Eli Lilly is a once-in-a-generation pharma franchise riding the GLP-1 wave with Mounjaro and Zepbound, but at a P/E of 78 and a price/IV ratio of 9.0 the market has already discounted a near-perfect decade. Buffett-Munger discipline says this is Too Hard at today's price.

Plain English

Eli Lilly makes prescription drugs. Their big winner is tirzepatide — branded as Mounjaro for diabetes and Zepbound for weight loss. It works extremely well, and people want it faster than Lilly can make it. The stock has gone up about ten times in five years because of it. The problem: at today's price, you have to believe the company will keep growing 24% every year forever. Drug patents expire. Competitors are coming. The government will negotiate prices. Even great medicines become cheap medicines eventually. Great company, wrong price.

Thesis

Eli Lilly (LLY) sells branded prescription drugs, with the franchise increasingly concentrated in tirzepatide (Mounjaro for type-2 diabetes; Zepbound for obesity), the GLP-1/GIP receptor agonist that, alongside Novo Nordisk's semaglutide, has become the fastest-selling pharmaceutical category in modern history. Secondary franchises in oncology (Verzenio), immunology (Taltz, Ebglyss, Omvoh), and Alzheimer's (Kisunla) round out the portfolio, supported by an oral GLP-1 (orforglipron) and a triple-agonist (retatrutide) in late-stage development.

The compounding case is real but priced. The scorecard puts the composite at 61/100, weighed down by a profitability sub-score of 11/100 and a valuation sub-score of 10/100. Reported ten-year average ROIC is 0.0% and five-year FCF conversion is 0.0% — distortions driven by a multi-year capital expenditure super-cycle (capacity build for tirzepatide) and IPR&D writedowns, but they make a clean owner-earnings read difficult. TTM owner earnings of just $9.6B against a market cap near $865B implies an owner-earnings yield of ~1.1%. Net debt to EBITDA at 20.8x is alarming on the surface, again partly an artifact of the capex cycle, but it removes the balance-sheet cushion you want at this multiple.

The price-to-intrinsic-value ratio is 9.04: $963.33 versus a base IV of $106.57 and a high-case IV of $153.72. The reverse-DCF implies the buyer must believe LLY can grow owner earnings at 23.9% per year forever to justify today's price. That is a bet on a single drug class continuing to be priced like a luxury good while doubling unit volume — exactly the scenario Buffett warns against in 'requires a superstar.' At a price under $200 the math improves toward the high-IV bull case; below $130 it becomes interesting on the base case. At $963 the margin of safety is negative.

Moat

Eli Lilly is, today, one of the strongest commercial pharma franchises on Earth. The question for a Buffett-Munger investor is not whether the moat exists, but whether it is wide enough — and durable enough — to justify a price/IV of 9x. We work the five moat types.

1. Intangibles — patents and brand. Damodaran [4] notes that the canonical pharma moat is legal: 'Firms may enjoy exclusive rights to produce and market a product because they own the patent rights on the product. This would be the case in the pharmaceutical and bio-technology businesses.' Tirzepatide has composition-of-matter patents extending into the 2030s in major jurisdictions. Mounjaro/Zepbound brand recognition is now category-defining among prescribers and patients. This is real, but Damodaran [4] also flags the trap: 'the companies that will see the greatest increases in value are not necessarily the companies that spend the most on R&D, but those who have the most productive R&D.' Lilly's R&D productivity has been excellent recently, but past pharma history is littered with patent cliffs (Lipitor, Humira) where the moat collapsed on a calendar date.

2. Pricing power. GLP-1s currently command list prices over $1,000/month in the US. Net realized prices are lower (rebates, PBM contracts), but pricing power has been demonstrated. The stress test: with a $10B war chest over five years, can a competitor erode this? Novo Nordisk already does. Pfizer, Roche, AstraZeneca, Amgen, Viking, Structure Therapeutics, and several Chinese biotech players have GLP-1, GIP, amylin, and oral programs in motion. By 2030 we will have 6-10 commercialized GLP-1-class molecules. Pricing power is high today, narrow tomorrow.

3. Switching costs. Damodaran [6] notes switching costs are powerful where the customer faces real pain in changing. For a chronic-disease patient who tolerates Zepbound, switching is annoying but not prohibitive — payers drive most switching at formulary renewals. Switching cost is LOW relative to, say, enterprise software.

4. Network effects. None meaningful. Drugs do not exhibit network effects.

5. Cost advantages. Manufacturing capacity for GLP-1s is currently the bottleneck of the entire industry. Lilly has invested ~$50B in capacity (Indianapolis, Lebanon IN, Concord NC, Research Triangle Park, Alzey Germany). In the short run this is a moat — competitors physically cannot make enough peptide. In the long run it is a depreciable asset that everyone is replicating. Cost advantage today: HIGH. Cost advantage in 2030: NEUTRAL.

The Buffett See's Candy test [3]. Buffett's framework is unambiguous: 'Long-term competitive advantage in a stable industry is what we seek in a business. If that comes with rapid organic growth, great.' Pharma is not a stable industry. Patent cliffs, regulatory shocks (Medicare price negotiation under the IRA), and biosimilar/generic entry mean every blockbuster is a depreciating asset whose runway you must constantly outrun. He continues [3]: 'if a business requires a superstar to produce great results, the business itself cannot be deemed great.' Today LLY largely requires Mounjaro/Zepbound to produce great results — concentration risk is severe.

The Damodaran R&D adjustment [1]. Damodaran [1] explicitly recasts pharma PE ratios using P/(E + R&D) to account for the fact that R&D is the real maintenance capex of a drug company. Lilly's headline P/E of 78 already looks expensive; once you correctly capitalize R&D as an investment in future earning power and recompute, the picture is more nuanced — but it does not get cheap. The dispersion in his Table 8 [1] (PE from 6 to 34 across pharma peers) reminds us that the market regularly pays very different multiples for ostensibly similar pharma cash flows; today it is paying the high end for LLY.

The moat exists, is real, and is currently very profitable. But it is a patent-and-capacity moat with a calendar-defined erosion curve, not a See's Candy moat. Moat verdict: NARROW.

Management

David Ricks has been CEO of Eli Lilly since January 2017. Under his tenure the share price has risen from roughly $75 to $963, a remarkable run substantially driven by the discovery, approval, and commercialization of tirzepatide. Capital allocation has five levers; we work each.

1. Reinvestment in the core business. Lilly is pouring an unprecedented amount into manufacturing capacity for incretin peptides — over $50B announced across Indianapolis, Lebanon (IN), Concord (NC), Research Triangle Park, and Alzey (Germany), plus contract manufacturing. This is the right call given that demand vastly exceeds supply and competitors are racing to expand. The risk: these are long-lived, single-use peptide plants. If demand for injectable GLP-1s rolls over (because oral GLP-1s like orforglipron win, or because alternative mechanisms displace tirzepatide), the assets become stranded capacity. Reinvestment in R&D is also heavy and, lately, productive: tirzepatide, donanemab, mirikizumab, lebrikizumab, retatrutide, orforglipron, muvalaplin. Grade for reinvestment execution: A. Grade for the reinvestment-cycle valuation read: B (the headline 0% ROIC is mostly a capex-cycle artifact, but management has not yet proven the new asset base will earn high returns at scale).

2. Acquisitions. Lilly has been disciplined relative to peers. Notable deals include Loxo Oncology ($8B, 2019), Dermira ($1.1B, 2020), Prevail Therapeutics, POINT Biopharma ($1.4B, 2023), Versanis ($1.9B, 2023, bimagrumab for obesity muscle preservation), Morphic ($3.2B, 2024), and DICE Therapeutics. None of these are bet-the-company deals; they are tuck-ins focused on platform extension. Goodwill on the balance sheet has grown but the firm has avoided the catastrophic mega-merger pattern that destroyed value at Pfizer (Wyeth) and others. Grade: B+.

3. Debt. Net debt to EBITDA at 20.8x looks alarming. The proximate cause is heavy capex outpacing operating cash flow over the last 24 months, plus debt issuance to fund capacity ahead of revenue. Interest coverage data is missing in the scorecard but credit ratings remain investment grade (A+/A1). This is a temporary-by-design balance sheet — appropriate IF the demand lasts. If it doesn't, the leverage will compound the equity drawdown. Grade: C+ (acceptable, not conservative).

4. Buybacks. Share count has decreased only 1.77% over ten years — Lilly is essentially not buying back stock at current prices, which is the right answer given a price/IV ratio of 9.04. Management has correctly let buybacks lapse rather than reflexively repurchase shares at $900. Grade: A. (Contrast: many pharma peers continued buying back at peak multiples and destroyed value.)

5. Dividends. Lilly pays a small, growing dividend (~0.6% yield). Reasonable given the capacity build cycle. Grade: B.

Communication. Quarterly disclosures are clear. Management has been measured about competitive risks, has acknowledged supply constraints openly, and has not over-promised on pipeline. Tone is closer to Buffett's 'plain speech' standard than the typical biotech management's. The exception: forward revenue framing for obesity has been aggressive in the way analysts model it, and management has not pushed back hard on the $150B+ TAM consensus.

Insider alignment. Insider ownership is small (typical for a 200-year-old firm). Compensation is performance-linked but not unusually owner-aligned. No notable red flags.

Overall, this is a competent, conservative-by-pharma-standards capital allocator running an unprecedented organic growth and capacity-build sprint. The big call — pour cash into capacity rather than blow it on buybacks — has been correct. The key unanswered question is what the post-capex steady-state ROIC will be on $50B of incretin manufacturing assets if competitive pricing materializes. Capital allocator: B+.

Industry

Branded pharmaceuticals, and specifically the metabolic/obesity segment in which LLY is now concentrated, is structurally attractive — but transitioning. We work Porter's Five Forces.

1. Threat of new entrants — HIGH and rising. New small-molecule entrants need Phase 3 data and FDA approval (high barriers, ~$2B and 10-12 years per drug). But the GLP-1/GIP/amylin/triple-agonist landscape has reduced biological risk: the mechanism is validated, the regulatory path is paved, and at least a dozen well-funded competitors (Novo, Pfizer, Roche, AstraZeneca, Amgen, Viking, Structure, Altimmune, Innovent, Hengrui, Akeso) are pushing programs. By 2028-2030, expect 5-10 new approvals. Capacity is also being scaled by CDMOs and competitors, so the manufacturing moat shrinks. Threat tier: HIGH for the next 5-10 years.

2. Bargaining power of buyers — HIGH and rising. US drug buyers are concentrated PBMs (Caremark, Express Scripts, OptumRx) plus Medicare/Medicaid. Under the IRA, Medicare can now negotiate prices on selected high-spend drugs after 9 years on market for small molecules, 11 years for biologics. Tirzepatide is a peptide that will face this clock. Internationally, single-payer systems already negotiate aggressively — German AMNOG, UK NICE, French HAS. As GLP-1 becomes a maintenance therapy for tens of millions, payers will push hard on price. Power: HIGH.

3. Bargaining power of suppliers — LOW. Lilly is largely vertically integrated for peptide manufacturing. API suppliers, fill-finish CDMOs, and excipient providers have low leverage. CMOs for auto-injector pens have some but not severe leverage. Power: LOW.

4. Threat of substitutes — MEDIUM and rising. Bariatric surgery and lifestyle programs are existing substitutes with much smaller share. New substitutes include emerging mechanisms (amylin agonists like CagriSema, MASH/NASH-specific agents, gene therapies for monogenic obesity, possibly oral peptides). The interesting wildcard is consumer behavior: the moment a competing pill (oral GLP-1, including LLY's own orforglipron) demonstrates equal efficacy and convenience, injectables face substitution pressure within their own class. Threat: MEDIUM today, HIGH by 2028.

5. Industry rivalry — INTENSE in metabolic. LLY versus NVO is a classic two-firm sprint with pricing discipline (so far). Once a third or fourth molecule enters, prisoner's-dilemma dynamics typically erode pricing within 18-24 months. Pharma history is unkind to oligopolies that exceed three players (statins went from premium to commoditized; PCSK9s never achieved their pricing). Rivalry tier: HIGH.

Value pool location and trajectory. Today, value sits with the two manufacturers (LLY ~50%, NVO ~45%, others ~5%), with PBMs capturing meaningful rebates. Over the next decade, value migrates outward toward (a) PBMs/payers via negotiation and substitution, (b) generic/biosimilar manufacturers post-LOE around 2034-2036, and (c) potentially toward oral and next-mechanism players. Manufacturers' share of the value pool likely contracts from ~70% today to 35-50% by 2032.

Industry verdict: Good today, with a clearly visible glide-path to Average within 5-7 years. This is a late-stage harvest moat, not a See's Candy compound. The math you can run on a 'Good' industry deteriorating into 'Average' does not justify a 9x price-to-IV multiple. Industry Verdict: Good (declining toward Average).

Inversion

I am now playing short-seller. The job is to make the strongest credible case that LLY at $963 is worth $200 inside five years. No hedging.

1. The single event that kills this. Orforglipron — Lilly's own oral GLP-1 — or a competitor's oral pill (Pfizer's danuglipron successor, Roche's CT-996, Structure Therapeutics' GSBR-209, AstraZeneca's ECC5004) reads out Phase 3 with non-inferior efficacy to tirzepatide and a safety profile that supports primary-care prescribing without titration friction. The injectable franchise begins to cannibalize within 18 months, and the value of the $50B injectable manufacturing footprint is permanently impaired. Lilly's own oral does NOT save them, because pricing for an oral pill is structurally lower than an injectable, and competition arrives faster. Mounjaro/Zepbound revenue plateaus in 2026-2027, then declines. The market re-rates LLY from a 78x P/E to a 22x P/E on lower earnings — a multiple compression of >70% combined with EPS that is 30% below current consensus.

2. Why the moat is narrower than bulls think. Damodaran [4] is explicit that pharma's moat is legal, not structural: 'they may enjoy exclusive rights to produce and market a product because they own the patent rights.' Patents expire. Tirzepatide composition-of-matter patents start expiring in major markets in 2036, but the more important date is the IRA negotiation date (2031-2032 for Medicare). Net realized prices in the US could drop 40-60% on negotiation. Bulls model 'high single-digit price erosion' — history says drug pricing under negotiation falls in a step function, not a glide. Manufacturing scale, the second pillar of the moat, is being directly attacked: NVO's Catalent acquisition, Hengrui-Eccogene, Pfizer, Roche, and CDMO expansions across India and China are all building peptide capacity. By 2027 there will be more capacity than demand at current prices.

3. Why management is worse than it appears. Ricks has been an excellent operator of an extraordinary tailwind. He has not yet been tested by the down-leg. The decision to lever to 20.8x net-debt/EBITDA and pour $50B+ into single-use peptide capacity is brilliant if demand sustains and ruinous if it doesn't. Pharma history shows that the same management teams that won the up-leg of a blockbuster cycle systematically destroyed capital in the down-leg trying to 'replace' the lost revenue (Pfizer post-Lipitor, Merck post-Vioxx, Bristol post-Plavix). The buyback abstinence at peak prices is praiseworthy, but the implicit substitution — capacity reinvestment — could prove worse. Watch how management responds when revenue first misses: if they accelerate M&A to defend the multiple, that's the canonical capital-destruction pattern.

4. What bulls are extrapolating that won't hold. Bulls are extrapolating: (a) 30%+ revenue growth through 2030; (b) gross margin expansion as scale kicks in; (c) limited price erosion through the IRA window; (d) a $150-200B obesity TAM with LLY taking 50%; (e) successful pipeline transitions to retatrutide and orforglipron; (f) low real-world discontinuation rates; (g) payer willingness to fund chronic GLP-1 use indefinitely. The historical base rate on simultaneously holding 7 favorable conditions is sub-10%. Real-world data already shows ~50% one-year discontinuation, and employers/PBMs are starting to push back on coverage. Reverse-DCF says the market is pricing 23.94% perpetual growth — that is more than Apple, Microsoft, or Google ever sustained over a decade.

5. Valuation trap (multiple compression / regime change). P/E TTM is 78.12 against a 10-year average of 53.17 — itself elevated for pharma. Damodaran [1] shows pharma normalizes at PE 6-15. If LLY re-rates to a 25x forward P/E on plausible 2028 EPS of ~$30 (post-competition, post-IRA), the price is $750. If competition is faster and EPS is $22 at 18x, the price is $400. If a tail-risk reads out badly (e.g., long-term cardiovascular safety signal), the price is $200. The price/IV ratio of 9.04 means a return to even the high IV ($153.72) is an 84% drawdown.

If I am right, the stock could be worth $200 within four years.

Lollapalooza Bias Check

Active analyst biases right now, in priority order:

1. Authority / social proof. Every major sell-side analyst has LLY as a Buy. The CNBC narrative is unrelenting. Index funds are forced buyers as LLY is now a top-10 S&P weight. The bias for me is to assume the consensus already incorporates the information — therefore the price must be roughly right. Buffett would not say that. The consensus has been wrong on every blockbuster cycle in pharma history (statins, biologics, Hep-C, oncology IO).

2. Recency. Tirzepatide's 2023-2025 ramp is unprecedented. Recency bias makes me extrapolate a 24% growth curve forward. I am consciously discounting decade-old base rates of pharma cycle behavior. The correction: weight 25-year pharma history more, 3-year tirzepatide history less.

3. Anchoring. I anchor on $963 as 'the price' and reason about whether it should be 10% higher or lower. The intrinsic-value math says I should anchor on $107 base IV and reason about how much premium to pay for growth. My initial draft trim/buy prices were too close to $963, not close enough to $107. Forced reset.

4. Confirmation. Once I formed the 'Too Hard at $963' thesis, I started reading the canon excerpts looking for pharma patent-cliff confirmations rather than evidence of durable franchise economics. Damodaran [1] R&D-adjusted PE is genuinely a both-sides argument I am underweighting because the conclusion would soften my bear case.

5. Deprival / FOMO super-reaction. Watching LLY appreciate while sitting on the sidelines is psychologically painful. The fear-of-missing-out machinery is the same one that produced the 1972 Nifty-Fifty peak and the 2000 Cisco peak. Cisco at the 2000 peak was a great business; the price still produced 50%+ permanent loss. Munger: 'It is remarkable how much long-term advantage people like us have gotten by trying to be consistently not stupid, instead of trying to be very intelligent.'

6. Commitment / consistency. I do NOT have a prior public position on LLY, so commitment bias is low. If I had been long since $400 it would be much higher; my analysis would be subtly skewed toward 'still attractive.'

7. Incentive (the analyst's compensation structure). Sell-side incentives reward bullishness on momentum names. Buy-side career risk is asymmetric: missing the next $200 of upside is fireable; warning before a 50% drawdown is rarely rewarded. I am writing as a Buffett-Munger analyst with no career risk — but I am aware most professional readers face the opposite incentive.

8. Liking. The drugs work. People I know personally have benefited from Zepbound. It is hard to be cold-eyed about a company whose product is unambiguously good for patients. I am separating 'good for the world' from 'good investment at this price.' These are different questions.

10-Year Outlook

Same fundamental business model in 2036? Mostly yes — Lilly will still be a branded-pharma manufacturer, but the revenue base will look very different. Tirzepatide will be facing peptide biosimilars (LOE around 2036 in major markets) or already eroded via Medicare negotiation. The decisive question is whether the pipeline successors (retatrutide, orforglipron, undisclosed Alzheimer's, oncology) replace the lost cash flow. Pharma base rates say ~30% of blockbuster revenue is replaced by next-generation in-class products from the same firm. That suggests LLY revenue is materially lower as a peak-tirzepatide multiple in 2034-2036 than today.

Customer base larger? Yes — global obesity and diabetes prevalence is rising, and access in middle-income countries is expanding. The patient pool will be larger. The question is whether LLY captures it or whether oral generics, Chinese biosimilars, and broader payer pushback hand most of the volume to lower-priced competitors.

Profit per customer higher? No. Real prices per patient will fall — IRA negotiation, biosimilar entry, and oral competition all push the same direction. Volume up, ASP per patient down.

Moat wider? No. The patent moat is on a known clock. The capacity moat is being replicated by competitors. Brand may persist for 'the original tirzepatide' but does not command much premium against AB-rated alternatives.

Single biggest threat? A tie between (a) faster-than-expected competition from oral GLP-1s, including Lilly's own orforglipron cannibalizing higher-margin injectables, and (b) IRA Medicare negotiation cutting US net realized prices 40-60% in the early 2030s.

The Buffett 10-year test. Buffett asks whether the company will be selling roughly the same product, to roughly the same customer, with roughly the same economics, in 10 years. For See's Candy, Coca-Cola, or American Express — yes. For LLY — only partially. The product mix in 2036 will be substantially different (tirzepatide is dominant today, will be commoditized then), and the economics per unit will be lower. This fails the durability test in the form Buffett requires.

Confidence on the 10-year forecast: LOW. Per the methodology, LOW confidence forces 'Too Hard.' Even if you believe the franchise is durable, a 9.04x price/IV ratio with LOW confidence on the terminal economics is not investable on Buffett-Munger principles. The honest answer is to put LLY in the Too-Hard pile until the price compresses or visibility on post-tirzepatide economics improves substantially.

CONFIDENCE: low

Position Guidance

• Recommendation: Too Hard
• Conviction: high (on the price discipline call, not on direction)
• Target buy price: $180 (~70% discount to current; meaningful margin of safety vs. base IV $106.57 with growth premium for the franchise)
• Target trim price: $250 (above this, even the high-case IV $153.72 plus a generous growth premium is exceeded)
• Position sizing: 0% at current price. If price reaches $180, consider 2-3% starter; full position only at $130 or below. Never more than 5% of portfolio given drug-concentration risk.
• Watch list triggers: (a) orforglipron Phase 3 readout, (b) IRA negotiation list inclusion of tirzepatide, (c) NVO/competitor pricing actions, (d) Q-over-Q net pricing trend in 10-Qs, (e) capex plateau and ROIC stabilization on new asset base.