Eli Lilly

Refresh delta (since 2026-06-15)

What changed

1. Q2 2026: revenue $22.97B (+48%), GAAP EPS $7.94, non-GAAP EPS $8.38 [8-K Q2'26]. EPS includes $3.03 of acquired IPR&D charges (Orna $1.23B, Ajax $0.91B, plus smaller items). Without them non-GAAP EPS would have been ~$11.41.
2. Guidance raised a second time. FY26 revenue went from $80-83B (Feb) to $82-85B (Apr) to $85-87B (Aug). The non-GAAP performance margin guide went from 46-47.5% to 49-50.5%. Non-GAAP EPS moved to $35.50-36.50. That is an underlying raise of $2.78 at the midpoint, offset by the $3.03 IPR&D hit [8-K Q2'26]. The EPS guide excludes IPR&D after 2026-06-30, so the July infectious-disease deals and AtaiBeckley can still lower the printed number.
3. Price is now a visible drag. Q2 volume rose 60% while realized price fell 13% [10-Q Q2'26]. In the US, price fell 3%, but about 9% before favourable adjustments to rebate estimates. Outside the US, price fell 36%, mainly because Mounjaro went onto China's reimbursement list (NRDL) [8-K Q2'26]. Self-pay was ~45% of all Zepbound scripts and ~55% of new ones.
4. Mounjaro now sells more outside the US than inside it. Q2 OUS Mounjaro was $5.15B (+172%) against $4.79B in the US [10-Q Q2'26]. OUS revenue overall grew 80%. China grew 93% at constant currency.
5. Retatrutide is ready to file. TRIUMPH-2 (obesity + T2D, 80 weeks: up to 20.8% weight loss, A1C down up to 1.6 points) and TRIUMPH-3 (severe obesity + heart disease: up to 22.6%) read out positive on 2026-07-23. Lilly says the package now supports global filings in obesity, sleep apnea and knee-osteoarthritis pain, with a US BLA in Q1 2027 [8-K Q2'26]. Ricks said Lilly is in active litigation and talks with FDA over whether the drug is classified as a biologic. This matters because a biologic gets 12 years of exclusivity and a new small-molecule drug gets 5.
6. Foundayo is launching slower than the June dossier assumed. Q2 sales were $98M [8-K Q2'26]. US prescribers grew from ~8,000 to ~36,000 between calls. Early uptake was described as slower than Novo's Wegovy pill. The UK approved it for obesity and T2D on 2026-08-11. The US T2D filing went in during Q2, with FDA action expected later this year [8-K Q2'26; call summary]. I could not confirm a US T2D approval as of today.
7. Medicare coverage started. The Medicare GLP-1 Bridge Program has covered discounted Lilly obesity drugs since 2026-07-01, running to 2027-12-31 [10-Q Q2'26]. Lilly said it lifted US coverage for its obesity drugs by ~35%, at $50/month out of pocket.
8. A deal spree, paid for with debt. Six months of business-development cash: $9.8B for acquisitions plus $3.5B for IPR&D [10-Q Q2'26, cash flow]. The main deals were Centessa (orexin agonists, up to $7.8B), Kelonia and Orna (in-vivo CAR-T), Ventyx and Ajax. Since June, Lilly has closed AtaiBeckley (~$2.8B upfront, up to ~$3.8B with CVRs, closed 2026-09-11), bought three infectious-disease companies for up to $3.9B combined, and signed a China discovery deal with InnoCare (up to $3.35B, $100M upfront). Net debt rose from $35.2B to $46.0B in six months [10-K FY25; 10-Q Q2'26].
9. Share price: ~$1,131 (June) to ~$1,182 (2026-09-24 close), +4.5%. The 52-week range is $712-$1,293.
10. June dossier errors, corrected here. (a) It called Q1 EPS of $8.26 non-GAAP. That was the GAAP figure [10-Q Q1'26]; non-GAAP was ~$8.56. The "+170%" was also the GAAP growth rate. (b) It said SURMOUNT-MMO timing was "not pre-specified". ClinicalTrials.gov lists completion as October 2027. (c) The Lebanon, Indiana API plant it expected "near end-2026" is now described as opening in 2027.

What held. Tirzepatide is still the company: Mounjaro plus Zepbound were $14.87B, 65% of Q2 revenue [8-K Q2'26], the same share as in June. The gross margin is still extraordinary at 85.8% GAAP / 86.3% non-GAAP in Q2 [8-K Q2'26]. Lilly still holds about 6 in 10 US obesity prescriptions and 7 in 10 injectable ones. Novo is weaker, not stronger: it announced 13,000 job cuts on 2026-09-21.

Phase A — Understand the business

Lens 1 · Overview: carried, see the previous dossier. Updates from the filings:

• Lilly defines "Key Products" as Ebglyss, Foundayo, Inluriyo, Jaypirca, Kisunla, Mounjaro, Omvoh and Zepbound. They were $15.7B in Q2. The non-incretin Key Products (immunology, oncology, neuroscience) grew 121% [8-K Q2'26].
• The pipeline widened well beyond metabolic disease this year: in-vivo CAR-T (Kelonia, Orna), sleep-wake (Centessa), psychiatry (AtaiBeckley), myelofibrosis (Ajax), vaccines (Curevo, LimmaTech, Vaccine Company). This answers June's Lens 14 question 10 ("what is the company after tirzepatide?") with a spread of bets rather than one.

Lens 2 · Supply chain: carried, with two updates.

• In May Lilly added $4.5B to its Indiana sites, taking Indiana commitments since 2020 past $21B. The Lebanon API site is to make tirzepatide, Foundayo and retatrutide and is due to open in 2027. The Limerick, Ireland site shipped its first commercial batch.
• CapexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. was $5.26B in H1 2026, against $3.21B in H1 2025 (+64%) [10-Q Q2'26]. The 10-Q says capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. "will result in meaningfully higher capital expenditures in the near term."

Lens 3 · Moats: carried. The efficacy ladder got a new rung: an eloralintide + tirzepatide combination showed 17% weight loss at 16 weeks against 10% for tirzepatide alone in Phase 2. The patent position is unchanged: the Mounjaro/Zepbound compound patent runs to 2036 in the US [10-K FY25].

Segments (re-run on filings)

The June table was built from press coverage. Here it is from the 10-Q.

[10-Q Q2'26; 8-K Q2'26]

For FY2025 the 10-K gives Mounjaro $22.97B (+99%), Zepbound $13.54B (+175%), Verzenio $5.72B (+8%) and everything else $22.95B (−1%). Total revenue was $65.18B (+45%), 67% of it in the US [10-K FY25].

The mix moved in two ways. Geography is swinging abroad: OUS was 33% of revenue in FY25 and 37% in Q2 2026. That is volume bought with price, since OUS price fell 36%. The non-incretin book also finally grew (+16% in "Other"), led by the newer immunology and neuroscience launches.

Phase B — Measure performance

Earnings result (latest print: Q2 2026, reported 2026-08-05)

• Revenue $22.97B (+48%). Volume +60%, price −13%, FX +1% [10-Q Q2'26].
• Gross margin 85.8% GAAP, up 1.5 points on better production costs and mix [8-K Q2'26].
• Opex: R&D $3.82B (+14%, 17% of revenue); marketing, selling and admin $3.43B (+25%) for the launches [8-K Q2'26]. Non-GAAP performance margin was 54.8%, up ~9 points.
• Below the line: $2.78B of acquired IPR&D and $0.70B of special charges (deal-related accelerated vesting) [8-K Q2'26]. The tax rate rose to 23.3% from 16.5% because the IPR&D is not deductible.
• EPS: GAAP $7.94 (+26%), non-GAAP $8.38 (+33%) [8-K Q2'26].
• Cash: Q2 operating cash flow was $10.69B, capex $2.93B, free cash flowFree cash flowCash left after paying to run and maintain the business. Unlike profit, it is hard to flatter with accounting choices. $7.76B [10-Q Q2'26, six-month less Q1]. H1 FCFFree cash flowCash left after paying to run and maintain the business. Unlike profit, it is hard to flatter with accounting choices. was $10.76B against H1 net income of $14.49B.
• Quality flags inside the beat: about $0.43B of Q2 US revenue came from changes in rebate estimates for earlier periods (3% of US revenue) [10-Q Q2'26, Note 2], plus the one-off $0.25B Jardiance milestone. Together that is ~3% of the quarter. Management warned that H1 benefited from both [8-K Q2'26, guidance section].

Unusual versus its own history: The volume/price split is the new pattern. In 2025 price was a small drag. In Q2 2026 it took 13 points off growth. Volume is still winning by a wide margin.

Pipeline by phase: what changed since June

Earnings calls (tone)

Transcript unavailable (503). From the summary:

• Still confident. Ricks opened with delivery ("strong business results… expanded access"). The release closes with "Lilly's future, after 150 years, has never been brighter" [8-K Q2'26, verbatim].
• New: management now talks openly about price. The CFO gave the −3% / −9% US split himself and flagged Q3 European seasonality and Q4 US diabetes seasonality. That reads as expectation-setting for a softer H2 after a flattered H1.
• Business development is now a talking point. Jake Van Naarden said Lilly will stay "opportunistic but disciplined" on deals.
• Not addressed in the material I could read: retatrutide tolerability, 2027 guidance, and any capex dollar figure.

Comps: carried

Carried from June. The forward P/E is hard to pin down: sources give 24x (GuruFocus, 2026-09-07), 28x (valueinvesting.io, undated) and 31.5x (2026-09-22). My own figure is ~32.8x the FY26 guide midpoint of $36.00, or ~30x if the Q2 IPR&D charge is added back. The competitive picture moved Lilly's way. Novo is cutting 13,000 jobs under its new CEO. CagriSema is expected to launch for obesity in early 2027. Amgen's MariTide has no Phase 3 readout yet.

Stock-price catalysts since June

• Up: the Q2 beat-and-raise (2026-08-05); the TRIUMPH-2/-3 readouts (2026-07-23); Berenberg's upgrade to Buy with a $1,400 target (2026-09-15); a +2.7% day on 2026-09-24 (the cause was not reported; EASD is the following week).
• Down / capping: HSBC stays at Reduce even after lifting its target to $940; worries about pricing and the slow Foundayo start.
• Consensus: Buy. 29 analysts (25 Buy / 3 Hold / 2 Sell). Consensus targets differ by aggregator: $1,305 (MarketBeat, 2026-09-18) versus a $1,393 median (Yahoo, 2026-09-16).
• Pattern, unchanged since June: the stock moves on readouts and on the price/TAM story more than on the quarter. A 48% quarter with a raise only moved the shares ~4.5% over three months.

Phase C — Judge people & books

Management and capital allocation (re-run: new numbers)

Ricks is still Chair and CEO. The new numbers are all about capital:

• Dividends: $1.73 per quarter (+15%), $3.09B in H1 [8-K Q2'26; 10-Q Q2'26].
• Buybacks: $3.96B in H1 2026 (4.4M shares). Q2 purchases averaged ~$976 a share. $7.0B remains on the $15B program authorized in December 2024 [10-Q Q2'26, Part II Item 2 and Note 10].
• Business development: ~$13.3B of cash in H1, then AtaiBeckley, three ID deals and InnoCare in Q3 (above). Goodwill plus intangibles rose from $12.4B to $27.0B in six months [10-K FY25; 10-Q Q2'26]. Contingent consideration adds up to $5.3B more (Kelonia $3.75B, Centessa $1.54B) [10-Q Q2'26, Note 4].
• Funding: H1 debt issuance was $8.9B of long-term debt plus $5.3B of new commercial paper. Almost all of the long-term debt was one ~$9B multi-tranche bond deal closed on 2026-05-20, with maturities from 2028 to 2066 [8-K 2026-05-20].
• Read: this is a new posture. In June the story was "reinvest in plants and do tuck-ins." Now Lilly is borrowing to buy a portfolio of platforms while also buying back stock at ~$976. Leverage is still modest against ~$30B+ of annualized operating cash flow, but it is no longer net-cash-light by choice. It is net-debt-heavier by choice.
• Insider: the chief accounting officer adopted a 10b5-1Rule 10b5-1 planA schedule an insider sets in advance saying when their shares will be sold, so the sales cannot be timed on private news. A sale made OUTSIDE such a plan is a live decision, which is why it carries more information. plan to sell up to 2,000 shares (2026-05-11) [10-Q Q2'26, Item 5]. That is immaterial.

Forensic red flags (re-run: filings now on the shelf)

• The rebate reserve is the number to watch, and it grew. "Sales rebates and discounts" rose from $17.38B to $21.12B (+22%) in six months, and it was $11.54B at end-2024 [10-K FY25; 10-Q Q2'26]. That is a reserve growing roughly in line with US revenue, which is the healthy direction. But in Q2 the company released estimates for earlier periods worth ~3% of US revenue, which flattered reported US price by ~6 points. One quarter of favourable true-ups is normal. Watch for it recurring.
• Receivables are fine. AR rose 13% ($17.76B to $20.08B) while H1 revenue rose 51% [10-K FY25; 10-Q Q2'26], so collection is keeping up.
• Inventory is a deliberate build. It rose 22% ($13.74B to $16.79B) ahead of the Foundayo launch and retatrutide. This is consistent with June's reading and is not a demand warning.
• Cash conversion recovered. H1 operating cash flow was $16.0B against net income of $14.5B; in H1 2025 it was only $4.8B against $8.4B [10-Q Q2'26]. Note that $3.4B of non-cash acquired IPR&D is added back, so conversion excluding it is about in line with earnings.
• Non-GAAP: Lilly does not add back acquired IPR&D. The gap between GAAP and non-GAAP EPS is small ($7.94 versus $8.38) and comes from amortization, special charges and equity-investment gains [8-K Q2'26]. That is conservative.
• Litigation [10-Q Q2'26, Note 9]: incretin product-liability MDLs (gastrointestinal and NAION eye injury) in E.D. Pa. with Novo as co-defendant; insulin-pricing MDL in D.N.J.; 340B suits; a Texas Medicaid qui tam that was widened to 15 products in 2025; a Brazil environmental case. Press reports put the GLP-1 MDL at 3,200+ federal cases, with bellwether trials scheduled from September 2026. Florida sued insulin makers and PBMs in September.
• SEC enforcement: our model found 0 SEC litigation releases and 0 AAERs since 2021-09-25 (regulatory/regulatory-findings.md). A web search found no FDA warning letter, FTC or DOJ action since June; that is an absence of results, not a confirmed clean record.

Net: no accounting red flag. The pricing reserve and the deal-funded balance sheet are the two things that deserve attention at each print.

Phase D — Project & stress-test

Forward projection (non-GAAP EPS, calendar year)

• FY27E moves up from June's ~$44 to ~$46, because the FY26 underlying base rose ~$2.78.
• Risk to the printed FY26 number: IPR&D after June is not in the guide. If the AtaiBeckley, ID and InnoCare deals are expensed as IPR&D, the reported FY26 EPS could land below $35.50 without anything going wrong in the business [inference].
• Valuation check: at ~$1,182, that is ~25.7x FY27E and ~21.9x FY28E. This is almost the same as June (~25.7x / ~21.3x).
• June forecasts, status: "FY2027 non-GAAP EPS ≥ $42, p=0.72" — the base moved up; I would now put it at p≈0.75, with IPR&D charges as the main way it misses. "Retatrutide filed by end of Q1 2027, p=0.78" — the company now guides to exactly that; I would hold p≈0.75, trimmed for the classification dispute. Neither was logged via our model; that is still open.

Bull vs bear

Bull, updated. Everything the June bull case needed happened on time. Retatrutide read out cleanly across four Phase 3 trials. Medicare coverage opened in July and added ~35% to US obesity coverage. OUS Mounjaro now outsells US Mounjaro. Novo is restructuring. Guidance went up twice. The stock only rose ~4.5%, so the multiple on forward earnings is unchanged while the certainty behind those earnings went up.

Bear, updated.

1. Price is now falling faster. It fell 13% consolidated and 9% underlying in the US. In the US, cash-pay is ~half of new Zepbound scripts at $299-449 a month, and TrumpRx/Medicare pricing points to ~$245 over time. Volume is outrunning price today; the bear case is that it stops doing so once the Medicare Bridge surge laps in 2027.
2. The deal spree is a new risk. Roughly $20B+ of committed deal value in 2026 so far, much of it Phase 1-2 platform bets (in-vivo CAR-T, psychedelics, vaccines), funded by debt. The June bear case never had to price capital-allocation risk; now it does.
3. Foundayo is behind. $98M in its first quarter, against a June street estimate of $1.5-2.8B for 2026. The oral franchise was the pillar of June's "small-molecule cost moat" argument.

Contrarian view the market is missing: The market treats the Q2 IPR&D charge as noise and the price decline as the signal. I think it is the other way round. The price decline is largely a China NRDL mix effect plus deliberate US access pricing, and volume is paying for it. The $3.03 of IPR&D, repeated each quarter, is the real hidden cost of the "what comes after tirzepatide" strategy.

Devil's advocate (short)

• "Guidance went up twice." Some of H1 was one-offs: the Jardiance milestone and ~3% of US revenue from rebate true-ups. Management itself flagged a seasonally softer H2. The implied H2 revenue of $42.2-44.2B is only flat-to-up on H1.
• "Retatrutide is de-risked." Clinically, largely. On regulation, not yet: the BLA-versus-NDA question decides whether it gets 12 years of exclusivity or 5. And a Q1 2027 filing means approval is likely around 2028, beyond most of the FY27 estimate.
• "The balance sheet is fine." It is, today. But net debt rose $10.7B in six months while Lilly kept buying back stock near $1,000. If the ramp slows, that buyback stops looking cheap.
• The single permanent-impairment scenario is unchanged: a class-wide GLP-1 safety signal. The product-liability MDLs going to bellwether trials make it a little more visible, not more likely.

Management questions

Carried from the previous dossier. Four new questions go to the front:

1. Q2 US price fell ~9% before rebate-estimate adjustments. What is the underlying US net price decline you plan for in 2027, once the Medicare Bridge population is in the base?
2. What is the timeline and your fallback in the retatrutide classification dispute, and does an NDA pathway change the Q1 2027 filing date?
3. You have committed $20B+ to deals in 2026, largely early-stage, funded partly with debt, while buying back stock near $1,000. What return hurdle do you use, and what would make you pause buybacks?
4. How much acquired IPR&D from the Q3 deals (AtaiBeckley, the three ID companies, InnoCare) will land in FY26 reported EPS?