A 36-year sleep apnea franchise compounds owner earnings on negative net debt and 16.75% ROIC. At $205 the price-to-IV ratio is 0.54, a 46% discount to base-case intrinsic value with the GLP-1 fear already in the multiple.

Plain English

ResMed sells the small machine and mask that millions of people wear every night to keep their airway open while they sleep. Without it they snore, stop breathing, and slowly damage their hearts. Once a doctor fits you, you buy a new mask from ResMed every six months for the rest of your life. Almost everyone with the disease hasn't been diagnosed yet, so the customer pool keeps growing. ResMed has one real competitor (Philips), who got recalled in 2021. Risk: weight-loss drugs may shrink the pool of patients who need the machine.

Thesis

ResMed designs, manufactures, and sells the cloud-connected CPAP devices and masks that 936 million people worldwide need to breathe safely at night. It is one of two scaled players in a true global duopoly (with Philips Respironics, which exited the market in 2021 after a respirator recall and is only now creeping back). Less than 20% of U.S. sufferers and under 10% of international sufferers are diagnosed, so the addressable market is multiples of today's installed base.

The economic engine is simple and beautiful: sell the device once, then sell the patient a replacement mask, cushion, and headgear every six months for 20 years. The recurring consumables stream is what generates the 10-year average ROIC of 16.75%, the 5-year ROIIC of 17.09%, and 5-year FCF conversion of 123% (working capital is releasing cash). The balance sheet carries net debt-to-EBITDA of -0.56 — i.e., net cash. Share count has grown only 0.45% over a decade. This is, financially, what compounding looks like.

Valuation. The scorer's reverse-DCF requires only 3.65% growth to justify $205, against a TTM owner-earnings base of $1.33B. EV/FCF is 20.2x and TTM P/E is 24.8x versus a 10-year-average P/E of 43.3x. The IV range is $251.6 (low) / $377.8 (base) / $478.2 (high). At $205 the price-to-IV ratio is 0.54: a 46% discount to base case and an 18% discount to the maintenance-capex-stressed low case. Composite score is 82/100. The market is treating ResMed like its terminal value died with Zepbound; the math says the GLP-1 panic is more than priced in. Buy under $230, trim above $460.

Moat

ResMed's moat is multi-layered, rare in medical-device land, and sits squarely in the categories Damodaran calls the most durable: brand/intangibles, switching costs, and cost advantages from scale [1][2][3].

Pricing power / brand intangibles. ResMed is the prescribing default for sleep physicians and the trusted brand for durable medical equipment (DME) channels. AirSense 11 and AirCurve 11 are spec'd into clinical workflows. The company supplied the chair of sleep medicine at Harvard Medical School and built a sleep research center at UCSD; it owns the academic mindshare that Damodaran identifies as the cause, not consequence, of brand value [1]. Pricing data: gross margins have hovered in the high-50s for a decade despite commodity component inflation. Verdict on this layer: strong but not Coke-strong, because reimbursement caps how much price you can extract.

Switching costs. This is the structurally interesting moat. A diagnosed OSA patient signs up to a CPAP regime they will use nightly for the rest of their life. Once titrated to a ResMed AirSense 11 with a particular AirFit mask, they are locked in three ways: (a) the cushion shape of a new mask must match the user's face after months of skin adaptation; (b) the device feeds AirView and myAir, the cloud platforms physicians and DMEs use to bill Medicare's compliance requirement (≥4 hours/night, 70% of nights — without compliance data, the patient loses insurance coverage); (c) muscle memory of pressure curves. Damodaran's exact framing of Microsoft's switching-cost moat applies here: "the most significant barrier to entry...is the cost to the end-user of switching from one product to a competitor" [3]. A patient who finally gets used to CPAP does not re-experiment.

Cost advantages from scale. ResMed manufactures masks and devices in Australia, Singapore, Malaysia, France, China, and the U.S. Mask cushions and headgear are high-volume, custom-tooled silicone pieces. A subscale entrant cannot match the per-unit cost. Combined with a global service network across 11+ countries, scale advantages are durable [2].

Network effects (modest). AirView accumulates data on millions of patients; the more patients on the network, the better the algorithms titrate pressure and predict mask leak. Not a Visa-class network effect, but a real flywheel for compliance.

Regulatory intangibles. Every device requires FDA 510(k) and ex-US equivalents. Every mask cushion design must be cleared. Patents on auto-titration algorithms, AutoSet, and humidifier integration. None of this is impenetrable individually, but the bundle takes years to replicate.

$10B / 5-year stress test. Could a $10B competitor — say, a coalition of Fisher & Paykel Healthcare ($14B), a re-engaged Philips Respironics, or a fresh entrant from Medtronic — kill ResMed's economics in five years? Fisher & Paykel has tried for a decade and remains the perpetual #3 in masks; Philips voluntarily withdrew its Respironics flagship in 2021 over a foam recall and is only tentatively returning under FDA consent decree. The DME channel is a scarce resource — pulmonologist relationships, payer contracts, distribution agreements with Lincare, Apria, AdaptHealth — and cannot be bought in five years. So no, the stress test is survivable.

Erosion risks. (1) GLP-1 weight-loss drugs (Zepbound has an OSA label as of 2024) reduce moderate OSA severity in obese patients, potentially shrinking the funnel. (2) Wearables and home-test screening (NightOwl, just launched April 2025) commoditize diagnosis but expand top-of-funnel — a wash. (3) Reimbursement compression by CMS or European payers — modest, ongoing, manageable. (4) Philips re-entry over a 5-year horizon — real, but late and damaged-brand.

Moat verdict: WIDE.

Management

Capital allocation under CEO Mick Farrell (son of founder Peter Farrell, in seat since 2013) has been disciplined though not flawless.

1. Reinvestment. R&D has run roughly 6–7% of revenue for a decade, funding AirSense 10 → AirSense 11, mask refresh cycles, and AI/ML in AirView. ROIIC of 17.09% over the last five years says reinvested capital is earning well above cost. The 10-year ROIC of 16.75% on a base that includes the 2022 Brightree/MatrixCare era confirms reinvestment is productive. Grade for reinvestment: A.

2. Acquisitions. Mixed record. Wins: Brightree (2016, $800M) — DME software with high attach to ResMed's hardware, cross-sells beautifully. MatrixCare (2018, $750M) — long-term care SaaS, expanded SaaS into a real second leg. VirtuOx (May 2025) — IDTF for home sleep testing, a smart vertical-integration play that captures diagnostic funnel. Caution flags: the SaaS roll-up has been steady but margins are dilutive to the device business and growth has been mid-single-digit, which is fine but not the home-run story management initially pitched. They paid full prices but did not destroy value. Grade: B+.

3. Debt. Net cash position (net debt/EBITDA = -0.56). They used revolver capacity in 2022 to fund MEDIFOX DAN (€1B) without issuing equity, then deleveraged within 18 months. Conservative, appropriate for a medical-device cyclical-on-reimbursement business. Grade: A.

4. Buybacks — and the critical avg P/IV question. Share count has changed +0.45% over 10 years. That means buybacks have just barely offset stock-based compensation; ResMed is not a great buyback story. They paused buybacks during the GLP-1 panic of 2023–2024 when the stock fell to ~$140 (P/IV well below 0.5) — a clear missed opportunity to retire shares cheaply. They have done the math correctly when they have bought, but they have under-utilized buybacks. The lesson Buffett would deliver: when your stock trades at 0.5x intrinsic value and you have net cash, you back up the truck. They didn't. Grade for buybacks: C+.

5. Dividends. Modest, growing, ~1% yield. Reasonable for a compounder that should be reinvesting. Grade: B.

Communication quality. The Farrell family has run quarterly calls with consistent, plain-language guidance for two decades. They explicitly discuss GLP-1 risk on every call (which is correct — confront, don't dodge). The 10-K is well-written, prevalence statistics are sourced, and management talks like operators rather than IR script-readers. They publicly state strategy: more diagnosis, more compliance, more software attach. They are not promotional. Long-tenured CFO. Insider ownership is meaningful but not heroic.

Track record on capital intensity. Owner earnings ($1.33B TTM) exceed reported earnings, and FCF conversion of 123% over five years means working-capital is releasing cash — a sign that growth is not capex-hungry. They re-shore selectively for tariff hedging without panic capex.

Things to watch. (1) The next CEO transition (Mick is mid-50s, no obvious lull). (2) Resistance to issuing equity for any megadeal. (3) Whether buyback discipline improves at lower P/IV ratios.

Capital allocator: B+.

Industry

Porter's Five Forces on the global respiratory-device industry.

1. Competitive rivalry — LOW to MODERATE. A near-duopoly. ResMed (~60% global PAP device share post-Philips recall, higher in masks) versus Philips Respironics (~25%, damaged) versus Fisher & Paykel Healthcare (~10–15%, focused on masks and hospital). Smaller players (BMC in China, Apex, Löwenstein) compete only in price-sensitive emerging markets. Pricing discipline has held through a decade. Critically, Philips's voluntary withdrawal in 2021 over the polyurethane foam recall handed ResMed a multi-year share gift, which the company converted into installed base. Even when Philips returns at full capacity, the displaced patients have been re-titrated onto ResMed devices and will not switch back without medical reason. Rivalry is restrained because the prize for share-grab via price war would be poisoned (DMEs and physicians punish reliability gaps).

2. Supplier power — LOW. Components (motors, sensors, microcontrollers, silicone, PCBs) are commodity-grade. Manufacturing footprint is six countries, providing supplier substitutability. The Texas Instruments / NXP semiconductor exposure is real but not concentrated.

3. Buyer power — MODERATE. Three-tier buying structure: (a) physicians prescribe — sticky, brand-loyal; (b) DMEs (Lincare, Apria, AdaptHealth) buy — consolidating, gaining leverage, but ResMed is must-have inventory; (c) payers (Medicare, private insurance, European national health) reimburse — they cap pricing but require compliance data which only ResMed and Philips deliver. Patient out-of-pocket portion is reasonable. Buyer power is real but balanced.

4. Threat of substitutes — MODERATE and rising. This is the live debate. (a) GLP-1 drugs (tirzepatide / Zepbound) got an FDA OSA label in December 2024. Trial data showed ~50% reduction in apnea-hypopnea index in obese patients. The bear case says the funnel shrinks 20–40%; the bull case says 70%+ of OSA patients are not the obese phenotype that GLP-1s help, and even responders typically still need PAP. The honest answer is somewhere in between, and that uncertainty justifies the multiple compression we have already seen. (b) Mandibular advancement devices (oral appliances) — second-line, ~10% of moderate cases, not new. (c) Inspire Medical's hypoglossal nerve stimulator implant — surgical, expensive ($30K+), suitable only for non-obese moderate OSA patients who cannot tolerate CPAP — a niche, not a replacement. (d) Wearables that just measure sleep — these are diagnostic, not therapeutic, and actually expand the funnel.

5. Threat of new entry — LOW. FDA 510(k) clearance on every device and every mask cushion, a decade of physician-relationship investment, DME channel agreements, and the cost-curve from scale all combine to keep entrants out. The Chinese players (BMC) compete only on price in emerging markets and have made no inroads in the U.S. over 15 years. Philips's recall demonstrates that even a peer scaled competitor can lose its license to operate overnight.

Value pool location and trajectory. Value sits in masks (recurring, 65%+ gross margin) and in the AirView/myAir cloud layer that creates compliance lock-in. The trajectory of the value pool is expanding (more diagnosis, aging population, awareness) net of GLP-1 share loss.

Industry Verdict: Good.

Inversion

I am a short-seller. Here is why ResMed at $205 is a multi-year value trap.

The single event that kills this. Eli Lilly publishes the SURMOUNT-OSA-3 readout: in patients with moderate-to-severe OSA who lose ≥15% body weight on tirzepatide, 71% achieve AHI <5 at 24 months and discontinue PAP therapy. CMS and major commercial payers respond by adding a step-therapy requirement: trial GLP-1 for 12 months before reimbursing CPAP for new diagnoses with BMI >32. The U.S. new-patient funnel — 60% of ResMed's incremental device units — collapses by 35% within 18 months. Mask attach to the existing installed base continues, but the installed base itself starts shrinking as compliant weight-losers exit therapy. Revenue growth turns negative for the first time since 1995. The market re-rates a -3% grower at 12x P/E.

Why the moat is narrower than bulls think. Bulls cite switching costs, but switching costs only matter inside the franchise. They do not protect against patients leaving the category entirely, which is exactly the GLP-1 risk. The mask habit-loop is real but the disease-modifying drug breaks the lock-in by removing the disease. Furthermore, the AirView cloud is not a true network effect — every competitor's cloud will work fine the moment a payer accepts it. ResMed's brand is strong in pulmonology but pulmonologists do not control prescriptions in a GLP-1 world; endocrinologists and primary care do, and they prescribe Zepbound. The intangible regulatory moat is real but Philips has clawed back FDA approval and is shipping AirSense-comparable devices in Europe today; in 24 months they will be back in U.S. retail at 10–15% lower prices. Fisher & Paykel has accelerated its Evora mask line and is winning new fits at 8% share gains in masks per year over the last two years.

Why management is worse than it appears. They missed the buyback window of 2023–2024 when the stock was $140 with P/IV of 0.4. A first-rate capital allocator would have retired 10% of the float; they retired ~1%. They paid full prices for SaaS roll-ups that have not produced the cross-sell magic management originally promised; MEDIFOX DAN's organic growth has decelerated. Mick Farrell has been CEO for 13 years and the operating playbook (more diagnosis, more software attach, hold price) has not evolved meaningfully. Insider selling has accelerated since 2024. R&D as a percent of revenue has crept up while incremental ROIC on R&D has stayed flat — the lazy compounder pattern.

What bulls are extrapolating that won't hold. Bulls extrapolate that 80% of OSA sufferers remain undiagnosed forever, so the funnel is bottomless. But the diagnostic infrastructure (home sleep tests, NightOwl, Apple Watch sleep apnea detection in iOS 18+) is finally arriving at the same time GLP-1s are removing patients from the pool. The two trends collide: more diagnosis of milder cases that will be steered to drugs first, not devices. Bulls extrapolate 16.75% ROIC forever; but ROIC has been falling since 2019 as software acquisitions diluted the device-only base — the trailing-five-year ROIC is below the trailing-ten-year ROIC. Bulls extrapolate 5% organic growth; in a post-GLP-1 world a flat-to-down decade is more plausible.

Valuation trap (multiple compression / regime change). Owner-earnings yield at $205 is ~4.5%. If growth goes to 0% for five years, the fair multiple compresses to 13–15x P/E (medical-device average for no-growth franchises). 13x on $1.33B owner earnings = $17.3B EV, ~$117/share. That's a 43% drawdown from here even before any earnings cut. If you also haircut owner earnings 15% for ASP compression and consumable volume erosion, the math gets to $95–$100. The composite score of 82 is backward-looking; the prospective score is closer to 60 once you adjust for the regime change in obesity treatment. The 10-year-average P/E of 43x was earned in an era when ResMed was the only credible pure-play sleep apnea franchise with a quasi-monopoly post-Philips recall — that regime is ending. Mean reversion of the P/E is the trap. Reverse-DCF requires only 3.65% growth, which sounds easy until you realize 3.65% requires net-positive volume growth in a post-GLP-1 world.

If I am right, the stock could be worth $100 within 3 years.

Lollapalooza Bias Check

Biases active in me as the analyst right now.

Anchoring (high). I am anchored to ResMed's $250+ trading range from 2021–2022 and the IV-base of $377.76. Every time I look at $205 I see a discount, but my anchor may be calibrated to a regime that is ending. The honest question is not "is $205 cheap vs $377?" but "is $205 cheap vs the post-GLP-1 fair value?" — which I have not independently computed.

Authority (medium-high). The scorer outputs a composite of 82/100 and an IV range, and I am instructed to "cite numbers verbatim." That is correct policy for math, but it can leak into qualitative deference: I treat the 82 as evidence the business is great, when in fact 82 is just a backward-looking mechanical aggregation. The scorer's own note flags maintenance capex as uncertain (>50% spread) — which I should not gloss over.

Confirmation (high). I started this analysis primed by the framing "Buffett-Munger compounder." That label biases me toward finding moats and dismissing bear arguments. I caught myself writing "WIDE" before fully stress-testing the GLP-1 case. The inversion section was a forced corrective.

Recency (medium). GLP-1 fear has been the dominant ResMed narrative for ~24 months. I am at risk of either over-weighting it (it has crushed the stock so it must be true) or under-weighting it (it has been the story so long it must be priced in). Both are recency-driven.

Social proof (medium). Multiple respected investors (Andvari, Polen, Blue Tower) own ResMed and have written compounder theses. Knowing they own it makes me more comfortable concluding the same. This is exactly Munger's warning about social proof in investment communities.

Commitment / consistency (low-medium). Once I framed this as a Buffett-style compounder in the headline, I felt pressure to keep the rest of the analysis consistent with that framing. Naming this bias here is the corrective.

Deprival super-reaction (low). Not active here — I have no position to defend.

Incentive (medium). I am rewarded (within this exercise) for producing a confident recommendation. "Hold" is the easiest defensive answer. I should not let that preference for clarity push me past appropriate uncertainty. The honest answer in a regime-change moment is to size the position smaller, not to be more certain.

Net correction. Take the bull thesis at 70% conviction, not 95%. Buy under $230, not under $251. Size at 2–3% of portfolio rather than 5–6%. Hold the buy discipline; do not back up the truck.

10-Year Outlook

Will ResMed in 2036 still be the same fundamental business? Probably yes. Sleep-disordered breathing will still be a chronic, lifelong condition for hundreds of millions of people; CPAP-style positive airway pressure will still be the gold-standard non-invasive therapy for the majority of moderate-to-severe cases; masks and humidifiers will still be sold as recurring consumables; and reimbursement will still gate access in developed markets. The shape of the cash flows ten years out looks recognizably like today.

Will the customer base be larger? Net, probably yes — but with more variance than at any point in the company's history. Three forces compete: (a) GLP-1 weight loss removes 15–35% of the obese phenotype from new prescriptions; (b) home sleep tests, consumer wearables (Apple Watch OSA detection), and primary-care screening expand diagnosis from <20% U.S. penetration toward 35–45%; (c) global aging keeps gross prevalence rising. Net of these, I expect the installed base to be 30–60% larger in 2036 than today — meaningful growth, but slower than the last decade.

Will profit per customer be higher? Roughly flat to modestly higher. ASP compression from Philips re-entry and DME consolidation offsets the mix-shift toward higher-margin masks and software attach. AirView/myAir SaaS layer adds margin if execution is good.

Will the moat be wider? Slightly narrower. Switching costs hold; brand holds; cost advantage holds. But the substitute threat (GLP-1, surgical implants) is structurally larger than it was a decade ago, and the network-effect layer is small. Rate the moat at 7/10 today, 6.5/10 in 2036.

Single biggest threat. A second-generation GLP-1 (oral, cheap, broadly reimbursed) that achieves >70% OSA resolution in obese patients across a 5-year durability study. This is not a tail risk; it is a 25–35% probability over a decade.

CONFIDENCE: medium.

Position Guidance

• Recommendation: Buy
• Conviction: medium
• Target buy price: $230 (below scorer's IV-low of $251.6, builds in 9% extra margin of safety for GLP-1 regime risk)
• Target trim price: $460 (just below IV-high of $478.2; trim into bull-case strength)
• Position sizing: 2–3% starter, scale to 4–5% only if price falls below $180 (P/IV < 0.48) or if a SURMOUNT-OSA durability readout removes meaningful GLP-1 overhang
• Holding period: 5–10 years; this is a compounder, not a trade
• Sell/avoid triggers: ROIIC drops below 10% for two consecutive years; share count grows >2% per year; CMS adopts GLP-1 step-therapy gating CPAP reimbursement; Mick Farrell exits without a credible internal successor