ResMed: AI Margin Pressure Analysis

Executive Summary

ResMed Inc. (RMD) generated $4.7 billion in fiscal year 2024 revenue (ending June 2024), with adjusted operating margin of approximately 27%. The company dominates the global market for positive airway pressure (PAP) therapy devices for obstructive sleep apnea (OSA) and chronic obstructive pulmonary disease (COPD) ventilatory support. ResMed's AI Margin Pressure Score of 4/10 reflects a balanced picture: the company is both an AI leader in respiratory care (myAir, AirSense 11, cloud-based therapy management) and faces meaningful AI disruption risk from GLP-1 drugs that treat sleep apnea, AI-diagnostic platforms that could change referral patterns, and digital therapeutics that compete with PAP therapy in the mild-to-moderate OSA segment.

Business Through an AI Lens

ResMed's competitive model is already deeply AI-integrated. The AirSense 11 — the company's flagship CPAP device — uses AutoSet intelligent pressure technology that automatically adjusts therapy pressure based on real-time respiratory event detection using machine learning. The myAir patient engagement app tracks therapy compliance, provides AI-generated adherence coaching, and synchronizes data with the U-Sleep cloud platform for provider monitoring. This makes ResMed one of the most AI-native large-cap medical device companies — its devices are effectively AI-powered therapeutic appliances.

The central AI challenge for ResMed is not that AI will displace its devices, but that AI will enable more accurate OSA diagnosis (potentially expanding the market) while simultaneously enabling competitors to offer comparable intelligent pressure technology at lower cost.

Revenue Exposure

ResMed's most significant revenue risk is the GLP-1 disruption. Drugs like Eli Lilly's Zepbound (tirzepatide) and Novo Nordisk's Wegovy (semaglutide) are demonstrating clinically meaningful reductions in OSA severity in obese patients. The SURMOUNT-OSA trial (published 2024) showed that tirzepatide reduced apnea-hypopnea index by 55–63% in moderately obese OSA patients — a magnitude of improvement that could enable some patients to discontinue PAP therapy. If GLP-1 drugs become a first-line OSA treatment for obese patients (estimated 50–60% of the OSA population), the PAP device addressable market could contract by 15–25% over a decade.

AI diagnostic platforms are the second revenue exposure. Companies like Sleepiz (radar-based sleep monitoring), Somnify, and SleepScore (smartphone-based acoustic monitoring) are developing AI-powered sleep apnea screening tools that operate without in-lab polysomnography. If these tools achieve FDA clearance for diagnostic use, they could dramatically expand OSA diagnosis rates (currently estimated at only 20% of the 936 million people with OSA globally) — a market expansion positive for ResMed. However, they also create pathways to diagnosis that bypass the sleep specialist referral channel, potentially accelerating competition from lower-cost PAP manufacturers at point of prescription.

Cost Exposure

ResMed's manufacturing operations — concentrated in Singapore, Australia, and the United States — produce approximately 2.5 million PAP devices and 40+ million mask headgear and cushion components annually. AI-powered manufacturing optimization (vision-based cushion seal quality inspection, predictive equipment maintenance) could improve manufacturing yields and reduce the approximately $1.7 billion annual cost of goods sold by 2–3%, worth $34–$51 million annually.

The Software as a Service business (AirView, U-Sleep, Brightree) generates approximately $500 million in annual recurring revenue with gross margins of 75%+. This segment benefits from AI investment — improving the algorithms powering myAir compliance coaching and AirView provider workflow tools increases net revenue retention from existing SaaS customers, a key metric for the cloud software portfolio.

R&D (approximately 7.5% of revenue, or $352 million) is focused on next-generation AirSense platforms, GLP-1 combination therapy protocols, and AI diagnostics. ResMed is actively investing in AI sleep diagnostics through its Stellar Sleep acquisition and internal research — hedging the GLP-1 risk by helping expand OSA diagnosis while defending the PAP therapy market.

Moat Test

ResMed's moat rests on three foundations. Market leadership: the company holds approximately 55% global market share in PAP devices by revenue (versus Philips Respironics at approximately 30% pre-recall and 15–20% post-recall, and Fisher & Paykel Healthcare at 15–20%). This leadership provides manufacturing scale and provider relationship advantages. Philips' recall of 5.5 million PAP devices in 2021 — due to foam degradation concerns — created a massive market share windfall for ResMed, which is still being fully absorbed as Philips slowly re-enters the market.

The AirSense 11 platform's AI-enhanced AutoSet algorithms, validated across 10+ million patients in the ResMed cloud, represent a proprietary data moat. The therapy effectiveness data informing pressure algorithms is trained on a unique longitudinal dataset that Philips, Fisher & Paykel, and new entrants cannot quickly replicate.

The moat's primary vulnerability is the GLP-1 disruption, which is outside ResMed's control. If obese patients successfully treat OSA through GLP-1 therapy, demand for PAP devices declines in exactly the segment (moderate-to-severe OSA in obese patients) that generates the highest value APAP and BiPAP device revenue.

Timeline Scenarios

1–3 Years

Near-term dynamics are favorable despite GLP-1 uncertainty. Philips' continued absence from the market (FDA import alert remains active as of mid-2024) is providing ResMed with above-market growth of 8–12% annually. The software business is growing at 15%+ driven by Brightree home health software acquisitions and U-Sleep platform expansion. GLP-1 OSA impact in the near term is modest — clinical adoption in sleep medicine is still emerging, insurance coverage is uneven, and most GLP-1 patients in OSA trials continue PAP therapy alongside drug treatment. Revenue growth of 10–12% is achievable through fiscal 2026.

3–7 Years

The medium-term scenario is defined by GLP-1 market penetration dynamics and AI diagnostic expansion. If GLP-1 drugs achieve broad insurance coverage for OSA indication by 2027 (likely), and 30–40% of newly diagnosed moderate-to-severe OSA patients choose GLP-1 therapy over PAP devices, annual PAP device market volume growth decelerates from the current 8–10% to 2–4%. ResMed's strategic response must include: expanding OSA diagnosis rates (currently 20% penetration globally — a 5-point improvement would add 47 million new patients), growing SaaS revenue to 20%+ of total (providing recurring revenue insulation from device volume cyclicality), and developing combination therapy protocols (PAP + GLP-1) that maximize patient outcomes and maintain device utilization among GLP-1 users.

7+ Years

Long-term, the AI opportunity for ResMed is building a respiratory health platform — using continuous monitoring data from PAP devices (sleep quality, respiratory event frequency, oxygen saturation) to predict and prevent adverse outcomes in COPD, heart failure, and obesity hypoventilation patients. ResMed already has 10+ billion nights of therapy data in the cloud — a dataset that could power predictive models for hospital readmission prevention, respiratory deterioration alerts, and integrated care management. If successfully monetized through a per-patient-per-month care management contract with health systems and payers, this data platform could generate $500 million–$1 billion in high-margin recurring revenue by 2035.

Bull Case

In the optimistic scenario, AI-powered diagnostics expand OSA diagnosis rates from 20% to 35% globally, creating 140+ million new potential patients. ResMed's superior therapy compliance algorithms (myAir AI coaching improves 90-day compliance from 65% to 80%) differentiate its devices in a market where payer reimbursement is conditioned on demonstrated compliance. The respiratory care data platform generates $600 million in SaaS revenue by 2030. GLP-1 impact is modest — most patients with moderate-to-severe OSA continue PAP therapy even if using GLP-1. Operating margin expands to 30–32%, and the stock re-rates from 28x to 32x forward earnings.

Bear Case

In the pessimistic scenario, tirzepatide and semaglutide achieve broad coverage for OSA indication by 2026, and sleep medicine specialists recommend a "GLP-1 first" approach for obese OSA patients. PAP device market volume growth decelerates to 0–2% annually as GLP-1 users achieve sufficient weight loss to eliminate OSA. Philips re-enters the US market with an AI-enhanced platform, intensifying pricing competition. ResMed's device ASPs decline 5–8% over three years. Software revenue growth slows as OSA patient volumes stagnate. Operating margin compresses to 22–24%. The stock de-rates from 28x to 18–22x forward earnings.

Verdict: AI Margin Pressure Score 4/10

ResMed earns a 4/10 AI Margin Pressure Score. As an AI-native PAP therapy company, it is better positioned than most medical device peers to benefit from AI-driven diagnosis expansion and therapy optimization. The primary risk is the GLP-1 drug disruption — an external clinical innovation that threatens device demand in the most valuable patient segment. ResMed's cloud data moat, SaaS revenue, and global market leadership provide meaningful resilience.

Takeaways for Investors

ResMed's investment thesis requires a clear-eyed view on GLP-1 disruption timing and magnitude. Investors should monitor: quarterly OSA device revenue excluding Philips recall benefit (organic demand signal — target: 5%+ growth indicates market expansion outpacing GLP-1 displacement), myAir app monthly active users (compliance coaching adoption indicates AI therapy platform stickiness), and SaaS revenue growth rate (target: 15%+ annually to reach 20% of revenue by 2027). The 28x forward earnings multiple pricing reflects both AI-native product positioning and GLP-1 uncertainty. A sustained 2–3 point expansion in ResMed's SaaS revenue mix would be a significant positive re-rating catalyst, confirming the data platform monetization thesis and reducing dependence on hardware device volumes.