Verdict: NO POSITION. Deal announced Feb 23, 2026. Alpha window closed.

LR Signal: 1.0 — No divergence from market pricing. The deal IS the price. Arb carry of 1-2% over 2-3 months is available but outside our edge zone.

What Happened

Gilead Sciences bid $115/share cash + $5 CVR (non-transferable) for Arcellx, announced pre-market Feb 23, 2026. Friday close was $64.11. 79.4% premium. Tender offer structure, Section 251(h) merger (no shareholder vote required), expected Q2 2026 close.

The stock was at 35% of its 52-week range with 21.5% short interest, 9.8 days to cover, and a 10.7x put/call OI ratio on Dec 2026 options. Twenty analysts had a mean target of $110.53.

Gilead, which had invested $678M across equity purchases and collaboration payments, owned 11.5% of outstanding shares, and had completed manufacturing transfer to Kite, bought the dip.

Deal Terms (Source: GILD 8-K, Feb 23, 2026)

Factor Decomposition (Pre-Deal, 1Y Trailing)

ACLX Factor Regression (n=250 daily observations)

Factor        Beta      Variance %
SPY          +0.16         1.9%      No edge. Negligible.
MTUM         -0.26        -3.1%      Anti-momentum. Not edge, but setup signal.
XBI          +1.10        34.8%      No edge. Unintentional biotech beta.
Idio                      66.4%      BELOW 75% THRESHOLD.

alpha = -22.9% annualized (stock destroying value after stripping XBI)
sigma_idio = 41.4%
R-squared = 33.6%

66.4% idiosyncratic variance. Below the 75% target. One-third of return variance was biotech sentiment, not company-specific. A standalone position required XBI hedging. Post-deal, this decomposition is irrelevant — the stock becomes a function of deal spread mechanics (≈99% idio).

Variable Templates

V1: Deal Close Probability

V2: CVR Payout ($6B Cumulative Sales by YE 2029)

CVR market share math (the missing piece):

The 4L+ RRMM market is ≈$3.5B (Kite head Cindy Perettie, GILD Q4 2025 transcript, line 170). Carvykti is currently running at ≈$2.2B annualized (Q4 2025: $555M). That's ≈63% of the 4L+ market already captured by one BCMA CAR-T.

For anito-cel to hit $6B cumulative by YE 2029 with ≈2.5 years of sales (mid-2027 launch → Dec 2029):

Scenario A: Share theft from Carvykti in 4L+ only
  4L+ TAM: ≈$3.5B (stable)
  Anito-cel share needed: avg ≈$2.4B/yr = 69% of market
  Carvykti drops to: ≈$1.1B (from $2.2B)
  Requires: anito-cel takes MAJORITY of Carvykti's share
  Probability: LOW (15-20%). J&J won't cede quietly.
  Precedent: Carvykti took share from Abecma but Abecma was
  inferior on data. Carvykti's efficacy (98% ORR, 83% CR)
  is close to anito-cel's (96% ORR, 74% CR). Safety
  advantage alone unlikely to flip 69% market share.

Scenario B: 4L+ share + market expansion (2L/3L label)
  iMMagine-3 Phase III in 2L+ RRMM ongoing, results 2028+
  If approved by 2029: expands addressable from $3.5B to
  ≈$8-10B (earlier lines have 3-4x more patients)
  Anito-cel share needed: avg ≈$2.4B/yr = 24-30% of expanded market
  Probability: MODERATE (40-50%) — but requires BOTH
  positive trial AND FDA approval within CVR window
  Constraint: Bispecifics (Tecvayli +19% YoY, Talvey +73% YoY
  per JNJ Q4 2025) are entrenching in 2L/3L where convenience
  matters. Community oncologists can prescribe bispecifics;
  CAR-T requires ATC referral.

Scenario C: 4L+ share + OUS ramp (no label expansion)
  Kite commercializes OUS. But Kite's existing OUS infrastructure
  for Yescarta is thin and declining. Carvykti OUS: $128M/Q
  (Q3 2025, JNJ transcript) and accelerating.
  Anito-cel OUS likely 12-18 months behind US launch.
  Modest contribution by 2029: maybe $300-500M cumulative.
  Not enough alone.

Blended CVR probability: Scenario A (20%) × 15% + Scenario B (50%) × 45% + Scenario C (30%) × 25% = ≈33%

Prior estimate was 35-45%. Lowering to 30-40% after explicit market share math. Fair value: ≈35% × $5 = $1.75.

V3: FDA Approval (PDUFA Dec 23, 2026)

Irrelevant to deal close (Q2 2026, before PDUFA). Gilead absorbs approval risk. Affects CVR only.

V4: Pre-Deal Implied Probability — Explicit Math

Implied P = (Current - Downside) / (Upside - Downside)

Downside assumption: $5/share
  Rationale: In a BLA rejection, ACLX has $576M liquidity
  (≈$8.50/share) but $200M/yr burn rate and no alternative
  revenue-generating asset. Pipeline (ARC-SparX) is early.
  In biotech BLA rejections, stocks typically trade at
  0.3-0.5x cash. $5 is conservative but defensible.

Upside assumption: $130/share
  Rationale: Pre-deal analyst range was $82-$134.
  Successful approval + launch → NPV of 50% US profit
  share + OUS royalties on $3-5B peak sales franchise.
  $130 approximates the probability-weighted success scenario.

Implied P = ($64.11 - $5) / ($130 - $5) = $59.11 / $125 = 47.3%

Sensitivity:
  If downside = $10: P = ($64.11-$10)/($130-$10) = 45.1%
  If downside = $20: P = ($64.11-$20)/($130-$20) = 40.1%
  If upside = $115: P = ($64.11-$5)/($115-$5) = 53.7%

The estimate is sensitive to the downside anchor. Range is 40-54% depending on assumptions. The point stands: market was pricing well below the 85-90% BLA base rate regardless of which anchor you use. The delta (35-45 points) is robust to reasonable assumption changes.

V5: Gilead's Acquisition Economics — Corrected

The prior version subtracted ≈$1-2B in "avoided contractual obligations" from the headline price. This double-counts: those obligations are contingent on anito-cel's success, and in the success scenario the acquisition is worth far more than $7.8B. You can't discount the price by payments that only exist in the upside case.

Corrected framing:

The honest frame: Gilead is spending $7.3B incremental cash for the right to capture 100% of a franchise that could generate $3-5B peak annual revenue, instead of sharing 50% of US profits. In the success case, this is a 2-3x return. In the failure case, it's an $8B write-off. The deal is a leveraged bet on approval + commercial execution.

Positioning (As of Feb 21, 2026)

Short interest: 21.5% of float, 9.8 days to cover, ≈11.2M shares. Forced covering over 10+ trading days at 0.9M ADV. Thirty percent of effective free float must be bought back.

Aggregate short loss — caveated: Maximum loss = ($115 - $64.11) × 11.2M = ≈$571M. But this assumes all shorts entered at $64 or below. The stock traded between $48-$94 over the past year; average short basis is unknowable. Some shorts entered at $80-90 during the decline and are less underwater. $571M is the ceiling, not the expected value. The forced covering is real regardless — at $115, every short is deeply underwater no matter when they entered.

Options: Dec 2026 P/C OI ratio 10.73x. Massive put OI at $120 (190 contracts), $110 (157), $75 (406). Bearish positioning obliterated by deal price. MM hedge unwinds add mechanical buying pressure.

Locked shares: GILD 11.5% + insider/NEA/SR One support agreements 10.3% = 21.8% committed to tender. Effective free float: ≈40.8M shares, of which 11.2M is short.

Forced buyers on deal: Shorts (11.2M shares, margin calls), options MMs (50-90K delta shares, hedge unwind), Gilead (46M shares via tender). Nobody forced to sell. 68% premium eliminates seller motivation.

Insider Form 4s (P-code only): Zero open-market purchases in 12 months. CFO and officers sold routinely (vest-and-sell). CEO gifted shares. The "insider buy" was Gilead's $300M in equity purchases across two tranches at $28.75 and $61.68 — visible in ACLX's 10-Q Note 5.

Kill Test

Test 1 — Is the edge consensus restated? Yes. Twenty analysts had mean target $110.53. The view that ACLX was undervalued at $64 was literally consensus. No sell-side analyst would disagree with anything in this memo. The market wasn't informationally mispriced — it was structurally mispriced by positioning (21.5% SI) and sector drag (34.8% XBI variance).

Test 2 — Bear case engagement:

The bears had three arguments. All were partially right.

Cash burn tripling ($60M → $171M 9mo loss YoY): Real. But ignored Kite bearing CMC costs and sharing 50% of development expenses. ACLX's standalone burn rate overstated the economic reality of a partnered asset.

Carvykti entrenched: Partially right. Carvykti's accelerating trajectory ($286M → $555M quarterly, Q3'24 to Q4'25) shows it's winning despite bispecific competition. But anito-cel's zero neurotox is a genuine differentiator for the 4L+ population (median age 69, multiple comorbidities where safety drives physician choice).

Bispecifics displacing CAR-T: Wrong in 4L+ (Carvykti still accelerating alongside Tecvayli/Talvey). Partially right for earlier lines where bispecific convenience advantage matters. This constrains 2L+ label expansion — the variable that determines whether the CVR pays.

Test 3 — Kill condition:

Pre-deal: No informational edge (consensus agreed). Below idio threshold (66.4%). BUT: identifiable forced actors (shorts + Gilead as latent acquirer) and quantifiable delta (market implied 40-54% vs 85-90% base rate). Two kills, two passes. Mixed.

Post-deal: No delta. Kill confirmed. Deal price IS the price.

Unfilled Gaps

Commercial execution (UNFILLED, UNFILLABLE): Kite has never launched a BCMA CAR-T product. Their existing cell therapy franchise is declining (-6% Q4, guided -10% FY2026 per GILD Q4 2025 transcript, line 147). 99% manufacturing reliability and 16-day turnaround are strong process metrics (GILD Q4 2025 transcript, line 173), but process ≠ commercial execution. No primary source can resolve this before launch. This is the key uncertainty on the CVR.

Bispecific 2L+ competition (PARTIALLY FILLED): Bispecifics coexisting with CAR-T in 4L+ (Carvykti accelerating). But no data on head-to-head in earlier lines. iMMagine-3 results needed — not available until 2028+. Affects CVR, not deal close.

Impact: Both unfilled gaps affect only the CVR ($5, non-transferable, ≈$1.75 fair value). Neither affects the $115 cash component or deal close probability.

Pattern Extracted: Strategic Partner as Latent Acquirer

The actionable output of this analysis is not a trade on ACLX. It's a pattern — currently a hypothesis, not a validated screen.

The signal, in order of appearance (all from primary sources):

1. Partner takes equity at fixed price ($28.75/share, Jan 2023) — ACLX 10-Q Note 5
2. Partner expands collaboration scope (lymphomas, ARC-SparX, Nov 2023) — ACLX 10-Q Note 6
3. Partner doubles equity investment at 2x the original price ($61.68/share, Dec 2023) — ACLX 10-Q Note 5
4. Manufacturing transfer completed ("tech transfer US IND to Kite now complete") — GILD Q2 2024 transcript, line 73
5. Partner uses "best in class" language about unowned asset — GILD Q2 2025 transcript, line 65
6. Partner's division head, on Q4 2025 earnings call (Feb 10, 2026 — six weeks before bid), verbatim: "We expect over time to become the market leader given our excellent efficacy profile and differentiated safety profile, in particular with the delayed neurotoxicity in enterocolitis." And: "We will have the ability to serve the market at launch. With 99% reliability and sixteen-day turnaround time, which, again, is very differentiated from the existing products on the market today." — GILD Q4 2025 transcript, Cindy Perettie (Kite head), lines 171-173

This is not "potential market leader" or standard pharma hedging. This is the head of Kite cell therapy saying "we expect to become THE market leader" and citing proprietary manufacturing metrics for a product they don't fully own. On a public earnings call. Six weeks before bidding.

Known false positives (the denominator problem):

The pattern has false positives. Equity + collaboration + escalating commitment does NOT always end in acquisition:

Differentiating feature: In every false positive, the partner did NOT (a) increase equity at rising prices, (b) expand scope, AND (c) complete manufacturing transfer. The GILD/ACLX pattern had all three PLUS ownership language on earnings calls. The false positives stopped at step 1 or 2.

This is still a hypothesis. Sample size of 1 true positive and 4 false positives is not a screen. It's a pattern to track. Needs more observations before it becomes a tool.

Operational trigger (if we decide to monitor):

• Trigger: Partner files Schedule 13D/G showing >10% ownership in clinical-stage biotech
• Action: Add partner's ticker to transcript monitoring list alongside the target
• Scan each earnings call for: (a) Equity increases at higher prices, (b) scope expansion language, (c) manufacturing transfer mentions, (d) ownership framing ("we expect to become market leader" vs "our partner is developing")
• Alert if: Steps 1-4 all occur AND target is trading below 50% of 52-week range AND short interest >15%
• Frequency: Every earnings call until partnership dissolves or acquisition occurs

Counterparty Summary

Capital Allocation Decision

Action: PASS.

No alpha on the $115 cash component (deal announced, gap closed). CVR is non-transferable with ≈$1.75 fair value and unfillable execution gaps. Merger arb spread of 1-2% over 2-3 months is not our game.

The only residual question is GILD as acquirer. At $151 / $188B market cap / 9x forward earnings, they're spending $7.3B incremental for a potentially $3-5B peak revenue franchise. If anito-cel hits, this is accretive by 2028. But GILD is fully covered by 30+ analysts, 70% idio (below threshold), and the acquisition's impact on EPS is modest (≈3-4% of revenue). No edge.

Log the pattern. Build the transcript surveillance trigger. Move on.