McKesson is one of three oligopolists routing roughly 90% of U.S. pharmaceutical volume, with genuine cost-scale and switching-cost moats. At 36.5x earnings, 46.6x EV/FCF, and 2.7x base-case intrinsic value, the market is paying for GLP-1, specialty, and oncology growth that may simply revert to industry mean.

Plain English

McKesson is a giant trucking company for medicine. When your pharmacy needs Lipitor, Ozempic, or chemotherapy drugs, McKesson moves them from drug factories to the pharmacy or hospital. They earn about three cents on every dollar of medicine they move, but they move hundreds of billions of dollars, so the pennies add up. Only three companies do this at scale in America, and starting a fourth would cost over ten billion dollars. The business is steady and boring in the best way. The problem today is the stock price has nearly tripled in three years, so you are paying a lot for steady and boring.

Thesis

McKesson Corp (MCK) is a Big-3 U.S. pharmaceutical distributor that, along with Cencora and Cardinal Health, controls roughly 90% of branded prescription drug flow in the United States. The business is a quintessential toll-bridge: razor-thin gross margins (~3-5%), enormous turnover, negative working capital float, and a duopoly-like cost structure that punishes any new entrant. The Compounder scorecard gives MCK a composite score of 68 — solid but not elite — driven by a balance-sheet score of 17, capital allocation of 20, and a punishing valuation score of 12.

What compounds here is not unit volume (prescriptions grow 1-3% a year) but mix: specialty drugs, oncology services (US Oncology Network), GLP-1 distribution, and biopharma services (RxTS) are growing faster than the legacy book and carry richer economics. Share count is down 5.95% over a decade — meaningful but not dramatic — and net debt/EBITDA of 0.81x signals a fortress balance sheet. FCF conversion of 287.6% of net income reflects working-capital release more than earnings power, which is a yellow flag for the durability of the FCF run-rate.

The problem is price. At $814, MCK trades at 2.70x base-case intrinsic value of $301.70 and well above even bull-case IV of $561.26. The reverse-DCF demands 23.15% owner-earnings growth in perpetuity — implausible for a 3-5% gross margin distributor. Owning McKesson at $814 requires believing the GLP-1 and specialty boom permanently re-rates a regulated commodity middleman. The price/IV math, not the business quality, is the disqualifier.

Moat

McKesson sits inside a regulated three-firm oligopoly. The five-moat audit produces a real but narrower moat than the multiple implies.

Cost advantages (primary moat). Drug distribution is a scale game: the marginal cost of routing one more pill through an existing 28-distribution-center network is essentially zero, while the fixed cost of building such a network — including DEA-licensed facilities, climate-controlled cold chain, and Drug Supply Chain Security Act (DSCSA) serialization infrastructure — is billions. A new entrant with $10B and five years could perhaps replicate the warehousing footprint, but not the manufacturer contracts, GPO relationships, or the working-capital scale that lets MCK extract roughly 2-3% gross spreads on hundreds of billions of revenue. Damodaran's framing applies cleanly: "Owning or having exclusive rights to a distribution system can provide firms with a cost advantage over its competitors" [6]. The Big-3 collectively distribute ~90% of U.S. branded volume; that share has been stable for 15+ years.

Switching costs (secondary moat). Independent pharmacies, hospital systems, and physician practices integrate MCK's order-management, inventory, and 340B compliance software (CoverMyMeds, RxTS) into their daily operating fabric. Switching distributors means re-papering thousands of NDC SKUs, re-validating cold-chain procedures, and re-integrating EHR systems. Damodaran's Microsoft Office analogy applies — once integrated, switching costs compound quietly [3][6]. However, switching costs in distribution are weaker than in software: the customer can dual-source, and contracts re-bid every 3-5 years (e.g., the periodic Walgreens, CVS, and Walmart re-negotiations).

Intangibles — DEA license + DSCSA scale. The DEA Diversion Control program effectively licenses scale: only a handful of firms can credibly distribute controlled substances at national scale post-opioid settlements. The 2021-2022 $7.4B opioid settlement, while painful, hardened this barrier. New entrants now face a regulatory regime designed around incumbents' compliance investments. This is closer to a legal-monopoly moat in Damodaran's taxonomy [1], but mixed-blessing: regulators preserve the right to cap economics (state pricing, PBM transparency, IRA negotiation flow-through).

Network effects — weak. MCK has data network effects through CoverMyMeds (prior-authorization platform connecting payers, prescribers, pharmacies) and Ontada (oncology real-world data). These are real but modest contributors to a $300B+ revenue base.

Pricing power — weak. This is the moat's Achilles heel. MCK is a price-taker on both sides: manufacturers set drug prices and pay a small handling fee; PBMs and large pharmacy chains negotiate fees down. Buffett's See's Candy framework [2] is the inverse here — MCK has scale and durability but no consumer-brand pricing power. Gross margins have hovered at 3-5% for 20 years.

$10B/5-year competitor stress test. A well-funded entrant (Amazon Pharmacy, a sovereign-backed logistics group) could plausibly take 3-5% share in generics or mail-order specialty within five years. They could not displace the Big-3 in branded distribution to independent pharmacies and hospitals — the contractual and regulatory mesh is too dense. The moat survives the stress test, but with erosion at the margin.

Erosion risks. (1) IRA Medicare negotiation reduces branded drug list prices, which mechanically shrinks distributor gross dollars on those SKUs unless contracts shift to fee-per-unit. (2) Manufacturer direct-to-pharmacy programs (e.g., Eli Lilly's LillyDirect for GLP-1s) bypass distributors entirely. (3) Amazon's slow-but-real encroachment on consumer mail-order. (4) Continued opioid-style litigation tail risk for any newly-recognized harm class.

The moat is genuine — Buffett would recognize the Mayo-Clinic-style durability [2] — but it is a low-margin moat. McKesson earns 3-5% gross on enormous volume; that produces decent dollar profits but constrains the multiple a rational buyer should pay.

Moat verdict: NARROW.

Management

McKesson's management under CEO Brian Tyler has been competent but not exceptional, and the capital-allocation record is a mixed bag of disciplined buybacks, expensive bolt-ons, and one structurally important divestiture (European pharmacy retail).

1. Reinvestment in the business. Capex is modest (~1% of revenue). The business is not capital-hungry — drug distribution is a working-capital game, and MCK's negative working capital position effectively turns suppliers into financiers. Where reinvestment matters is in specialty distribution capacity, biopharma services (RxTS, CoverMyMeds), oncology platforms (US Oncology Network, Ontada), and prescription technology. These investments have been steady but not transformational. Returns on incremental capital are obscured by the net-capital-return cycle (the scorer flagged ROIIC as not meaningful — "Net capital return period; ROIIC not meaningful"). The 10-year ROIC average of 3.41% looks weak in isolation but is misleading because invested capital includes substantial intangibles from past M&A; on tangible operating capital, returns are far higher (negative WC distorts conventional ROIC).

2. Acquisitions. History is mixed. The 2017 Rexall-Health acquisition wrote down. CoverMyMeds (2017, ~$1.4B) has been a strong cultural and strategic fit, becoming the prior-authorization standard. The 2024 PRISM Vision acquisition into ophthalmology and the continued tuck-ins in oncology (US Oncology) are sensible adjacencies. The 2023 divestiture of European pharmacy retail (sold to Phoenix Group) was a smart reversal of an earlier expansion mistake — recognizing that low-margin retail in fragmented European markets was not a sustainable add. Tyler has shown a willingness to admit and unwind prior errors, which is meaningful.

3. Debt. Net debt/EBITDA of 0.81x is conservative for a stable-cashflow oligopolist; capacity for opportunistic share repurchase exists. Interest coverage was unreported in the scorecard but is comfortably high based on historical patterns.

4. Buybacks — the bright spot, with a caveat. Share count is down 5.95% over 10 years, with most of the reduction concentrated in the last 3-4 years as the stock re-rated. Aggressive buybacks at $200-400 in 2020-2022 were excellent capital allocation; buybacks continuing at $700-800 in 2024-2025 are far less attractive given Px/IV of 2.70x. The average P/IV on buybacks over the last decade is probably ~1.0-1.3x — adequate but no longer additive at current prices. Continued buybacks at >2x IV would be a Buffett red flag: "truly great businesses ... can't for any extended period reinvest a large portion of their earnings internally at high rates of return" [2] — but neither should they buy back overpriced stock.

5. Dividends. Modest, ~0.4% yield. Consistent, growing, but not a primary capital-return vehicle. Sensible for a tax-aware shareholder base.

Communication quality. Investor day presentations are clear, segment disclosure is reasonable, and the move to disclose Specialty, RxTS, US Pharmaceutical, and Medical-Surgical separately has improved transparency. Management does not over-promise; they do not, however, openly discuss the IRA flow-through risk or manufacturer direct-distribution threat with the candor a Buffett-style operator would.

Net assessment. Tyler is a steady, disciplined operator. He does not destroy capital, he occasionally creates it (CoverMyMeds, European exit), and he runs a fortress balance sheet. He has not, however, demonstrated the rare capital-allocation genius (Henry Singleton, Tom Murphy) that would justify the current valuation premium. The buyback discipline at current prices will determine whether his grade slips.

Capital allocator: B.

Industry

U.S. pharmaceutical distribution is one of the most stable industry structures in healthcare, but it sits inside a value chain whose pricing power has been migrating away from middlemen for a decade.

1. Threat of new entrants — LOW. The Big-3 (McKesson, Cencora, Cardinal) have a regulatory and scale moat that makes greenfield entry impractical. DEA controlled-substance licensing, DSCSA serialization compliance, manufacturer contracts negotiated over decades, and the working-capital intensity required to finance ~30 days of inventory across thousands of SKUs all combine into a barrier exceeding $10B. The opioid settlements paradoxically hardened this barrier by raising the compliance bar. Amazon Pharmacy is the credible challenger but operates in a different lane (direct-to-consumer mail order generics) rather than three-tier branded distribution. Verdict: very protective.

2. Bargaining power of suppliers — MEDIUM-HIGH. Drug manufacturers (Pfizer, Merck, Lilly, Novo Nordisk, etc.) have substantial leverage because they control the molecules. They set list prices, set distribution-fee schedules, and increasingly experiment with direct-to-pharmacy or direct-to-patient programs (Lilly Direct for GLP-1s being the watershed). Distributors essentially bid for the privilege of routing molecules. However, no manufacturer can bypass the Big-3 entirely without sacrificing access to the long tail of independent pharmacies and health systems. Net: suppliers have meaningful but bounded power.

3. Bargaining power of buyers — HIGH and rising. Customers split into chains (CVS, Walgreens, Walmart, Kroger) and the long tail (independent pharmacies, hospitals, physician practices). Chain customers re-bid contracts periodically and extract every basis point. The 2013 Walgreens-Cencora alliance (extending in 2024) and the 2013 CVS-Cardinal Red Oak JV are examples of vertical integration squeezing distributor economics. Independent pharmacies have less power individually but band together via GPOs (Health Mart, Good Neighbor). Hospital systems negotiate aggressively. Verdict: this is the structural pressure point.

4. Threat of substitutes — MEDIUM and growing. Substitutes include manufacturer direct-distribution (Lilly Direct), specialty pharmacy direct-fulfillment, mail-order, and 503B compounding. None substitute for the bulk of the business today, but the trajectory is unfavorable as biologics and self-administered injectables grow.

5. Industry rivalry — MEDIUM (rational oligopoly). The Big-3 compete but have learned not to compete on price destructively. Customer churn between the three is measurable but contained. Margins have been remarkably stable at 3-5% gross for two decades, which is the signature of a rational oligopoly.

Value pool location and trajectory. The value pool in pharma has been steadily migrating: from manufacturers (still the biggest pool, now squeezed by IRA and biosimilars), to PBMs (peak power 2010-2020, now under regulatory siege), to integrated payer-providers (UnitedHealth/Optum, CVS/Caremark/Aetna). Distributors have held a small but stable share. The growing pools — specialty pharmacy, biopharma services, real-world data, prior authorization — are pools MCK is investing in, but where it competes with Optum, Evernorth, and Cigna's CareCentrix on a less favorable footing.

Industry verdict: Good. Stable, oligopolistic, defensible — but with structural buyer pressure and a value pool that is migrating slowly away. Not Excellent because the long-arc trajectory is sideways-to-down on margin per script.

Inversion

I am playing a short-seller of McKesson at $814. My thesis is that this is a structurally low-margin, regulated commodity intermediary trading at a software multiple driven by a transient mix-shift narrative. Here is why I think the bear case is real.

1. The single event that kills this. The killing blow is not one event but a synchronized two-step: (a) Eli Lilly Direct or a copycat manufacturer-direct distribution program meaningfully scales for GLP-1s (the highest-growth, highest-dollar drug class in pharmacy today), removing $5-15B of MCK's branded gross dollars over 3-5 years; combined with (b) IRA Medicare-negotiated drug pricing reducing list prices on top selling branded drugs, mechanically shrinking the percentage-of-WAC fees distributors earn on those molecules. Either alone would be a 10-15% earnings drag. Together, they would force a re-rating from 36x to 18-20x earnings. At unchanged earnings, that is a $400 stock. At lower earnings, it is closer to $300 — close to the base-case IV of $301.70.

2. Why the moat is narrower than bulls think. Bulls cite "Big-3 oligopoly, 90% share, stable margins for 20 years." That history is real but contains a hidden assumption: that the manufacturer-distributor-pharmacy three-tier system is the only viable architecture. It is not. Specialty pharmacy already bypasses traditional distribution for many high-cost biologics. Lilly Direct demonstrates that a manufacturer with a direct relationship to consumers (via telehealth platforms like Ro and LifeMD) can fulfill prescriptions without using Big-3 distribution. The moat against new entrants in branded oligopoly distribution is wide; the moat against value-chain disintermediation is narrow. Damodaran's warning applies: "Unless these firms can come up with a compelling strategy for increasing switching costs, assuming that growth will continue for extended periods seems dangerous" [6]. That was about search engines; the parallel here is real.

3. Why management is worse than it appears. Brian Tyler runs the company well operationally but has accelerated buybacks materially in the 2024-2025 window when the stock has been trading at 2-2.7x base-case IV. This is the inverse of disciplined buyback behavior. Buffett's framing is unambiguous: buy back stock only below intrinsic value, otherwise return cash via dividends or hold. MCK's continued ~$3-4B annual buyback at >$700 will, in retrospect, look like the late-2000 GE-style buyback at the top of the cycle. Additionally, the executive compensation plan rewards EPS growth — and buybacks at any price grow EPS — which is precisely the bad-incentive cocktail Munger warns about. "Show me the incentives and I'll show you the outcome."

4. What bulls are extrapolating that won't hold. Bulls extrapolate three things: (a) GLP-1 dispensing volume will compound at 25-30% for five more years; (b) specialty drug share of pharmacy spend will continue to grow toward 60%; (c) MCK's RxTS biopharma services will scale to a Visa-like data toll. Each is partially true and entirely overpriced. GLP-1 volumes will normalize as supply catches up, generic semaglutide arrives in major markets in 2026, and Lilly Direct grows. Specialty growth is real but earnings flow disproportionately to specialty pharmacies (Optum, Accredo) not traditional distributors. RxTS is a $1.2-1.5B business; even at 10x revenue and triple in 5 years, it is $40-50B of value, a fraction of MCK's $103B EV.

5. Valuation trap — multiple compression and regime change. The reverse-DCF embeds 23.15% owner-earnings growth — a number wholly inconsistent with a 3-5% gross margin distributor in a low-single-digit-volume industry. EV/FCF of 46.6x is a software multiple. The 10-year average P/E is 25.97x, and TTM is 36.47x — the stock is a full turn above its own decade average on optimism. FCF conversion of 287.6% over 5 years is mostly working-capital release; that is a one-time effect that cannot recur. Owner earnings TTM of $1.55B against an enterprise value north of $100B is an owner-yield of ~1.5%. You are paying software prices for a regulated logistics middleman.

Mathematical inversion. Base case IV is $301.70. Bull case IV is $561.26. Current price is $814.02. Even if the bull case is correct, you are 45% above bull-IV. If the bear case plays out — Lilly Direct scales, IRA bites, multiple reverts to 18-20x on 10% lower earnings — the stock prints $250-300 within 3 years. That is a 60-70% drawdown.

If I am right, the stock could be worth $300 within 3 years.

Lollapalooza Bias Check

Several biases are pulling on me as I write this analysis at $814.

Authority and social proof. McKesson is a Big-3, S&P 500, $103B-EV blue chip. Sell-side coverage is uniformly positive; recent price targets cluster $850-950. The implicit message is "smart institutional money owns this; who am I to disagree." I notice I want to soften the bear case to land closer to consensus. I am consciously rejecting this pull because the math (Px/IV = 2.70x) does not care about consensus.

Recency bias. MCK has compounded at ~30% IRR for 4 years from the 2021 lows. Every recent data point — earnings beats, GLP-1 tailwind, $7.4B opioid resolution — has been a positive surprise. Recency bias whispers that the trend will continue. I am cross-checking with the 10-year P/E average of 25.97x (vs. 36.47x today) which reminds me we are 40% above decade-mean valuation, not at it.

Anchoring. I anchored early to "this is an oligopoly compounder" framing and have to consciously stress-test against the disintermediation thesis. The Lilly Direct evidence is recent and underweighted in my mental model. I am compensating by giving the manufacturer-direct argument more weight in the inversion.

Confirmation bias on the bear side. Once I started writing the inversion, I noticed I was hunting for confirming evidence (IRA, Lilly Direct, opioid tail). I am attempting to balance by noting genuine bull-case strengths: durable balance sheet (0.81x net debt/EBITDA), real switching costs in CoverMyMeds, rational oligopoly behavior, defensible RxTS investments.

Deprival super-reaction (loss aversion). Selling MCK now means losing future GLP-1 upside if the bull case is right. The deprival sting of "missing the next leg" tempts a Hold rating instead of Trim/Avoid. The discipline check: a stock trading at 2.7x IV does not need to be owned to be admired. Berkshire owns very few stocks above 1.0x IV.

Incentive-caused bias in management. I am mindful that Brian Tyler's compensation rewards EPS growth, which buybacks mechanically deliver regardless of price. This is exactly the conflict Munger warned about — and it should make me discount management's continued buyback enthusiasm rather than treat it as a bullish signal.

Net effect. The strongest pulls are authority/social proof and recency. I am leaning into the inversion to counterweight, while acknowledging the business itself is genuinely good and the balance sheet genuinely strong. The disagreement is purely about price.

10-Year Outlook

Same fundamental business model in 2036? Largely yes. The U.S. three-tier drug distribution architecture has structural inertia — regulatory, contractual, and physical — that resists rapid change. McKesson in 2036 will still be moving molecules from manufacturers to pharmacies and providers, still earning 3-5% gross margin, still operating ~30 distribution centers. The mix will be more specialty, more biologics, more services. This is fundamentally a stable business.

Customer base larger? Modestly yes. U.S. drug spend grows at GDP-plus, and demographics are favorable (aging population, chronic-disease prevalence). Branded volume growth will be 1-3%, total dollar growth 4-6%. Customer count (independent pharmacies) is shrinking due to consolidation, but customer-dollar concentration is rising in oncology, specialty, and hospital systems.

Profit per customer higher? Uncertain. RxTS, biopharma services, and Ontada data products could lift profit per customer if MCK successfully monetizes the data layer. Counterweighted by IRA flow-through, manufacturer-direct programs, and chain re-bidding. Plausible base case: profit per script flat to slightly down, profit per customer flat to slightly up via mix.

Moat wider? No. Likely flat to slightly narrower. The regulatory moat (DSCSA, DEA) is already maxed. Switching cost moat (CoverMyMeds, integrated PA) widens slowly. Pricing power is structurally weak and will likely weaken further as direct-to-patient channels grow.

Single biggest threat? Manufacturer-direct distribution scaling (Lilly Direct as the template) for a substantial share of high-margin specialty drugs. Secondary: continued PBM/payer vertical integration (Optum-style) that pulls distribution in-house.

Confidence assessment. I have HIGH confidence the business will exist and earn money in 2036. I have MEDIUM confidence about whether margins/multiples justify $814 today. The qualitative picture is clear; the price is the problem.

CONFIDENCE: medium

Position Guidance

• Recommendation: Avoid (Trim if held)
• Conviction: medium
• Target buy price: $260 (margin-of-safety entry below base-case IV of $301.70, ~14% discount)
• Target trim price: $560 (just below bull-case IV of $561.26 — above this, even the optimistic case is exhausted)
• Position sizing: 0% at $814. Watch-list only. If price drops to $300-350 range on IRA/Lilly-Direct narrative shock, size to 2-4%. If price drops to $200-260 on broad healthcare drawdown, size to 4-6%. Never exceed 6% given narrow moat and structural buyer power.
• Holding-period framework: Long-term holder candidate at the right price; not a forever holding given moat erosion trajectory.
• Catalyst watch: Lilly Direct dispensing volume disclosures, IRA negotiated-drug list expansion, Big-3 quarterly gross-margin trends, MCK buyback pace at current prices.