Madrigal Pharmaceuticals, Inc. (NASDAQ: MDGL) — A First-in-Disease Blockbuster Priced for a Market It Just Taught Everyone Else to Want

Independent equity research · 2026-07-10

⚡ Claude’s Take

This block is the author’s own subjective opinion and general information only — not investment advice. The analysis that follows it is deliberately position-free and carries no price target.

Verdict: HOLD — a genuinely great single-asset launch at a full-but-not-crazy price; accumulate on weakness, don’t chase near the all-time high. Not a short. Fair-value zone ~$460–560; I’d get actively constructive under ~$440 (≈7x FY26 sales / ~5x 2028E sales). Tag: “They built the market — now everyone’s moving in.”

Madrigal did something rare: it ran the first Phase 3 trial to beat MASH, won the first-ever FDA approval (March 2024), and turned Rezdiffra into a $1.1B-run-rate blockbuster in under two years — 42,250 patients, 92%+ gross margins, a market it is <10% penetrated into and that is growing ~50% partly because competitors are now spending to build it. On execution this is close to a best-case specialty launch, and the stock’s own price-to-sales sits at the cheapest level in its history (4.5th percentile) precisely because revenue outran the share price through 2026. That is the bull case in one sentence: a mega-blockbuster still in its infancy, patent-protected to ~2045, at ~10x forward sales near an all-time high.

What keeps me at HOLD rather than BUY is the other side of the same coin. This is a 100%-single-asset company whose ~$13B enterprise value already discounts the mega-blockbuster path — and that path is now contested by the deepest competitive set in biopharma: Novo’s Wegovy already has a MASH label (Aug 2025), Lilly’s tirzepatide/oral orforglipron loom, Akero and Roche/89bio are racing Madrigal to the F4c cirrhosis prize that management is counting on to “double” the market. Gross-to-net has structurally stepped to the high-30s, handing payers a lever that only tightens. And the tell I can’t ignore: insiders have sold ~$150M with literally zero open-market buying, the founders cashing out post-approval, the CEO holding just 24,195 shares. Nobody inside is underwriting $560 with new money. This is a quality-growth-at-a-price / momentum name whose fundamentals are proven but whose terminal value is genuinely uncertain — the empirical tape confirms it (RS 12-mo +95%, beta 0.84, Sharpe 1.7, alpha positive; an uptrend, not a falling knife). Conviction: medium. The single fact that flips me bullish: a positive MAESTRO-NASH-OUTCOMES F4c readout in 2027 while Rezdiffra holds F2/F3 share against Wegovy. The single fact that flips me bearish: hard evidence of GLP-1 share capture or an accelerating GTN slide — the moment “competition grows the market” becomes “competition takes the market.”

📈 Stock Price Action — Five-Year Event Map

Madrigal has made one of the great biotech re-ratings of the decade — but unusually, the round-trip has been almost all one direction. Over five years the stock traveled from ~$99 (mid-2021) to a low of $55.89 (Jan 2022), then re-rated ~10x to an all-time high of $602.83 (Dec 22, 2025), and sits at $560.01 today (Jul 9, 2026) — just -7.1% off its all-time high, near the top of a 52-week range of $289.88–$602.83. The entire re-rating rests on two binary events (a Phase 3 readout and an FDA approval) followed by a commercial ramp that has, so far, validated the bull case.

Cycle narrative. (1) For eighteen months MDGL was a pre-data option on a THR-β agonist in a disease that had buried every prior developer; the January 2022 low of $55.89 marked peak skepticism. (2) On December 19, 2022, MAESTRO-NASH became the first Phase 3 MASH trial to hit both biopsy endpoints (NASH resolution and fibrosis improvement) — the stock nearly quadrupled in a week ($63.80 → $289.43), permanently repricing the equity from “binary option” to “likely first-to-market drug.” (3) Through 2023 the stock gave back more than half that gain as the market discounted dilution and fretted over launch timing — a textbook “sell-the-readout” fade to $131 by October 2023. (4) The FDA’s March 14, 2024 accelerated approval of Rezdiffra — the first drug ever approved for MASH — re-anchored the stock to ~$270. (5) 2024 was a “prove-the-ramp” holding pattern as the Street waited for the launch curve. (6) Then the 2025 prints landed: revenue compounded from $137M (Q1’25) to $321M (Q4’25), management declared “blockbuster status,” and the addressable market itself grew ~50% — driving the stock to its $602.83 ATH in December 2025 even as Novo’s Wegovy won a MASH label (August 2025) without denting Madrigal’s adds. (7) 2026 has been digestion: a February dip to $432 on GLP-1-substitution fear and a BofA downgrade to Neutral, recovered to $560 as the Q1’26 print (+127% YoY, gross-to-net better than feared) reaffirmed the trajectory.

1. Executive Summary

Madrigal Pharmaceuticals is a commercial-stage biopharmaceutical company with a single product — Rezdiffra (resmetirom), the first and only FDA-approved therapy for MASH (metabolic dysfunction-associated steatohepatitis). Approved March 14, 2024 under accelerated approval for non-cirrhotic MASH with moderate-to-advanced fibrosis (F2–F3), Rezdiffra is an oral, once-daily, liver-directed thyroid hormone receptor-beta agonist. It has produced one of the best specialty-pharma launches of the last decade: net sales of $180.1M (FY24) → $958.4M (FY25, +432%) → $311.3M in Q1’26 (+127% YoY, ~$1.25B annualized run-rate; ~$1.13B trailing-twelve-months), with >42,250 active patients and >10,000 prescribers.

The economics behind the ramp are excellent: gross margin runs 92–96% (COGS is essentially a Roche royalty on a small molecule), and the operating loss has narrowed sharply from -$497.9M (FY24) to -$300.1M (FY25) even as revenue quintupled — proof of real operating leverage. Operating cash burn improved from -$456M (FY24) to -$190M (FY25). The company holds ~$813M in cash and securities against ~$340M of senior debt; management says profitability is “inevitable” beyond 2026.

The market opportunity is genuinely large and under-served. The US addressable pool (diagnosed F2/F3 under specialist care) grew from 315,000 (YE2023) to 460,000 (YE2025), yet the diagnosis rate is only ~10% and Rezdiffra penetration is <10% of the addressable base — a long runway. A separate ~245,000-patient F4c (compensated cirrhosis) opportunity, with no approved therapy, could “double” the market if the MAESTRO-NASH-OUTCOMES trial reads out positive in 2027.

The debate is not whether Rezdiffra works or sells — both are answered — but how much of the market it keeps, at what net price, and whether the ~$13B enterprise value already prices the mega-blockbuster outcome. The competitive set is the deepest in biopharma: Novo’s Wegovy already holds a MASH label (Aug 2025), Lilly (tirzepatide, oral orforglipron), Akero and Roche/89bio (racing Madrigal in cirrhosis), and others are converging on a profit pool Madrigal proved exists. Gross-to-net has stepped structurally to the high-30s. And insiders have sold ~$150M with zero open-market buying. This is a proven, high-quality, single-asset growth franchise whose terminal value is contested — expensive on near-term multiples, reasonable only if the mega-blockbuster-plus-F4c thesis holds.

The analysis below discusses valuation solely as embedded expectations and scenarios. It contains no recommendation and no price target.

2. Business Overview

Madrigal is a one-product company. One hundred percent of revenue comes from Rezdiffra; there is no partnered revenue, no second commercial product, and no diversification. The entire equity is a levered position on a single molecule and its label-expansion path.

The disease. MASH (formerly NASH) is the progressive, inflammatory form of metabolic-dysfunction-associated steatotic liver disease — fat accumulation in the liver driving inflammation and fibrosis that, untreated, progresses to cirrhosis, liver failure, hepatocellular carcinoma, and transplant. It is the #1 cause of liver transplant in US women and #2 in men (management, Q1’26 call). It is chronic, largely asymptomatic until late, and — until March 2024 — had no approved pharmacotherapy; standard of care was diet and exercise. That combination of high prevalence, progressive severity, and zero prior treatment is what makes the commercial opportunity so large.

The product. Rezdiffra (resmetirom) is an oral, once-daily, liver-directed thyroid hormone receptor-beta (THR-β) agonist. THR-β is expressed predominantly in the liver; selective agonism raises hepatic fat metabolism, lowers liver fat, improves lipids, and — critically — improves fibrosis, without the systemic thyroid effects of THR-α. Three properties define its commercial positioning: it is weight-loss-independent (works in lean MASH and in patients regardless of GLP-1 use), requires no titration, and is well-tolerated (mild, transient GI effects; manageable LDL/drug-interaction profile). It is dosed by body weight (two dose strengths). Approval rested on MAESTRO-NASH, the pivotal biopsy-confirmed Phase 3 histology trial that hit both accelerated-approval surrogate endpoints: MASH resolution without worsening fibrosis, and ≥1-stage fibrosis improvement without worsening MASH.

How it makes money. Rezdiffra is a chronic-therapy annuity priced at a launch WAC of ~$47,400/year (list has since risen to ~$49,400), within the ICER cost-effectiveness band ($39,600–$50,100). Net revenue per patient runs ~$28–30k/year after a blended gross-to-net discount that stepped to the high-30s% in 2026. Distribution flows through a limited network of specialty pharmacies (a concentration risk flagged in the 10-K), supported by a patient-access hub and copay assistance. Because MASH is lifelong and progressive, the model compounds on patient persistence — management reports strong adherence and “increasing depth of prescribing.” Revenue is recognized in the US today; an EU leg opened in late 2025 but is deliberately negligible near-term.

Segmentation. There is effectively one segment (Rezdiffra, US), with a nascent, deliberately throttled EU contribution and a >10-program pipeline (all early-stage, BD-sourced) that generates no revenue. Recurring vs. non-recurring: revenue is ~100% recurring chronic-therapy; the pipeline is pure optionality.

Verdict: A genuinely differentiated, high-margin, chronic-annuity product executing one of the better specialty launches in recent memory — but with total single-asset concentration. Every dollar of enterprise value depends on Rezdiffra’s trajectory and label expansion.

3. Industry Dynamics

A large, genuinely undertreated market. Management sizes the US “addressable” pool — diagnosed F2/F3 patients under the care of target specialists — at 460,000 (YE2025), up from 315,000 (YE2023), roughly +50% in two years. The striking figure is the funnel leakage: the diagnosis rate is “just over 10%,” implying a true F2/F3 population on the order of 4–4.6 million in the US, the vast majority undiagnosed. Rezdiffra penetration is <10% of the 460,000 addressable (42,250 ÷ 460,000 ≈ 9.2%). Layer on F4c well-compensated MASH cirrhosis: ~245,000 US patients under specialist care, no approved therapy — a pool management says could “double” the opportunity. On the demand axis this is structurally rich: a chronic, progressive, high-prevalence disease with ~92–96% gross margins, high persistence, and a diagnosis bottleneck that is both a near-term headwind and a multi-year runway.

The diagnosis bottleneck is the real gate, not demand. MASH is asymptomatic until late; identification requires FibroScan/VCTE, MRE, or biopsy, and most primary-care and endocrine patients are never staged. This caps near-term penetration but — perversely — protects the incumbent: building diagnostic and referral infrastructure is slow and costly, and Madrigal has a two-year head start “wiring the system” practice by practice. Rising point-of-care FibroScan/NIT adoption in specialist offices is a durable tailwind that converts the 1.9M diagnosed into stageable, treatable patients.

Greenwald “Competition Demystified” lens. On the demand side the industry is attractive (large, growing, sticky, high-margin). On the supply/barriers-to-entry side it is not structurally protected: the moat is regulatory/IP plus first-mover, and those barriers erode as approvals arrive. There is no economies-of-scale-plus-captivity dynamic locking the market to a single supplier — a superior molecule can enter and take share drug-by-drug. This is the crucial distinction: a good market is not the same as a defended franchise.

Marathon capital-cycle lens — the flashing warning. Madrigal proved MASH is a multi-billion-dollar profit pool, and capital is now flooding in exactly as the capital-cycle framework predicts. Within ~18 months of Rezdiffra’s approval: Novo Nordisk won a competing MASH approval (Aug 2025); Roche paid up to $3.5B for 89bio/pegozafermin (Sept 2025); and Lilly, Boehringer/Zealand, GSK, Akero, Viking, and Inventiva all pushed programs into Phase 2b/3. That is textbook supply-side deterioration of a newly proven profit pool — high returns attracting capital that will, over time, compete away some of the incumbent’s economics. The counter-nuance is that the market is expanding fast enough that new entrants have so far grown the pool rather than carved into Rezdiffra’s volumes — but capital-cycle logic says the pressure builds.

Verdict: structurally good demand, deteriorating supply-side. A large, chronic, undertreated market with excellent product economics — but a profit pool being contested faster than almost any recent therapeutic category. Good market, contested moat.

4. Competitive Position

Name the moat honestly: a genuine but time-limited first-mover franchise lead, not a structural fortress. Rezdiffra’s defensible assets are real: (1) first-in-disease approval with an ~18-month-plus head start; (2) the only oral, liver-directed, weight-loss-independent, well-tolerated agent — a mechanistically distinct niche the GLP-1s do not occupy; (3) established first-line commercial-payer access with no step edits, built over two years; (4) a real-world-evidence flywheel (tens of thousands of patients, 40+ congress abstracts, RWD on liver stiffness, liver fat, enzymes, LDL, Lp(a)); and (5) a 2027 (F4c) / 2028 (F2–F3) outcomes-data lead that no competitor can replicate near-term. In Greenwald’s taxonomy these are intangible / regulatory barriers with a clock on them, reinforced by modest chronic-therapy switching costs — not scale or network-effect barriers. A clearly superior competitor can still relegate or displace Rezdiffra; incumbency buys time, not permanence.

The competitive threat is unusually deep. Pressure-testing each:

• Novo Nordisk — Wegovy (semaglutide 2.4mg, injectable GLP-1). FDA-approved for non-cirrhotic MASH F2–F3 on Aug 15, 2025 (accelerated, ESSENCE Part 1: MASH resolution 63% vs 34% placebo; both resolution + fibrosis improvement 33% vs 16%). The most direct threat — a global commercial machine with the cardiometabolic/weight-loss halo patients and payers want. The pivotal question is whether it takes share or grows the market. Management’s data-point: three quarters into Wegovy’s MASH availability, Rezdiffra adds have not slowed (Q1’26 was the best NBRx month since launch), ~25% of Rezdiffra patients are concurrently on a GLP-1 and >50% previously exposed — i.e., real-world use is combination, not substitution. But there is no independent share data yet, and mechanistically GLP-1s attack the upstream driver (obesity), so many patients arrive on a GLP-1 first. Treat “market-growing, not share-taking” as a corroborated hypothesis, not proof.

• Eli Lilly — the biggest long-term threat. Tirzepatide (GIP/GLP-1) SYNERGY-NASH Phase 2 posted MASH resolution up to 73.3% (15mg) vs 13.2% — best-in-class efficacy. Add retatrutide (triple agonist, Phase 3) and orforglipron, an oral GLP-1 (approved as Foundayo for T2D, April 2026). An oral GLP-1 with a future MASH label would erase Rezdiffra’s convenience edge.

• Akero — efruxifermin (FGF21, weekly injectable): the F4/cirrhosis leader. SYMMETRY Phase 2b in compensated cirrhosis showed cirrhosis reversal 39% (50mg) vs 15% placebo at week 96 (NEJM, May 2025) — a striking result in F4, where Rezdiffra’s own F4c data are open-label/biomarker-based. Akero’s Phase 3 SYNCHRONY program runs in both F4 (Outcomes) and F2–F3 (Histology), directly contesting the F4c “doubling” opportunity Madrigal is counting on.

• Roche / 89bio — pegozafermin (FGF21). Roche’s up-to-$3.5B acquisition (Sept 2025) puts big-pharma weight behind Phase 3 ENLIGHTEN (F2–F3) and ENLIGHTEN-Cirrhosis (F4).

• Boehringer / Zealand — survodutide (glucagon/GLP-1 dual, injectable). FDA Breakthrough Therapy; Phase 3 LIVERAGE + LIVERAGE-Cirrhosis.

• Viking — VK2809 (THR-β, the same class as Rezdiffra). VOYAGE Phase 2b positive (resolution up to 75%, fibrosis improvement, LDL −20–25%) — a direct mechanistic copy, though Viking has de-prioritized it behind its VK2735 obesity asset.

• Inventiva — lanifibranor (oral pan-PPAR). NATiV3 Phase 3 fully enrolled (~1,009); topline expected H2 2026 — a near-term binary that could add another oral option.

The likely endgame is combination therapy — and this is Madrigal’s strategic hedge. KOL consensus is that MASH will ultimately be treated with combinations (a liver-directed agent plus a systemic/weight agent). Management is explicitly building for exactly that: an in-licensed oral GLP-1 (MGL-2086 / CSPC SYH2086), a D2/THR combination, and the ARO-PNPLA3 siRNA from Arrowhead targeting a genetic driver present in ~30% of F2–F3 patients (I148M homozygotes, Hispanic-enriched, 2x liver-event risk; Phase 1 showed −46% liver fat) — all positioned around Rezdiffra as the “foundational backbone.” If that thesis holds, Rezdiffra survives even as GLP-1s dominate obesity, anchoring combos. If a single GLP-1 or triple agonist delivers enough fibrosis benefit alone, Rezdiffra risks relegation to a shrinking mono-therapy niche. Notably, management observes that big pharma has exited MASH after failures (Pfizer, J&J, BMS), which both reduces some near-term entry risk and validates the disease’s difficulty.

Intellectual property. Six Orange Book patents. The composition-of-matter patent (US 7,452,882) expires September 12, 2026 — a near-term lapse — but method-of-use, combination, formulation, and a new weight-threshold-dosing patent (issued Aug 5, 2025) extend protection to 2037–2045 (management’s base-case exclusivity is 2045). Franchise protection therefore likely holds well into the 2030s on the surviving claims, though the composition-of-matter lapse invites 505(b)(2)/generic probing that must be monitored.

Verdict — be direct. This is a genuine, differentiated, but time-limited first-mover franchise lead — not a structural moat. For the next ~2–3 years Rezdiffra owns a defensible position: the only oral liver-directed agent, entrenched first-line access, a real-world-evidence flywheel, and an outcomes-data advantage nobody can replicate soon. But it is a single-asset company in a market that just attracted every large pharma, facing GLP-1 machines with vastly bigger reach and a cardiometabolic halo. Durability hinges on three things breaking Madrigal’s way: (1) the F4c outcomes readout (2027) landing positive and roughly doubling the market before Akero/Roche get there; (2) outcomes data (2027/2028) cementing standard-of-care status; and (3) the combination/backbone thesis proving correct. Miss on those and the moat is thin. Best characterized as a multi-year lead, not a permanent franchise — a first-mover that must keep running to avoid being leapfrogged.

5. Growth History and Forward Opportunities

The ramp, decomposed. Rezdiffra’s revenue arc is one of the best specialty launches of the past decade, and it is entirely a volume story on a roughly fixed price. Quarterly net sales ($M): Q2’24 14.6 → Q3’24 62.2 → Q4’24 103.3 → Q1’25 137.3 → Q2’25 212.8 → Q3’25 287.3 → Q4’25 321.1 → Q1’26 311.3 (FY24 $180.1M; FY25 $958.4M; LTM ~$1.13B). The engine is the patients-on-therapy ladder (end-of-quarter, net of discontinuations):

Patients grew 2.5x year-over-year into Q1’26. Prescriber breadth is now largely built out: >10,000 HCPs have written Rezdiffra (milestone hit Q3’25), covering ~80% of the top-6,000 targets. Management has explicitly pivoted from breadth to depth — deeper penetration within an already-wired base. Net revenue per patient runs ~$28–30k/year, implying a blended gross-to-net of ~40% off the ~$49.4k list, consistent with the guided high-30s% GTN for 2026.

The Q1’26 sequential dip is GTN reset plus seasonality, not demand. Revenue fell ~3% sequentially ($321.1M → $311.3M) despite ~6,000 net patient adds. Two mechanical factors explain it fully: (1) a GTN step-up — after deliberately not contracting for the first six quarters (to preserve net price), Madrigal signed commercial first-line contracts effective 1/1/26, moving overall GTN into the high-30s but securing first-line access with no step edits (a genuine win); and (2) the “Q1 effect” — benefit-plan resets, deductible restarts, and re-verifications that hit ~100% of the book each January. Management cited the best NBRx week and month ever in April 2026 as evidence the demand curve never bent. The price move down is a fact; the “demand is fine” read is corroborated by rising patient counts and the no-step-edit access win. The bear counter: high-30s GTN is now structural and gives payers a lever that only tightens as competition matures.

Forward drivers — real TAM expansion plus large but unproven optionality.

• (a) F2/F3 core penetration. The strongest fact in the thesis: an addressable market that grew ~50% in two years to 460,000, with only ~10% diagnosis rate and <10% Rezdiffra penetration — durable runway in the core alone, expanding partly because competitors invest in awareness.
• (b) F4c compensated cirrhosis — the “doubler,” but a 2027 binary. ~245,000 US F4c patients, no approved therapy, sicker and higher-urgency. The event-driven MAESTRO-NASH-OUTCOMES Phase 3 reads out in 2027; if positive it supports F4c and full approval across F2–F4c. A 2-year open-label bridge (122 F4c patients from MAESTRO-NAFLD-1) is encouraging — mean −6.7 kPa liver stiffness, 35% regressed F4→F3, 65% of baseline-CSPH patients improved risk category — but it is small-N open-label supportive data, not the controlled outcomes trial, and Akero is running a competing F4 outcomes program.
• © EU/international — approved but deliberately de-emphasized. EC conditional marketing authorization granted Aug 19, 2025 (first/only approved MASH drug in EU), self-commercialized, Germany launch Q4’25, EU F2/F3 pop ~370,000 — but management guides ex-US as “negligible” in 2025 and “not much” in 2026 given MFN/reference-pricing. A multi-year, lower-margin call option, not a near-term needle-mover.
• (d) Diagnosis-rate expansion. Rising point-of-care FibroScan/NIT adoption feeds diagnosis and staging — a durable tailwind.
• (e) Pipeline (>10 programs, all BD-sourced, all early). CSPC oral GLP-1 (MGL-2086, $120M upfront, clinic 1H’26); Arrowhead ARO-PNPLA3 siRNA; Pfizer ervogastat (DGAT-2) + two early assets; and the Ribo/Ribocure siRNA platform ($60M upfront, up to $4.4B milestones, first candidate-nomination milestone hit July 2, 2026). All pre-clinical-to-early-clinical.

Verdict — high-quality but concentrated growth. The core F2/F3 US franchise is high-quality growth: volume-led, ~92–95% gross margin, first-line access without step edits, a genuinely expanding and under-penetrated TAM, and a patent wall to ~2045. That alone can compound for years. But this is a single-asset company (~100% of revenue is Rezdiffra US), and every “doubler” — F4c, EU, the combo/siRNA pipeline — is an option, not earnings: F4c is a binary 2027 readout against a credible competitor, EU is throttled by pricing, and the pipeline is early. Fast, durable-in-the-core, but capped-until-proven — the gap between a $1.25B run-rate drug and the “$2.6–2.7B by 2028 / mega-blockbuster” narrative rests on catalysts not yet in hand.

6. Financial Quality

Verdict up front: the economics are among the best a single-product launch can show — 92%+ gross margins, revenue compounding >400% annually, and a burn collapsing toward breakeven — but every dollar depends on one drug, and GAAP still shows a loss because Madrigal is deliberately outspending the ramp to entrench it.

Revenue. Rezdiffra reached a $1.13B trailing-twelve-month run-rate by Q1’26 — one of the fastest specialty launches on record. Annual: FY2024 $180.1M → FY2025 $958.4M (+432%) → Q1’26 $311.3M (+127% YoY, ~$1.25B annualized). The lone sequential dip (Q4’25 → Q1’26) is the “Q1 effect” plus a GTN step-up, not demand . One hundred percent of revenue is one product; there is no diversification.

Margins and unit economics. Gross margin runs 92–96%; COGS is almost entirely a royalty owed to Roche (Madrigal licensed resmetirom from Roche in 2008), i.e., ~8% of sales, not manufacturing cost — so incremental gross margin is effectively fixed and very high. Economics unquestionably improve with scale: operating loss narrowed from -$497.9M (FY24) to -$300.1M (FY25) even as revenue 5x’d, and operating margin went from -276% to -31%. The drag is deliberate spend, not poor unit economics: FY25 SG&A was $813.8M (commercial build-out, endocrinology field-force expansion, DTC) and R&D $388.5M (pipeline, outcomes trials, and ~$170M of one-time BD upfronts). Q1’26 continued the pattern — SG&A $268.5M, R&D $108.7M (inflated by a $54.3M one-time BD upfront) against $311M revenue, for a -$92.7M operating loss.

Cash generation. Operating cash burn improved dramatically: OCF -$455.6M (FY24) → -$189.6M (FY25); FCF -$193.0M (FY25). Capex is trivial (~$3.5M) — an asset-light small-molecule model. Management states profitability is “inevitable” beyond 2026 but explicitly not in 2026 (Q2 carries a $25M PNPLA3 upfront). Working capital swings with the AR/inventory build of a scaling launch (receivables $187M, inventory $112M by Q1’26 as they pre-buy API).

Balance sheet. Solid but no longer net-cash. Q1’26: cash + marketable securities $812.8M ($817.9M incl. restricted), against ~$340M of senior term debt + leases, for net debt of ~+$113M (they were net-cash as recently as Q2’25). GAAP equity is $543.5M with a $2.18B accumulated deficit — the cumulative cost of a 16-year pre-approval history. Tangible book is ~$26.5/share (FY25); the reported P/B (~24–30x depending on share count) is not analytically meaningful for a pharma whose value is the drug franchise. Liquidity is ample for the current burn (current ratio ~3.5x; multi-year runway even before breakeven), but the fortress net-cash days are over — they now lean on debt and equity to fund both launch and pipeline.

Quality-of-earnings flags: (i) the GAAP loss overstates the true cash economics somewhat — $98.1M FY25 SBC and ~$170M one-time BD upfronts are real drags, but the underlying gross-profit engine is highly profitable; normalized “core” R&D is closer to ~$220M; (ii) share count has risen steadily (16.5M FY20 → ~23.0M Q1’26) from serial equity raises — a genuine dilution cost; (iii) revenue quality is high (chronic persistence, first-line access, net of a high-30s% GTN) but concentrated in one SKU with a Roche royalty on top. Net: high-quality gross economics, improving cash trajectory, one-product concentration, and a balance sheet that has shifted from fortress to merely adequate.

7. Capital Allocation

The dilution machine — well-timed, not reckless. Madrigal is a textbook “dilute-to-fund-the-asset” biotech: share count grew from 15.5M (FY20) to ~22.9M (FY25), roughly +48%, funded by a cadence of large equity raises to carry resmetirom through Phase 3 and the launch:

The critical judgment: dilution was timed near or after the value inflection, not at the bottom. The largest raise (~$688M) priced at $260 in the days after the March 14, 2024 approval de-risked the asset; the stock later traded to ~$560. Selling equity when the option value was already recognized minimized dilution per dollar raised — competent sequencing. The 2025 pivot to non-dilutive senior debt once real revenue existed is the right order of operations.

The debt — clarifying the label. There is no convertible bond. On July 17, 2025 Madrigal signed a Financing Agreement with funds managed by Blue Owl (LSI Financing agent): up to $500M senior secured — a $350M Initial Term Loan (funded that day) plus $150M delayed-draw and a $250M uncommitted incremental. Terms: 3-month term SOFR (1.0% floor) + 4.75% (~9.5–10% all-in), bullet maturity July 17, 2030, no amortization, carrying value ~$339.9M net of discount. Proceeds repaid the $250M Hercules facility ($2.8M extinguishment loss). The only genuinely convertible instruments are Series A/B non-voting convertible preferred held by Baker Bros behind a 9.9%/19.99% ownership blocker — a modest dilution overhang.

BD deployment — where the capital-allocation risk actually sits. With the balance sheet flush, management pivoted hard into licensing (no whole-company M&A, no buybacks, no dividend): CSPC (oral GLP-1, $120M upfront, up to $2.0B); Pfizer (ervogastat DGAT-2 + two early assets, $50M upfront); Ribocure/Suzhou Ribo (six siRNA programs, $60M upfront, up to $4.4B). That is ~$230M of upfronts (~$170M expensed into FY25 R&D) deployed into pre-clinical/Phase-2 combination bets while the company still burns cash. This is the aggressive, unproven part of the record — spending the Rezdiffra windfall to build a MASH-combination franchise before the core asset has demonstrated durable self-funding. It is a defensible strategic hedge against future GLP-1 competition, but the quality of the spend is entirely unproven.

Spend intensity. SG&A exploded 7.5x in two years — $108M (FY23) → $435M (FY24) → $813.8M (FY25, ~85% of revenue) — larger than R&D. The launch machine is enormous; the bull case requires it to show operating leverage as revenue scales, and the improving OCF trajectory (-$324M → -$456M → -$190M) suggests it is beginning to.

Comp and incentives (DEF 14A, Apr 2026). CEO Bill Sibold (joined 9/8/2023, ex-Sanofi/Dupixent) earned $19.2M in 2025, after a $32.3M sign-on year (2023). LTI is an equal mix of options/RSUs/PSUs tied to relative TSR vs the Nasdaq Biotech Index (2025–2027) — shareholder-friendly design. But the 2025 annual cash bonus paid at 194.7% of target against a -$288M GAAP net loss, on revenue/access/patient goals (not profitability), with specific targets redacted. Alignment is adequate on equity design but weak on the cash side and on skin-in-the-game: Sibold owns just 24,195 shares.

Verdict — adequate, tilting to a watch item. Capital raising was intelligently timed (dilution near/after the value inflection; sensible pivot to debt). Capital deployment is now the risk: ~$230M of upfronts into unproven combination pipeline plus an $814M SG&A base, funded by a company still losing ~$190M of operating cash. No buybacks (correct for a burn-stage name). The verdict on allocation quality cannot yet be positive — it hinges on Rezdiffra reaching self-funding before the BD bets need more capital.

8. Changes and Headwinds — Last Two Years

Leadership and strategy. Bill Sibold became CEO in September 2023 and built a full US commercial organization from scratch ahead of approval (CFO Mardi Dier, CMO Dave Soergel). The strategic pivot is explicit: from single-product launch execution to a “maximize Rezdiffra + build the pipeline” two-track story, funded by aggressive BD (CSPC, Pfizer, Arrowhead, Ribo) and non-dilutive debt (Blue Owl $500M facility, July 2025).

Regulatory milestones. FDA accelerated approval March 14, 2024 (first-ever MASH drug); EC conditional approval August 2025; the AASLD practice guidance was updated to incorporate resmetirom, and EU guidelines list it first-line; a new dosing patent pushed exclusivity to ~2045. Guideline inclusion plus first-line labeling is a meaningful, under-appreciated moat reinforcement.

Competition — the defining change. The pivotal event is Novo’s Wegovy MASH approval on August 15, 2025 (ESSENCE: resolution 63% vs 34%; fibrosis improvement 37% vs 22%), positioned as a cheaper alternative to Rezdiffra’s list price. Madrigal frames semaglutide as market-growing (screening/awareness) rather than share-taking, citing Rezdiffra’s consistency across subtypes including T2D (~60% of patients), no titration/better adherence, and the ~25% concurrent GLP-1 use / >50% prior exposure that make the two complementary in practice. Three quarters in, management insists it is “not to Rezdiffra’s detriment.” This is the central bull/bear fault line and remains genuinely unresolved. Separately, Akero’s efruxifermin posted Phase 2b Week-96 F4 cirrhosis data (39% vs 15%, NEJM May 2025) and is running Phase 3 SYNCHRONY including an F4 outcomes study — a direct future threat in both F2/F3 and Madrigal’s prized F4c expansion. Lilly and Boehringer add to a crowding field.

Street reaction. Madrigal gives no formal revenue guidance. On June 11, 2026, BofA maintained Neutral and trimmed its PT to ~$542, cutting its 2028 Rezdiffra estimate from $2.7B to $2.6B, citing “rapidly evolving competitive landscape [that] raises the risk of price and market-share erosion.” The stock, which peaked at $602.83 (Dec’25), consolidated to ~$560 through H1’26 as the debate shifted from “can it launch” (answered) to “can momentum survive GLP-1 competition plus the GTN step-up.”

Headwinds (interpretation). GTN erosion (high-30s now structural); GLP-1 substitution/step-edit risk (benign today, but Novo’s price and scale are the overhang, with 2027 payer renewals the key test); the event-driven F4c OUTCOMES binary (2027) against a competing Akero readout; single-asset, single-geography concentration; EU/MFN pricing caps; and China-sourced BD (CSPC, Ribo) geopolitical/quality optics.

Verdict: the last two years strengthened the thesis operationally (blockbuster launch, first-line access, ~2045 patent, EU approval, guideline inclusion, a real pipeline) while simultaneously raising terminal-value uncertainty (Novo/Akero competition, structural GTN step-down, single-asset risk). Net-positive on execution, but the investment question has shifted from “will Rezdiffra work” to “how much of the market does it keep, and at what net price.”

9. Risk Analysis

The dominant, correlated risks are competitive (GLP-1 share + Akero cirrhosis) and valuation (a full multiple that prices success), amplified by single-asset concentration. A catastrophic total loss is very unlikely — this is an approved, ~$1.25B-run-rate, well-funded franchise — but a 40–50% de-rating on a competitive/GTN disappointment is entirely plausible given where the multiple sits.

10. Valuation Discussion (Embedded Expectations)

Where the multiple sits. At $560, market cap is ~$13.0B and enterprise value ~$12.4–13.2B. On sales that is ~13x trailing (FY25 $958M), ~10x the Q1’26 annualized run-rate (~$1.25B), and ~8.4x FY26 Street consensus (~$1.48B). There is no meaningful P/E (GAAP loss), and P/B (~24–30x) is analytically empty for a franchise-value pharma. The single most informative valuation datum is the own-history price-to-sales percentile: 4.5th — MDGL is at the cheapest price-to-sales in its history even near an all-time-high stock price, because revenue outran the equity through 2026. That is the crux of the valuation tension: expensive in absolute multiples, but the multiple has compressed as the business scaled.

Embedded-expectations framing. A ~$12.5B EV on a ~$1.25B run-rate implies the market is underwriting a clear path to a multi-billion-dollar franchise. Back-of-envelope: if Rezdiffra reaches the Street’s ~$2.6B (2028) and, at maturity, a ~92% gross margin business settles SG&A toward ~30–35% of sales and R&D ~15%, operating margins of ~40–45% are achievable → ~$1.0–1.3B operating income, perhaps ~$0.9–1.1B net, on ~24M+ shares → mature EPS on the order of $35–45. At $560 that is ~13–16x mature (2028+) earnings — reasonable if the drug gets there and holds share. The embedded assumptions the buyer is accepting: (1) the ramp continues to a mega-blockbuster (~$2.5–3B+); (2) GLP-1 competition grows rather than carves the market; (3) GTN stabilizes in the high-30s rather than sliding; and (4) F4c and/or the pipeline add incremental value. The market is arguably pricing (1)–(2) correctly given the evidence so far and giving only partial credit for (3)–(4) optionality.

Scenario analysis (illustrative, not a target).

• Bear (~$300–400): GLP-1s capture meaningful share, Rezdiffra plateaus ~$1.5–1.8B, GTN slides, F4c disappoints/delays; multiple compresses toward ~6–7x on a slower-growth single asset.
• Base (~$500–650): mega-blockbuster path intact, ~$2.5–3B by 2028–29, F4c positive, share held; the current ~8–10x forward-sales multiple is roughly fair and re-rates modestly on de-risking.
• Bull (~$800+): F4c roughly doubles the TAM, combos work, franchise scales toward $4B+ with durable pricing; a genuine standard-of-care compounder.

Comp context. Against commercial-stage specialty biopharma with a single dominant, patent-protected franchise (e.g., the annuity-drug comps in the sector — Jazz’s oxybate, Neurocrine’s INGREZZA, Axsome’s Auvelity), MDGL trades at a higher forward-sales multiple, justified by faster growth and a larger untapped TAM but exposed to more competitive intensity. It is priced as a grower, not a value name — the opposite of the deep-value single-asset pharma the sector often throws up.

Verdict: not cheap on any near-term multiple, but the own-history compression and the low market penetration mean the valuation is defensible for the mega-blockbuster path and stretched for anything less. The market is underwriting success; the risk is in what it is not fully pricing — competitive share loss and GTN erosion — not in an obviously silly multiple. No price target; no recommendation.

11. Variant Perception

Consensus. The Street views MDGL as a high-quality, first-in-class MASH franchise executing a best-in-class launch, with a large under-penetrated TAM and real F4c/pipeline optionality — tempered by an intensifying competitive set (BofA at Neutral captures the marginal-buyer hesitation; others remain constructive on the ramp). Consensus roughly = “great drug, great launch, fair-to-full price, watch the GLP-1s.”

Strongest bull case. A genuinely rare setup: first-in-disease approval, a foundational, differentiated, patent-protected (~2045) asset, a market <10% penetrated and growing ~50%, a blockbuster already at ~$1.25B run-rate with 92%+ gross margins and a burn collapsing toward breakeven — and, crucially, the own-history cheapest price-to-sales in company history. Add F4c (“doubles the market,” 2027), EU, and a >10-program combination pipeline as free options. If competition keeps growing the pool (as it has so far) and F4c lands, this is a multi-year standard-of-care compounder that grows into and past its valuation.

Strongest bear case. A 100%-single-asset company at ~$13B EV / ~10x forward sales near an all-time high, in a profit pool it just proved and that every large pharma is now attacking — Novo already MASH-approved, Lilly’s oral GLP-1 coming, Akero/Roche racing to cirrhosis. GTN has structurally stepped down and payers have a competitive lever that only tightens. The mega-blockbuster is already in the price, leaving asymmetric downside if share erodes or F4c slips. And insiders are voting with their wallets — ~$150M sold, zero bought, CEO holding 24,195 shares.

The 3–5 assumptions that matter most: (1) GLP-1 competition grows the market rather than taking Rezdiffra’s share; (2) gross-to-net stabilizes in the high-30s rather than sliding through 2027 renewals; (3) the F4c OUTCOMES trial reads out positive in 2027 and roughly doubles the TAM before Akero/Roche arrive; (4) outcomes data (2027/2028) cement standard-of-care status ahead of the composition-of-matter lapse and generic probing; (5) the market’s ~$2.5–3B peak assumption is right, not optimistic.

What would falsify each side. Bull falsifier: independent script data showing Wegovy/tirzepatide capturing incremental F2/F3 patients at Rezdiffra’s expense, or a GTN print that slides below the mid-30s — i.e., “market-growing” becomes “share-taking.” Bear falsifier: a positive 2027 F4c OUTCOMES readout with continued 2.5x-ish patient growth and stable GTN — the mega-blockbuster-plus-doubler thesis confirmed, which would re-rate the stock through its current range.

Factor-positioning read (empirical overlay). FactorsToday shows MDGL as a biotech-and-market-beta name (Industry: Biotech ~1.16, Market ~0.75) with strong, positive risk-adjusted momentum: RS 12-month +94.8%, 1-year return +81% (Sharpe 1.74, Sortino 3.1), 3-month +30%, with 6-month roughly flat (the H1’26 consolidation), beta 0.84 and positive alpha (~+0.19). This is the signature of an uptrend digesting gains near highs — not a falling knife and not an abandoned value name. Factor-similar peers (Viking VKTX, Roivant, Immunovant, Axsome) are volatile single-catalyst biotechs, underscoring that the tape treats MDGL as a high-beta, event-driven growth name whose direction is set by competitive/clinical newsflow. The positioning corroborates the “momentum-name-with-contested-terminal-value” framing: consensus is long the ramp; where it may be offsides is in under-weighting how quickly the capital cycle (Marathon) can compete away a newly proven profit pool.

12. Fact vs. Interpretation

13. Open Questions

1. Is Wegovy taking share or growing the market? No independent script/share data yet isolates Rezdiffra’s F2/F3 share trend post-Aug 2025. This is the single most important unknown.
2. Where does gross-to-net settle by 2027? High-30s is guided for 2026; the 2027 payer-renewal cycle under full GLP-1 competition is untested.
3. Will MAESTRO-NASH-OUTCOMES (F4c) read out positive and on time (2027)? Event-driven timing and outcome are both uncertain; Akero races the same prize.
4. Does the combination/backbone thesis hold — do the BD-sourced assets (oral GLP-1, ARO-PNPLA3, Ribo siRNA) generate data that make Rezdiffra a durable anchor, or is the ~$230M of upfronts value-destructive?
5. What is normalized peak revenue and mature margin structure? The gap between ~$1.8B (bear plateau) and ~$3B+ (bull) drives most of the valuation dispersion.
6. Does the composition-of-matter lapse (Sept 2026) invite meaningful 505(b)(2)/generic challenge despite surviving method/formulation/dosing claims to 2045?

14. What Must Be True

Bull case — what must be true:

• Rezdiffra keeps its F2/F3 share as GLP-1s scale (competition grows, not carves, the pool), and net revenue/patient holds ~$28–30k as GTN stabilizes.
• The addressable market keeps expanding (diagnosis rate climbs from ~10%) and penetration rises from <10% toward mid-teens+ over the next 2–3 years.
• MAESTRO-NASH-OUTCOMES (F4c) reads out positive in 2027, roughly doubling the TAM and enabling full F2–F4c approval, ahead of Akero/Roche.
• The launch machine shows operating leverage (SG&A falls as a % of the growing revenue base) and the company reaches sustained self-funding before the BD bets require more capital.
• Falsification test: independent data showing Wegovy/tirzepatide capturing F2/F3 patients at Rezdiffra’s expense, OR a GTN print sliding below the mid-30s, OR an F4c OUTCOMES failure/material delay — any one breaks the bull.

Bear case — what must be true:

• GLP-1s (injectable now, oral orforglipron later) substitute for, rather than complement, Rezdiffra in F2/F3, and/or Akero/Roche win the cirrhosis race — eroding both share and the F4c “doubler.”
• GTN slides through 2027 renewals as payers use competition as leverage, compressing net revenue/patient.
• The ~$13B EV de-rates as the mega-blockbuster assumption is trimmed, with single-asset concentration amplifying the move.
• Falsification test: a positive 2027 F4c OUTCOMES readout accompanied by continued ~2.5x-type patient growth and a stable high-30s GTN — confirmation of the mega-blockbuster-plus-doubler thesis would break the bear and likely re-rate the stock through its current range.

15. Source Appendix

See the Source Appendix below for the full source list.

APPENDIX A — Standard Diligence Questionnaire

Madrigal Pharmaceuticals, Inc. (NASDAQ: MDGL) — as of 2026-07-10

Supplemental to the memo. Grounded in the research log; Fact/Interpretation/Assumption labels applied where material.

General

What thoughtful questions have other investors asked about this company? The core debate everywhere is substitution vs. complementarity with GLP-1s: does Novo’s Wegovy (MASH-approved Aug 2025) and Lilly’s coming oral orforglipron grow the MASH market or take Rezdiffra’s F2/F3 share? Secondary questions: where does gross-to-net settle after the 2026 step to the high-30s; will the F4c MAESTRO-NASH-OUTCOMES trial read out positive/on-time in 2027 (the “doubler”); is ~$13B EV / ~10x forward sales already pricing the mega-blockbuster; and is the ~$230M of BD upfronts into an early combination pipeline good capital allocation or windfall-spending. On the Q1’26 call, sell-side pressed on 2Q trends, the Q1 patient-add slowdown (GTN/seasonal), prescriber breadth (GAstro vs hepatology vs endocrinology), GLP-1 combination use, and F4c event-tracking.

Cyclicality & Earnings Nature

Cyclical high or low? Neither in the macro sense — this is a launch-curve business, structurally early (penetration <10%). Earnings (a GAAP loss) are depressed by deliberate launch/pipeline investment, not by a down-cycle. Driven by external environment or internal actions? Overwhelmingly internal (commercial execution, label expansion) plus a demand tailwind from competitor-funded disease awareness. Revenue stability? High and recurring — chronic lifelong therapy with strong persistence; the only volatility is gross-to-net resets and quarterly seasonality (the “Q1 effect”). Outlook for the product? Volume-led growth with a long runway; the debate is terminal share/price, not near-term direction. Market size? Large and growing: US addressable F2/F3 460k (from 315k in two years), ~10% diagnosed of a ~4–4.6M pool; F4c ~245k more; EU ~370k; global, expanding double-digit.

Business Quality & Competitive Moat

Industry more or less competitive? Rapidly more competitive — Madrigal proved the profit pool and capital is flooding in (Novo, Lilly, Akero, Roche/89bio, Boehringer, Viking, Inventiva). Profitability (ROIC/ROE)? Not yet meaningful — GAAP losses and a $2.18B accumulated deficit; ROIC/ROE are negative and not analytically useful until profitability (management: “inevitable” beyond 2026). Underlying gross economics are excellent (92–96% GM). Industry profitability / barriers? High gross margins, real regulatory/IP barriers, but no scale/network moat — barriers erode as approvals arrive. Easily understood? Yes — one drug, one disease, a volume ramp. Undermined by low-cost foreign labor? No (patent-protected branded pharma), though generic/505(b)(2) risk emerges after the composition-of-matter lapse (Sept 2026); surviving claims run to ~2045. Do brands matter? Yes — first-in-disease status, guideline inclusion (AASLD), and real-world-evidence build a genuine standard-of-care brand. Nature of competition? Efficacy, route (oral vs injectable), tolerability, payer access, and — increasingly — combination potential. Switching costs? Modest chronic-therapy stickiness; a clearly superior competitor can still displace.

Financial Condition & Balance Sheet

Assets not on the balance sheet? The Rezdiffra franchise itself (internally generated; carried at near-zero) is the entire economic value — tangible book (~$26.5/sh) vastly understates it. Off-balance-sheet liabilities? The Roche royalty (in COGS) and milestone obligations under BD deals (up to $2.0B CSPC, $4.4B Ribo, etc.) — contingent, not on the balance sheet. Accounting conservatism? Reasonable; ~$170M of BD upfronts were expensed (not capitalized) into FY25 R&D — conservative and it depresses the reported loss. CapEx-hungry? No — asset-light small molecule (~$3.5M capex); the “capex” equivalent is SG&A and BD spend.

Capital Allocation & Management

FCF generation and use? Still FCF-negative (-$193M FY25, improving from -$462M FY24); no FCF to allocate yet. Philosophy: reinvest everything into the launch and a combination pipeline. Recent acquisitions? No whole-company M&A; three licensing deals (CSPC $120M, Pfizer $50M, Ribo $60M upfronts). Buying back shares? No (correct for a burn-stage name). Issuing shares to insiders? SBC $98M FY25; share count +48% over five years from raises — genuine dilution, but raises were well-timed (largest at $260 post-approval). Compensation policy? CEO Sibold $19.2M (2025); PSUs tied to relative TSR vs Nasdaq Biotech (shareholder-friendly), but cash bonus paid 194.7% of target on non-profitability goals. Motivations of management? Growth/market-leadership focused; alignment adequate on equity design but weak on skin-in-the-game — Sibold owns only 24,195 shares, and insiders sold ~$150M with zero open-market buying.

Valuation & Market Data

ADR/MLP/K-1? No — US C-corp common stock; 1099, not K-1. Dividend policy? None (nor expected — reinvesting). Profitability? Not yet on a GAAP basis; high gross margins with a narrowing operating loss. Net income vs. cash from operations? Both negative; OCF (-$190M FY25) is less negative than net loss (-$288M) because of ~$98M SBC and working-capital timing — cash burn is running ahead of accounting profitability improvement.

Risks & Downside

What would cause the stock to decline? Evidence of GLP-1 share capture; a GTN slide; an F4c OUTCOMES failure/delay (2027); a de-rating of the full forward-sales multiple; disappointing pipeline data. Catastrophic loss risk? Low — an approved, ~$1.25B-run-rate, well-capitalized franchise; not a binary pre-revenue biotech. Total loss risk? Very low near-term; the plausible bad outcome is a 40–50% de-rating on competition/GTN, not a wipeout.

Recent News & Events

Business environment changed recently? Yes — Novo’s Wegovy MASH approval (Aug 2025) reset the competitive frame; EU approval (Aug 2025) opened a throttled ex-US leg; Blue Owl $350M term loan (Jul 2025) added non-dilutive capital; three BD deals (late 2025–early 2026) built a pipeline; BofA cut to Neutral (Jun 2026). Significant acquisitions? Licensing only (above). Accounting-policy changes? None material. Recent changes — markets/facilities/management? Endocrinology field-force expansion (Q4’25); DTC campaign; Germany launch (Q4’25); Sibold-led commercial org matured; board-class reshuffle (Jun 2026).

APPENDIX B — Source Appendix

Madrigal Pharmaceuticals, Inc. (NASDAQ: MDGL) — as of 2026-07-10

Sources prioritized primary-first. Quantitative figures reconciled to SEC filings; third-party data aggregators used as cross-checks, not authority.

Primary — SEC filings (EDGAR, CIK 0001157601)

• Form 10-K, FY2025 (filed 2026-02-19) — revenue $958.4M, net loss -$288.3M, balance sheet, Roche royalty, patent/Orange Book disclosures, risk factors, share count.
• Form 10-Q, Q1 2026 (period ended 2026-03-31) — net sales $311.3M, SG&A $268.5M, R&D $108.7M, cash $812.8M, Blue Owl debt carrying value.
• Form 10-K FY2024 / 10-Qs — Rezdiffra launch (approval 3/14/24), prior-year comparatives.
• DEF 14A proxy (filed 2026-04-28) — CEO/NEO compensation, PSU design (relative TSR vs Nasdaq Biotech), bonus payout 194.7% of target, insider ownership.
• Form 4 corpus (Jan 2025–Jun 2026) — insider transactions: 259 sales, 62 exercises, 36 grants, 0 open-market purchases; founders Friedman ~$101.8M, Taub ~$48.8M; CEO Sibold ~$7.8M, 24,195 shares held.
• 8-K filings — FDA approval (3/14/24); $600M+ follow-on at $260 (3–4/24); Blue Owl $350M term loan (7/17/25); CSPC & Pfizer in-licensing (late 2025); FY25 results (2/19/26); annual meeting/board changes (6/17/26).
• SC 13D/G — 5% holders: Baker Bros 9.9% (Julian Baker on board), Avoro 9.9%, RTW 8.7%, Paulson 7.4%, Janus 6.9%, Vanguard 6.8%, BlackRock 5.2%.

Primary — company & regulatory

• Madrigal Q1’26 / Q4’25 / Q3’25 / Q2’25 earnings-call transcripts (via ROIC.ai) — patient counts (>42,250), prescribers (>10,000), TAM 315k→460k, F4c ~245k, GTN high-30s, pipeline (>10 programs), “profitability inevitable beyond 2026.”
• FDA — Rezdiffra (resmetirom) accelerated approval, March 14, 2024 (first MASH drug); label F2–F3.
• European Commission / EMA — conditional marketing authorization, Aug 19, 2025 (Germany launch Q4’25).
• AASLD practice guidance — updated to incorporate resmetirom.

Competitive / clinical (public)

• Novo Nordisk / Wegovy (semaglutide) MASH approval, Aug 15, 2025 — ESSENCE Part 1 (resolution 63% vs 34%; fibrosis 37% vs 22%). [PRNewswire]
• Eli Lilly — SYNERGY-NASH (tirzepatide) Phase 2; orforglipron oral GLP-1 (Foundayo, T2D, Apr 2026); retatrutide Phase 3.
• Akero Therapeutics / efruxifermin — SYMMETRY Phase 2b Week-96 cirrhosis (39% vs 15%, NEJM May 2025); Phase 3 SYNCHRONY. [Akero IR]
• Roche / 89bio (pegozafermin) — up-to-$3.5B acquisition, Sept 2025; ENLIGHTEN Phase 3.
• Boehringer/Zealand survodutide; Viking VK2809; Inventiva lanifibranor (NATiV3, topline H2 2026).
• Ribo–Madrigal siRNA collaboration — $60M upfront, up to $4.4B; first candidate-nomination milestone July 2, 2026. [PRNewswire/FierceBiotech]
• BofA Securities — Neutral, PT ~$542, June 11, 2026 (2028 est. $2.7B→$2.6B). [Investing.com]
• ICER — Rezdiffra cost-effectiveness band $39,600–$50,100. BioPharma Dive — launch WAC $47,400/yr (Mar 2024).

Quantitative aggregators (cross-check, not primary)

• ROIC.ai MCP — income statement, balance sheet, cash flow, profitability ratios, enterprise value (~$12.4B), valuation multiples, per-share data.
• AZI valuation_index — own-history percentiles: P/S 4.5th pctile, P/B 90.8th pctile, no P/E (loss).
• FactorsToday — factor loadings (Biotech ~1.16, Market ~0.75), leaderboard (RS 12m +94.8%, 1yr +81%, Sharpe 1.74, beta 0.84, alpha +0.19), related peers (VKTX, ROIV, IMVT, AXSM).
• AZI price history CSV — 5-year OHLCV: low $55.89 (Jan’22), high $602.83 (Dec 22 '25), current $560.01 (Jul 9 '26); MAESTRO-NASH day $63.80→$234.83 (Dec 19 '22).