Apex Capital Partners · Phase 3 Discovery · Deep Dive

Schrödinger
The Computational Engine

Physics-based molecular design at the convergence of quantum mechanics, machine learning, and the largest AI pharma partnerships on earth. The pick-and-shovel play on AI drug discovery.
Price$11.81
Market Cap$870M
Cash Position$352–391M
Enterprise Value~$479–518M
Revenue (TTM)$207.5M
52w Range$11.15 — $28.47
EarningsFeb 25 ★
Alpha v271
Section I

What Schrödinger Actually Does

Schrödinger is not a biotech company. It is not a SaaS company. It is both — and that hybrid identity is precisely what the market fails to value correctly.

Founded in 1990, Schrödinger built a physics-based computational platform that simulates molecular behavior using quantum mechanics. Where most AI drug discovery companies train models on historical data and predict outcomes statistically, Schrödinger calculates from first principles — the actual physics of how atoms interact, how proteins fold, how small molecules bind to targets. Their core technology, FEP+ (Free Energy Perturbation), computes binding free energies with unprecedented accuracy, allowing chemists to design optimized drug candidates in silico before synthesizing a single molecule.

Think of it this way: most AI drug companies are pattern matchers. Schrödinger is a physics simulator. Pattern matching breaks on novel targets where there's no historical data. Physics doesn't — because physics is physics. This is the moat.

Software Platform — The Recurring Engine

  • FEP+ (Free Energy Perturbation): Gold standard for binding affinity predictions. Thousands of GPU-hours per campaign. Accuracy within 1 kcal/mol — sufficient for medicinal chemistry decisions.
  • LiveDesign: Cloud-native enterprise informatics. Real-time collaboration for drug design teams. Now integrating Lilly's TuneLab AI models.
  • Maestro: Flagship molecular modeling GUI. New AI-powered conversational interface. Used daily at all top-20 pharma companies.
  • Predictive Toxicology: Beta solution covering ~50 kinases. Gates Foundation–funded ($19.5M grant).
  • Materials Science: Same physics engine applied to catalysts, polymers, semiconductors. Neptune ammonia catalyst achieved 2x yield vs competitors.

Drug Discovery — The Optionality

  • Two clinical-stage oncology programs (SGR-1505, SGR-3515) plus preclinical assets (SGR-5573, SGR-6016)
  • Strategic pivot (2025): No longer advancing programs independently through late-stage trials. Discovery-focused model — design molecules, license to pharma partners. ~$70M annual savings.
  • Collaborative pipeline: Novartis multi-target deal. Ajax Therapeutics (co-founded) in Phase 1.
  • Revenue model: Upfront payments + milestones + royalties. Software licensing provides the steady base.
"Every pharmaceutical company on earth is Schrödinger's customer. That's not a pipeline bet — it's an infrastructure position on the digitization of drug discovery itself."
Section II

What Makes Schrödinger Novel

Five Structural Advantages No Competitor Replicates

1. Physics + AI Hybrid. Pure AI companies (Recursion, Insilico, Isomorphic Labs) train models on data and predict outcomes. Schrödinger calculates outcomes from quantum mechanical first principles, then enhances with ML. When AlphaFold predicts a protein structure, FEP+ simulates how millions of candidates interact with it at atomic resolution. Physics provides the accuracy floor; AI accelerates the search. No other company has 30+ years of validated physics code layered with modern ML.

2. All Top-20 Pharma Are Customers. Not prospects. Customers. Novartis, Lilly, Pfizer, Roche, Merck, J&J, AstraZeneca, GSK, Sanofi, BMS — every single one. This creates an information flywheel: as more companies use the platform, Schrödinger gains insight into which approaches work, improving the platform, attracting more usage.

3. Dual Revenue Model = De-Risked Biotech. Traditional biotechs burn cash praying for clinical success. Schrödinger generates $160-170M in recurring software revenue while also running a drug pipeline. If every clinical program fails tomorrow, the software business alone justifies $1.3-2.0B at peer multiples. The pipeline is free optionality on top of a real business.

4. Materials Science Expands TAM Beyond Pharma. Neptune ammonia catalyst — 2x yield improvement, 47% capacity increase, 23% energy reduction — proved the physics engine works on industrial chemistry. Catalysts, polymers, battery materials, semiconductors — the TAM expansion from pharma ($70B+) to materials ($200B+) is completely unmodeled by analysts covering SDGR as "healthcare technology."

5. The Jensen Huang Signal. NVIDIA's CEO hosted Schrödinger's CEO at his home and told him to "think bigger." There was "shouting." This isn't casual — it's strategic alignment between the GPU maker and the software that consumes them. Every advance in NVIDIA compute makes Schrödinger's simulations faster. SDGR is a direct beneficiary of Phase 1 infrastructure investment.

Section III

Partnership Ecosystem

Novartis — The Landmark Deal

$150M upfront + $2.3B in milestone payments + royalties + expanded software license.

Multi-target research collaboration to discover and develop novel therapeutic candidates. Novartis gains industry-leading access to Schrödinger's platform. Schrödinger receives milestone cascade that could generate revenue for a decade. This single deal validates computational drug discovery at the highest level of pharma.

Eli Lilly — TuneLab Integration

Lilly's TuneLab AI platform integrated into LiveDesign. Participating biotechs access TuneLab through Schrödinger's interface — making LiveDesign the de facto OS for AI-assisted drug design. Prior Lilly relationship yielded $47.6M (Morphic acquisition).

NVIDIA — Compute Partnership

Jensen Huang's $1B AI pharma lab. FEP+ is a primary GPU workload in life sciences. BioNeMo integration. More powerful GPUs = more simulations = Phase 1 → Phase 3 feedback loop.

Ajax Therapeutics — Venture Creation

Co-founded by Schrödinger. Series C complete. AJ1-11095 in Phase 1. Demonstrates the venture-creation model: design molecules, spin out companies, retain equity upside.

Additional Partners

  • Manas AI: Strategic AI drug design agreement
  • Copernic: Neptune ammonia catalyst (materials)
  • Gates Foundation: $19.5M predictive toxicology grant
  • Takeda: Prior $111M (Nimbus acquisition)
  • All top-20 pharma: Annual software licenses
Section IV

Clinical Pipeline

SGR-1505 · MALT1 Inhibitor
Relapsed/Refractory B-Cell Malignancies · Waldenström Macroglobulinemia
PHASE 1 — DOSE ESCALATION COMPLETE

Lead asset. 22% ORR across 45 patients. Well-tolerated — no DLTs, no treatment-related deaths. FDA Fast Track + Orphan Drug Designation for Waldenström. Dose escalation complete; recommended P2 dose to be discussed with FDA. Seeking licensing partner for mid/late-stage development — becomes milestone + royalty asset.

SGR-3515 · Wee1/Myt1 Co-Inhibitor
Advanced Solid Tumors
PHASE 1 — DOSE ESCALATION ONGOING

Key upcoming catalyst. Initial clinical data expected H1 2026 — within our window. Differentiated mechanism in solid tumors. Preclinical data strong at EORTC-NCI-AACR. This readout is the most important binary event for SDGR in 2026.

SGR-5573 · Brain-Penetrant EGFR Inhibitor
Osimertinib-Resistant NSCLC + Brain Metastases
PRECLINICAL

Potent against resistant EGFR variants. Strong wild-type selectivity. Robust anti-tumor activity in brain metastases models (ESMO). Addresses massive unmet need — osimertinib resistance in ~40% of NSCLC patients.

SGR-6016 · Brain-Penetrant NLRP3 Inhibitor
Neuroinflammation
DEVELOPMENT CANDIDATE

Newly selected. NLRP3 inflammasome validated neuroinflammation target. Potential licensing candidate for pharma partners.

Discontinued: SGR-2921 · CDC7 Inhibitor

Discontinued Aug 2025 after emergent events linked to two patient deaths. Pipeline narrowed to two clinical assets but accelerated pivot to discovery-focused licensing model. The $70M in annual savings improves runway by ~2 years.

Section V

Financial Architecture

Revenue Structure

Annual Revenue (FY24)$207.5M
Q3 2025 Revenue$54.3M (+54% YoY)
Software Revenue (Q3)$40.9M (+28% YoY)
Drug Discovery Rev (Q3)$13.4M
Gross Margin58.2%
Software Gross Margin73–75%
FY25 Software Guide+8–13% growth
FY25 Discovery Guide$49–52M

Cash & Runway

Cash + ST Investments$352–391M
FCF (TTM)-$31.9M
Net Margin-68.5%
EPS (TTM)-$2.39
Cost Savings (Pivot)~$70M annual
Est. Post-Pivot Burn~$80–100M/yr
Est. Cash RunwayThrough 2027–28
Dilution RiskLow

The Valuation Anomaly

At $870M market cap with $352-391M cash, EV is ~$479-518M. Software revenue at ~$165M growing 28% with 73-75% gross margins would trade at 8-12x for any pure SaaS company — implying $1.3-2.0B for software alone. The market is assigning approximately negative $800M to $1.5B of value to the Novartis $2.3B milestone cascade, the entire clinical pipeline, materials science applications, 30 years of physics IP, and the Lilly/NVIDIA partnerships. That's the dislocation.

Section VI

Alpha Score v2 — Forward 6-Month

71
Composite Alpha Score
STRONG ALPHA BUY 1-1.5%
Moat Durability
72
Catalyst Runway
78
Supply/Demand Traj.
68
Edge Decay Rate
65
Regime Alignment
68

= (72×.30)+(78×.25)+(68×.20)+(65×.15)+(68×.10) = 71.0

Factor Reasoning

FactorScoreRationale
Moat Durability7230yr physics codebase. All top-20 pharma customers. FEP+ unmatched by pure-ML. But Isomorphic Labs (Google) and internal pharma AI teams investing heavily. Moat widening from physics side, testable from ML side.
Catalyst Runway785 catalysts in window: Q4 ER Feb 25, SGR-3515 Ph1 data H1, SGR-1505 licensing partner, Novartis milestones, Neptune pilot. Majority new-information events. Densest calendar in Phase 3 space.
Supply/Demand68Pharma AI adoption accelerating. Patent cliff driving large pharma to computational approaches. GPU cost decreases expand simulation capacity. But software guide lowered 2pts — demand strong, not explosive.
Edge Decay65At 52w low — generalists have abandoned. Hybrid model confuses both software and biotech analysts. Most institutions have sold. Edge intact because nobody is looking.
Regime Alignment68Biotech outperformance trend (Morgan Stanley). Rate cut tailwinds. Patent cliff driving pharma M&A + licensing. Some FDA uncertainty under Trump creates headwind.
Section VII

Risk-Reward Scenarios — 6 to 12 Months

Bull Case — $22–28 (+86–137%)

Probability: 25%

  • SGR-3515 data shows strong signal → pipeline re-rated
  • SGR-1505 licensing deal ($100M+ upfront)
  • Software growth re-accelerates to 20%+
  • Novartis milestone payment disclosed
  • Materials science partnership expands TAM narrative

Base Case — $16–20 (+35–69%)

Probability: 40%

  • SGR-3515 data mixed but encouraging
  • Software growth sustains 10-15%
  • Cost savings materialize, burn improves
  • Stock recovers from extreme oversold

Bear Case — $8–11 (-7–32%)

Probability: 25%

  • SGR-3515 disappoints → pipeline narrative collapses
  • Software growth decelerates below 5%
  • No licensing deal materializes
  • Pharma macro headwinds worsen

Crash Case — $5–7 (-41–58%)

Probability: 10%

  • Second clinical safety event
  • Major customer loss / Novartis restructuring
  • Broad biotech liquidity crisis
  • Cash floor ~$5/share provides hard bottom

Expected Value

E[V] = (0.25 × $25) + (0.40 × $18) + (0.25 × $9.50) + (0.10 × $6) = $16.43+39% expected upside

65% probability of positive outcome. Bull upside (+86-137%) significantly exceeds Bear downside (-7-32%). Cash floor ~$5/share bounds the crash scenario.

Section VIII

The Acceleration Thesis

"Software compounds exponentially while atoms compound linearly. Computational drug discovery follows software timelines, not pharmaceutical timelines. The market applies pharma discount rates to a software acceleration curve."

The Compute Curve: Every GPU NVIDIA ships feeds into Schrödinger's simulation capacity. FEP+ consumes thousands of GPU-hours per candidate. As compute cost falls ~40% per FLOP annually, molecular simulations per dollar double every 18 months. This doesn't make robots faster — it makes molecular discovery faster. Phase 1 infrastructure compounds directly into Phase 3 speed.

The Data Curve: Biological datasets for ML training double every 12-18 months. Schrödinger's top-20 pharma customers feed results back into model improvement. Each dataset improves the next model, which targets the next simulation, which generates better data. The flywheel accelerates.

The Integration Curve: Schrödinger's FEP+ engine + Lilly TuneLab + NVIDIA BioNeMo + Ginkgo/OpenAI autonomous labs + AlphaFold structures = an ecosystem where each platform amplifies the others. This interoperability creates the step function no linear extrapolation captures.

Implication: If AI compresses target identification from 4 years to 18 months and autonomous labs compress experiments from weeks to days, the market's discount rate on Phase 3 is systematically too high. Clinical events arrive sooner than modeled. Stocks re-rate when speed becomes undeniable — before drugs are approved.

Section IX

Execution Plan

Entry Zone$11.00 – $12.50 (current)
Target (6-mo)$18 – $22 (+52–86%)
Target (12-mo)$22 – $28 (+86–137%)
Stop Loss$9.50 (-20%)
R:R Ratio2.6:1 to 4.3:1
Position Size1 – 1.5% ($75 – $115K)
ConvictionMEDIUM Phase 3 Option Book

Tranche Strategy

Tranche 1 (Now): 0.5% position ahead of Feb 25 earnings. Downside bounded by cash floor. Upside from earnings beat could gap stock 15-25%.

Tranche 2 (Post-ER): If earnings confirm software growth + cash runway, add 0.5% to 1% total.

Tranche 3 (SGR-3515 Data): When H1 2026 clinical data releases — add 0.5% if positive (→ 1.5%). If negative, hold existing but tighten stop to $10.

Section X

Catalyst Calendar

DateEventImpactType
Feb 25 ★Q4/FY25 EarningsSoftware growth, cash update, FY26 guide, pipeline updateNew Info
H1 2026SGR-3515 Phase 1 DataBiggest binary catalyst. Wee1/Myt1 in solid tumors. Data = pipeline re-rate.New Info
2026SGR-1505 Licensing PartnerUpfront payment + milestone cascade. Validates discovery-to-license model.New Info
OngoingNovartis MilestonesRevenue catalyst with zero R&D cost to SchrödingerConfirm
2026-27Neptune Catalyst PilotValidates materials science TAM expansionNew Info
Q2 2026ASCO / Medical MeetingPotential venue for clinical data presentationsNew Info
Section XI

What Could Kill This Thesis

  • Clinical failure cascade: SGR-2921 already discontinued (2 deaths). If SGR-3515 also fails, pipeline narrative collapses. SDGR becomes pure software co at $870M — potentially still undervalued, but optionality premium vaporizes.
  • Software deceleration: Guide lowered from 10-15% to 8-13%. If growth drops below 5%, the "SaaS floor" thesis weakens.
  • Google competition: Isomorphic Labs is the existential threat. Unlimited compute, AlphaFold, deep pockets. If they achieve FEP+-competitive accuracy with superior speed, moat narrows.
  • Novartis underperformance: If collaboration doesn't advance to milestone-triggering stages, $2.3B looks aspirational.
  • The chart: 90% decline from ATH. Every MA above. Maximally negative momentum. Even if fundamentals are right, stock could grind sideways for months. This is a patience trade.
Section XII

The Bottom Line

"Schrödinger is the computational engine of Phase 3 Discovery — the only company that monetizes everyone else's AI drug discovery acceleration through software licensing while retaining its own pipeline optionality. At $870M with $370M in cash, you're buying the engine for the price of the fuel."

The software platform — used by every major pharma company — generates real revenue growing 28% with 73% gross margins. Novartis validates the approach at the highest institutional level. The clinical pipeline retains two active programs with near-term catalysts. Materials science expands TAM beyond pharma. The strategic pivot transforms the cash equation from burning to survive to investing to compound.

At $11.81, the market prices Schrödinger as if computational drug discovery has failed. The evidence says it's just beginning to work. The acceleration thesis — compute, data, and integration curves all steepening simultaneously — means Phase 3 validation events arrive sooner than consensus models. Position accordingly: 1-1.5% in tranches, stop at $9.50, let the catalysts work.