame confirmed and sharpening, posture unchanged

  • Type: refresh (subtype: continuity-annotation)
  • event_id: 2026-05-19-ai-power-bottleneck-r14-ipp-leg-deeper-oversold
  • Source: R14 next-day research desk fanout — workflow/RUNS.md 2026-05-19 r14-next-day-research-desk-fanout. Inputs: published/research-notes/2026-05-19-r14-ai-power-ipp-leg-deeper-oversold.md; published/scan-result-buckets/2026-05-19-r14-next-day-research-desk-fanout.json.
  • Decision context: R13 (earlier today) concluded the IPP-leg de-rate was valuation unwind + IPP-specific regulatory overhang (PJM capacity collar; FERC co-location process uncertainty), NOT a demand collapse — equipment leg held on contracted backlog. One trading session later, the IPP leg is DEEPER oversold (CEG RSI 38 → 32, VST RSI 27 unchanged, TLN RSI 38; CEG 7D -10.1% → -11.1%). Does the deeper de-rate flip R13's frame, or simply confirm it?
  • Verdict — confirms R13's frame and sharpens it. NOT a flip. Posture unchanged. Three confirmations:
    1. No new fundamentals catalyst broke in the intervening 24 hours. No IPP earnings miss, no PJM capacity-collar extension news, no FERC Crane decision, no demand-side commentary from hyperscalers, no negative analyst house-view shift. The deeper de-rate is what R13's frame predicts when the rationale is regulatory/valuation and no counter-catalyst arrives.
    2. Equipment-vs-IPP relative performance widened, exactly per the R13 sharpening. Equipment leg held -6%/-6% 7D (BE/GEV) vs IPP -10% to -15% 7D (CEG/VST/TLN) — a 4-9 percentage-point spread in one session. The relative-performance dispersion IS the thesis playing out.
    3. Cohort split clean at the watchlist level. R14 ai-power watchlist breakdown is 33 up / 3 down / 9 flat across 45 names — the 3 down names ARE CEG/VST/TLN. The IPP-leg weakness is concentrated; not spread.
  • Effect on thesis: HOLD — posture unchanged. IPP leg stays weakening (not broken); CEG/VST/TLN kept in key_tickers (regulatory/valuation drag, not demand-side); equipment leg holding on contracted backlog (BE/GEV +18% / +2% 30D). R12 cohort-split frame still structurally correct. R13 sharpening (equipment leg is the cleaner expression for new entries) reinforced.
  • Operational gate clarification (for next refresh): the implicit "broken" threshold is approximately "RSI sub-25 + 30D < -25% for at least one of CEG/VST/TLN" — would be worth stating in the perspective body to avoid future ambiguity. Currently VST RSI 27 / -14.3% 30D is the closest to that gate; CEG RSI 32 / -6% 30D is approaching the RSI side but not the 30D side; TLN RSI 38 / -6.4% 30D is the mildest.
  • Watches unchanged (→ SCANS): FERC Crane / Three Mile Island injection-rights decision (June/July 2026); PJM capacity-collar extension through 2029/30 (13-governor push); CEG Calpine debt + interest-expense refresh. Both regulatory catalysts remain the cleanest near-term resolution catalysts.
  • Per-ticker (R14 tape):
    • CEG → $260.67 (2026-05-19), RSI 32 (was 38), -11.1% 7D, -6.0% 30D — deeper oversold; CEG-specific Calpine/Crane drag compounding
    • VST → $136.75 (2026-05-18), RSI 27 (unchanged), -10.1% 7D, -14.3% 30D — past deep-oversold; no buyer yet
    • TLN → $324.21 (2026-05-18), RSI 38, -15.5% 7D, -6.4% 30D — biggest 7D drop; the R13 "proof case" is now deeper underwater
    • BE → $261.34 (2026-05-19), RSI 41, -6.9% 7D, +18.3% 30D — equipment leg cooled but holds trend (Oracle 2.8 GW backlog cushion)
    • GEV → $1012.25 (2026-05-18), RSI 39, -5.7% 7D, +2.2% 30D — same pattern (83 → 100 GW gas backlog)
3 events
DateTypeEvent
May 4 ArticleCambrian Thesis — AI supply chain 2x2 May 29 UpdateEnergy Candidates May 29 ScanNVDA Ecosystem Scan

No direct external sources are attached to this read.