Tirzepatide Franchise Compounds, FY Guide Raised $1.5B at Midpoint, Orforglipron ATTAIN-1 Reads Below the Injectable Bar — Initiating LLY at Hold

Rating Action

Initiating coverage of Eli Lilly & Company (LLY) at Hold. This is our coverage initiation; we have not previously published on LLY. We are starting at Hold rather than Outperform because while the Q2 print is genuinely excellent and the tirzepatide franchise is the most powerful organic-growth engine in large-cap pharma, the orforglipron ATTAIN-1 readout disclosed pre-market on the same day re-shapes the multi-year setup in a way the headline beat-and-raise obscures. Two assets — tirzepatide injectables (Mounjaro, Zepbound) and orforglipron oral — were the basis of the bull case for an unbroken decade of GLP-1-led compounding. ATTAIN-1 keeps the first leg fully intact but shrinks the second leg from a premium oral GLP-1 to a competitive but unremarkable monotherapy oral. That asymmetry in our view leaves LLY balanced rather than long-biased at current expectations. We want to see (a) ATTAIN-2 and the ACHIEVE diabetes readouts later in 2025, (b) early commercial signal on the Zepbound CVS exclusion impact, and (c) the SURPASS-CVOT and SURMOUNT-MMO regulatory trajectory crystallize into a label expansion before moving constructive.

Results vs. Consensus

Management presented Q2 2025 as another quarter of strong financial performance, and the line items support that characterization. Revenue grew 38% YoY, with US revenue also up 38% (volume-led), Europe up 77% in constant currency, and China up 19% in constant currency. The $5.2B Mounjaro and $3.4B Zepbound prints are both materially above sell-side standalone consensus models we track for the franchise. The bigger signal in the print is gross margin expansion of 300 bps to 85% — a function of improved cost of production and favorable mix, partially offset by the 8% US price erosion. Performance margin (gross margin less R&D and SG&A) of 45.9% is up over 6 points YoY despite SG&A growing 30% and R&D growing 23%. Operating leverage is genuinely working at this revenue scale.

The hardest single read in this print: US realized prices fell 8% YoY in Q2, and management explicitly framed single-digit net price erosion as the going-forward structural pattern (“single-digit erosion like we do other chronic medications in net pricing, while maintaining a value point on list”). This is a more conservative pricing posture than the implicit assumption in some of the more bullish sell-side decade-out volume-led models. It is not a step-function downgrade, but it is a tax on terminal-value math and is the single biggest reason we are cautious on multiple expansion from current levels.

Segment Performance

Lilly reports as a single Pharmaceutical Products segment, but management consistently breaks out commercial performance by therapeutic area. We follow the same framework, ordered by 2025 revenue contribution and forward-cycle catalyst density.

Cardiometabolic Health (Diabetes & Obesity) — the franchise

Mounjaro at $5.2B and Zepbound at $3.4B together represent over half of Q2 revenue and account for almost all the YoY growth. Three operational signals worth flagging:

• Supply has normalized. Management produced more than 1.6x the salable incretin doses in 1H 2025 versus 1H 2024 and guided to at least 1.8x in 2H. The Research Triangle Park, NC facility is contributing meaningfully. Two additional US manufacturing sites are to be announced this quarter. The 2024 supply-shortage narrative is now a 2026 capacity-headroom narrative.
• Cash-pay channel (LillyDirect) is scaling fast. Zepbound vials hit 1.1M TRx in Q2 (~20% of total Zepbound TRx; over 35% of new prescriptions). Management positioned this as a hedge/bridge for employer coverage gaps rather than a permanent channel pivot, but the data suggests the cash-pay segment is structurally larger than the original framing implied.
• SURPASS-CVOT cardiovascular outcomes win is a real label-expansion catalyst. Tirzepatide showed an 8% lower MACE-3 rate vs. dulaglutide head-to-head, with 16% lower all-cause mortality. The management-cited indirect-comparison analysis vs. putative placebo (28% MACE-3 reduction, 39% all-cause mortality reduction) is a stronger marketing story than the head-to-head primary readout. Submission to global regulators is planned by end of 2025, opening a 2026 cardiovascular indication.

Counter-balancing those positives: (a) the CVS template formulary excluded Zepbound effective July 1, with management explicitly guiding it as a Q3 volume-growth rate headwind; (b) US employer opt-in for anti-obesity coverage has been “steady around 50% to 55%” for ~18 months by management's own framing; (c) the Canadian semaglutide generic launch expected in early 2026 is a fresh competitive vector for the cash-pay channel that did not exist in the 2024 setup.

Oncology — growing methodically

Verzenio global sales grew 12% YoY (volume-led) and remained the US NBRx and TRx leader in early breast cancer. Jaypirca uptake within its CLL/MCL later-line label is “encouraging” per management; the BRUIN CLL-314 Phase III readout (pirtobrutinib non-inferior to ibrutinib on response rate, with a positive PFS trend favoring pirtobrutinib) sets up a label-expansion submission with BRUIN CLL-313 expected later this year. Olomorasib, the KRAS G12C asset, is now in four simultaneous indications across the SUNRAY-01/02 programs — a high-conviction internal bet on combinations with IO agents in earlier-line lung cancer.

Immunology — Ebglyss inflecting, Omvoh slow

Ebglyss (atopic dermatitis): TRx nearly doubled from Q1, and the drug now has formulary coverage from all three of the largest PBMs representing 90% of commercial lives. This is the textbook biologic launch curve. Omvoh (Crohn's): one quarter into the launch in a competitive market, with positive new-patient-start trends but no inflection language yet. Management description was “making progress” — as much as they ever say about a launch they are not yet excited about.

Neuroscience — Kisunla launch curve

Kisunla (donanemab) for early Alzheimer's: over 1,500 physicians and 150 of the top healthcare organizations have started patients. Two structural unlocks since the last call: (a) FDA approval of a modified dosing schedule strengthening the safety profile, (b) positive CHMP opinion in EU pointing to launch later in 2025. Long-term extension data from TRAILBLAZER-ALZ 2 showed sustained and increasing clinical benefit at 3 years despite most patients having completed treatment within the first 18 months — the limited-duration dosing thesis appears to hold up at extended follow-up.

Key Topics & Management Commentary

1. Orforglipron ATTAIN-1: the asymmetric framing problem

The single highest-stakes data point in the print was disclosed pre-market: ATTAIN-1, the first Phase III obesity (without diabetes) readout for the once-daily oral GLP-1 orforglipron. Top-dose patients lost 12.4% of body weight (about 27 pounds) at 72 weeks, with discontinuations due to adverse events of 5–10% and a GI side-effect profile consistent with the GLP-1 class. Management framing was unambiguously positive:

“We're really pleased with the data we've disclosed this morning, really the idea that you can get 27 pounds of weight loss from a single pill and also get really encouraging effects on other important biomarkers, things like blood pressure, lipids, inflammatory biomarkers and fasting glucose.” — Ken Custer, President, Lilly Cardiometabolic Health

The CSO was more analytically direct about the architecture of the molecule:

“I think, actually, this is as good as it gets for GLP-1 monotherapy here in a once-a-day small molecule oral.” — Dr. Dan Skovronsky, Chief Scientific Officer

That framing — “as good as it gets for GLP-1 monotherapy” — is the analytical key. It is honest, and it is the single most important sentence in the call. Orforglipron is delivering what a pure GLP-1 mechanism without GIP co-agonism can deliver. The injectable bar (Zepbound, a dual GLP-1/GIP) is structurally higher because of the dual agonism, not because of injection vs. oral. The implication: orforglipron is a serious commercial asset for the price-sensitive, primary-care, scalable-global-supply segment, but it is not a tirzepatide-equivalent oral — and it was not engineered to be one.

2. Pricing posture — explicit single-digit erosion as the run-rate

CEO Ricks framed the going-forward US pricing environment in unusually explicit terms:

“Inside of the health care system, which is most of the business, we think that we'll expect single-digit erosion like we do other chronic medications in net pricing, while maintaining a value point on list that makes good sense. That's not considering any new policy environment, but that's our philosophy going forward.” — Dave Ricks, Chair and CEO

This is a more conservative tone on price than the post-Q1 framing implied. The 8% US net price decline this quarter is consistent with that guide. Investors modeling the franchise should not assume price-led upside; the entire growth case is volume + mix + label expansion.

3. Capacity inflection

The capacity story has shifted from a constraint to a tailwind in two quarters:

“We produced more than 1.6x the amount of salable incretin doses during the first half of 2025 when compared to the first half of 2024 ... We expect to produce at least 1.8x the number of salable incretin doses in the second half of 2025 compared to the second half of 2024.” — Lucas Montarce, CFO (Dave Ricks framing in prepared remarks)

This is the structural unlock that makes the +38% revenue print sustainable. Lilly is no longer rationing demand — it is allocating supply to channels and geographies it chooses.

4. MFN / drug pricing engagement — carefully constructive, no concessions

Ricks devoted material prepared-remarks time to the administration's drug pricing posture, framing the company as constructive on the goal of US/Europe price rebalancing but explicit that any imported price-control mechanism into the US system would be value-destructive without parallel structural reform of the gross-to-net environment. This is a careful posture that gives the company room to engage without committing to anything specific. Investors should treat MFN headline risk as live but not management-acknowledged as a 2025 P&L event.

Guidance

FY 2025 guidance was raised on revenue, performance margin, and EPS:

The raise is genuinely operational and not just an FX tailwind, though management did acknowledge an FX component. The Q3 implicit math: with 1H run-rate around $30B (last reported), a $61B FY midpoint implies ~$31B in 2H — back-half-loaded but achievable given the 1.8x salable-doses guide. The CVS Zepbound exclusion is the visible Q3 hairsplitter management chose to flag explicitly.

Analyst Q&A

The Q&A session was overwhelmingly orforglipron-focused (six of the twelve questions touched the asset directly), with secondary focus on pricing, the CVS exclusion, employer coverage, and the cash-pay channel. We paraphrase the most analytically loaded exchanges below.

• Chris Schott (JPMorgan) opened by framing what we view as the single hardest question of the call: how does Lilly position orforglipron in the treatment landscape given a profile that “clearly looks efficacious but maybe a touch below Wegovy”? Ken Custer's response leaned hard on convenience (oral pill, no refrigeration, manufactured at scale) and adjacent biomarker improvements (BP, lipids, inflammation) rather than directly engaging the head-to-head efficacy gap. The non-engagement is itself the answer.
• Seamus Fernandez (Guggenheim) pressed on the pricing trajectory for orforglipron and compounding. Ricks' answer on price-to-value philosophy and single-digit net erosion (quoted above) was the most pricing-explicit exchange of the call.
• Geoff Meacham (Citi) asked about the GI adverse event rate trajectory through the orforglipron studies. Skovronsky was direct that the pattern was as expected for a GLP-1 monotherapy — early-onset, dose-escalation-linked, declining over time — with no notable patient-characteristic predictors.
• Tim Anderson (Bank of America) raised the Canadian semaglutide generic vector for early 2026 and whether it represents a durable cash-pay headwind even if compounders are shut down. Ilya Yuffa redirected to current Zepbound vial momentum (1M+ TRx in Q2, 65% NBRx share) without directly engaging the 2026 generic competitive question. Another non-engagement worth marking.
• David Risinger (Leerink) asked the structural question on US employer coverage: has net covered lives been flat for ~18 months, and what is the outlook into January 2026? Yuffa confirmed the steady 50–55% employer opt-in framing and pointed to new benefit designs (Evernorth out-of-pocket cap, simplified prior auth) as the unlock vector. This is the most candid acknowledgment we have heard from the company that employer coverage has stalled.
• Terence Flynn (Morgan Stanley) asked about expectations for ATTAIN-2. Custer set up “similarly encouraging results” expectation and confirmed the regulatory-submission timeline for chronic weight management remains end-of-2025 with potential approval in 2026.
• Courtney Breen (Bernstein) pressed on apparent differences in nausea/vomiting/constipation rates between ATTAIN-1 and ACHIEVE-1. Skovronsky framed both profiles as in line with prior GLP-1 monotherapy experience in their respective populations (obesity vs. T2D).
• Asad Haider (Goldman Sachs) asked whether the LillyDirect framing has evolved from hedge/bridge to permanent channel given administration interest in DTC. CFO Montarce held the original framing — LillyDirect remains a bridge to employer coverage, not a substitute for it — while acknowledging the 1.1M TRx scale.
• Umer Raffat (Evercore) raised the most analytically direct ATTAIN-1 cross-trial comparison: orfo at ~9% placebo-adjusted ITT vs. oral semaglutide at ~14% placebo-adjusted in OASIS 4, and whether a 45-mg cohort is warranted. Skovronsky declined to engage the exact numbers, framing the result as “what GLP-1 agonism can give you” and pushing back on Wall Street's cross-trial-comparison instinct.
• Steve Scala (TD Cowen) framed ATTAIN-1 as “widely viewed as disappointing” and pushed on the disconnect between the Q1 framing of the Novo/CVS deal as “modest” impact and the Q2 prepared-remarks call-out of Q3 volume-growth pressure. Yuffa quantified the impact as “a couple of hundred thousand TRx volume that may vary” on a base of ~1.7M Q2 TRx adds — meaningful but not thesis-changing.
• Alex Hammond (Wolfe Research) asked why total discontinuation rates were included in the press release this time. Custer benchmarked the 22–24% orforglipron and 29.9% placebo discontinuation rates against analogous obesity studies (placebo ~20s, active ~5 points lower) and called the pattern unremarkable.
• Akash Tewari (Jefferies) asked about US/EU net price parity launches and the path to a bespoke administration solution. Ricks held the line on rebalancing being “a rational thing” long-term but explicit that European reimbursement structures need to rise in parallel and that a sudden collapse of the US gross-to-net is not workable.
• Kerry Holford (Berenberg) asked whether Lilly will follow Novo in offering Ozempic via a cash-pay channel by extending Mounjaro to LillyDirect. Yuffa noted Mounjaro has >90% commercial and Part D coverage and so the cash-pay rationale that drives Zepbound LillyDirect does not apply, while leaving the door open for future brand additions.
• Evan Seigerman (BMO) closed on practical industry actions to support administration MFN goals. Ricks reiterated the on-ramp framing.

What They're NOT Saying

Three notable absences in management framing — useful as forward-cycle watchpoints rather than red flags:

• No direct head-to-head efficacy comparison of orforglipron with Zepbound or Wegovy in commercial-positioning terms. The company has not framed where orfo sits relative to Zepbound for the “same-patient, lower-acuity” segment vs. for net-new primary-care patients. The decision tree for prescribers and the implied cannibalization assumption are conspicuously absent. We expect this to crystallize at ATTAIN-2 readout and approval.
• No quantification of the CVS Zepbound exclusion impact in dollars or TRx beyond “a couple of hundred thousand TRx” range. No mention of compensating reaccess actions with other PBMs. This silence reads as ongoing negotiation rather than acceptance.
• No specific commentary on retatrutide/triple-G commercial positioning or Phase III timelines beyond the new TRIUMPH-7 pain study and MASLD design. Retatrutide is the asset that would, in principle, replace tirzepatide as the franchise driver in the early 2030s. Its silence on the call is structural — readouts are still some quarters out — but it leaves a gap in the post-tirzepatide narrative that the orforglipron readout did not fill.

Market Reaction

The post-print stock reaction was muted-to-negative despite the headline beat-and-raise — almost entirely driven by the orforglipron readout, which was disclosed pre-market and was already in the price by the time the earnings call started. The Q2 numbers and the FY guide raise were re-confirmation of an already-known operational story; ATTAIN-1 was the new information, and it landed below the upper end of the buy-side whisper range. The market response treats this print as: tirzepatide franchise unchanged, orforglipron contribution to terminal-value math marked down. We agree with that read.

Street Perspective

The Street perspective coming out of this print breaks roughly three ways. The bull case being made is that the Q2 print confirms a multi-year +30%-plus revenue run-rate is sustainable on tirzepatide alone, that orforglipron is still a multi-billion-dollar global oral GLP-1 even if it is not a Zepbound-equivalent oral, and that the SURPASS-CVOT cardiovascular indication and SURMOUNT-MMO morbidity/mortality outcome trial open further label-expansion legs. The bear case being made on the Street is that orforglipron's profile shifts the cumulative GLP-1 franchise terminal value materially, that net pricing erosion is now an explicit management-acknowledged single-digit headwind, that employer coverage has stalled, and that the Canadian generic semaglutide vector is a real 2026 competitive event. The middle view — which we share — is that LLY is now a tirzepatide-led-franchise story with orforglipron as a respectable but secondary contributor, and the multiple needs to re-rate to that more concentrated view rather than the previous dual-engine framing.

Model Implications

• FY 2025 revenue: tracking management's $60–$62B with an own modeled point estimate near $61B. The 1.8x 2H salable-doses guide is the supply-side anchor; demand-side conviction is the constraint.
• FY 2025 non-GAAP EPS: tracking $21.75–$23.00, modeled near $22.50. Performance margin upside (toward 45.5%) is feasible if SG&A growth moderates from the +30% YoY Q2 pace.
• 2026 setup: the Canadian semaglutide generic launch in early 2026 and the orforglipron approval/launch (assuming end-2025 submissions land on schedule) are the two largest model-relevant inputs. Net pricing assumption: explicit single-digit decline embedded in our forward run-rate.
• Long-dated terminal-value math: we mark down the orforglipron peak-sales contribution to a level consistent with a competitive-but-undifferentiated oral GLP-1 monotherapy, and rely more heavily on the tirzepatide CVOT-driven cardiovascular indication and SURMOUNT-MMO morbidity/mortality outcome to extend the franchise-revenue tail. Retatrutide is the next-cycle insurance policy but is not yet model-relevant in 2025/2026.

Thesis Scorecard (Initiation)

Net assessment: the franchise (tirzepatide) is exceptional and the operational execution is best-in-class. The bridge to a longer-dated growth narrative (orforglipron oral GLP-1, retatrutide triple-G) is now narrower than it appeared three months ago. Initiating at Hold reflects our view that the multiple is fair-to-full at current expectations and that we want to see ATTAIN-2, the ACHIEVE diabetes program, and early retatrutide TRIUMPH readouts before underwriting an Outperform thesis. We expect to revisit at the Q3 2025 print with the CVS exclusion impact crystallized, additional ATTAIN/ACHIEVE readouts, and clearer 2026 setup language from management.