Initiated 2026-09-27 · Price $70.50 (as of 2026-09-25 close, stockanalysis.com) · Mkt cap $2.04bn · Healthcare / medical devices (sleep apnea implant) · Model: verified

Rating: Hold — Conviction: Low

Inspire has an 85% gross margin, no debt and $415mm of cash and investments, 20% of its market cap. But 2026 revenue is shrinking because Medicare contractors dropped the billing code surgeons used for the implant. The price sits almost exactly on the model's base case: a partial recovery from 2027 with GAAP margins that stay thin because stock compensation costs about twice FY2025's GAAP EBITDA. The outcome depends on reimbursement and GLP-1 drugs, which no one can yet call. The range is wide in both directions, so Hold with Low conviction.

Model value range vs price
Bear $20.26Base $68.79Bull $148.63Price $70.50

Business overview

Inspire sells an implanted hypoglossal nerve stimulator for patients with moderate-to-severe obstructive sleep apnea who can't tolerate CPAP. It is effectively a single-product company, now moving from Inspire IV to the higher-margin Inspire V. Revenue was $912.0mm in FY2025 (+13.6%), but Q2 2026 revenue fell 7.6% to $200.6mm on US weakness. Earnings turn on three things: US implant volume, which depends on surgeon payment and prior authorization; gross margin (85.5% in Q2, +150bp on Inspire V mix); and a heavy sales and marketing cost base. FY2026 guidance, raised in August, is $835–875mm of revenue with a 4–6% adjusted operating margin.

Competition

The implant competitor is Nyxoah (Genio), which launched in the US in 2025. In Q2 2026 its US revenue was €5.2mm (+22% sequentially), and it says early accounts reached 15% share within six months. That is still tiny next to Inspire's ~$200mm a quarter, but it shows where pressure would appear first: share in high-volume centers, and later price. The bigger substitute is drug therapy, since Lilly's Zepbound is approved for OSA in adults with obesity. It shrinks the pool of candidates who fail CPAP, though no source quantifies the effect on referrals yet. On one definition (stockanalysis TTM EV/EBITDA, 2026-09-25), ResMed trades at 14.9x, Stryker at 15.8x and Medtronic at 12.7x; none is a pure comparable.

Bull case

  1. Reimbursement resets and the headwind laps. CMS created new C-codes for HGNS procedures effective 2026-01-01, and the company estimates a $120–130mm FY2026 coding hit ($40mm in Q2). Plays out if surgeon payment stabilizes by early 2027 and growth returns to low teens. Model: rev_growth, exit_ev_ebitda
  2. Inspire V and Project Horizon lift margins. Inspire V raised gross margin 150bp in Q2, and Project Horizon targets ~$30mm of annual savings to redeploy. Plays out if revenue recovers and the savings reach GAAP margin. Model: ebitda_margin
  3. Net cash cushions the downside. Cash and investments are $415.2mm ($14.37 per share), with no funded debt and positive operating cash flow ($23.2mm in Q2). Model: market.cash (FY0 fact)

Bear case

  1. Surgeon economics stay broken. Palmetto, Noridian and CGS removed CPT 64568 for HGNS, and fallback billing is less certain. If surgeons are paid less, they implant fewer, and inventory built for Inspire V ($145.3mm at FY2025) turns slowly. Plays out if US revenue keeps falling into 2027. Model: rev_growth, ebitda_margin, nwc_pct_rev
  2. GLP-1s and Genio shrink the funnel. Fewer patients reach surgery, and Genio takes share at high-volume centers. Plays out if growth stays low single digits after the coding headwind laps. Model: rev_growth, exit_ev_ebitda
  3. Stock compensation is the gap. SBC was $130.3mm in FY2025 against GAAP EBITDA of $64.9mm, so "adjusted" metrics overstate earnings, and the 0.70 beta understates single-product risk. Plays out if SBC stays near $120mm while revenue is flat. Model: ebitda_margin, wacc

Valuation & balance sheet

Metric (definition) Current Own history (range or 5y avg) Peers Source, as-of
EV/EBITDA, aggregator (TTM GAAP, lease-inclusive) 24.5x n/a (EBITDA negative FY2021–FY2023) RMD 14.9x · SYK 15.8x · MDT 12.7x stockanalysis.com, 2026-09-25
EV/EBITDA, model basis (EV $1.62bn / FY2025 EBITDA $64.9mm) 25.0x n/a (not restated) n/a (peers not restated) model-inputs.json
EV/FCF (TTM; FCF before SBC cost) 14.9x n/a (FCF near zero FY2021–FY2023) RMD 18.9x · SYK 24.8x · MDT 21.6x stockanalysis.com, 2026-09-25
Forward P/E 52.0x n/a (unverified) RMD 18.4x · SYK 17.0x · MDT 14.8x stockanalysis.com, 2026-09-25

Model-implied value range (from model-summary.json; generic module, Gordon-growth DCF and exit-EV/EBITDA DCF, midpoints): Bear $20.26 · Base $68.79 · Bull $148.63 per share, i.e. implied returns of −71.3% / −2.4% / +110.8% vs $70.50. These ranges show how the bull and bear drivers translate into value; they are not price targets. The price sits on the base case: the market expects partial recovery, not a return to the old growth rate. The upside if reimbursement resets is larger in dollars than the downside, but it rests on events the company doesn't control.

Re-rating. The exit method assumes a de-rating: 16x against today's 25.0x (−36%), applied to a Y5 margin of 15%, twice FY2025's. It gives $83.17. The Gordon method gives $54.41, 35% below it, so the Gordon method embeds a lower terminal multiple still.

Tail, quantified. The company is a single product, so a harsher case was run outside the committed model. It assumes reimbursement never recovers and share is lost: revenue −8.4%, −10%, −5%, 0%, 0%; EBITDA margin 1.5% falling to 0% then 3%; NWC 25% of revenue; 8x exit; 11.2% WACC; 2% terminal growth. That gives $11.82 (−83%), below net cash per share, because FCF turns negative.

Balance sheet: net leverage not meaningful (net cash $415.2mm, no funded debt); coverage not meaningful; liquidity $415.2mm; no material maturities; ratings n/a (unverified). Leases (~$32mm) are excluded because EBITDA is after rent.

Model note:

Scenario stress test

Reasoned from the bull/bear drivers above. The model column comes from the scenario overlays (Base case + shock).

Scenario Effect Mechanism Magnitude Model Δ value vs Base ($/sh)
S1 Fast equity crash − Small-cap growth medtech de-rates; procedure demand unaffected Med −$4.93
S2 Slow bear / recession − Elective, prior-authorized implant with patient cost-sharing: patients defer and sales costs deleverage. Milder than Base − Bear ($48.53) on purpose: transitory, while the bear case is permanent High −$10.78
S3 Rapid rate shock − Higher discount rate on cash flows back-loaded behind the reset High −$8.47
S4 Slow rate grind − Same channel, grinding Med −$5.03
S5 Soft-landing cuts + Lower discount rate; consumers schedule elective procedures High +$10.40
S6 Recession-driven cuts − Elective volume loss outweighs the lower discount rate Low −$3.34
S7a Credit liquidity shock − Net cash, no balance-sheet channel; forced selling of small-cap growth Low −$2.96
S7b Slow default cycle − Stressed hospitals and ASCs slow elective scheduling Low −$2.42
S8 Stagflation − Higher discount rate; squeezed consumers defer an implant with out-of-pocket cost Med −$7.57
S9a Dollar spike − Small international business (share not verified); translation only Low −$0.16
S9b Dollar slide + Mirror of S9a Low +$0.16
S10 Melt-up + Speculative bid lifts beaten-down small-cap growth Med +$5.92
S11 Energy supply shock − Fuel prices squeeze consumers at the margin Low −$0.26
S12 Mega-cap/AI derating 0 No material effect, not modeled Low $0.00

Currently active/on watch per the playbook: state.md lists S3 as partially active, and the macro log records all three legs crossed on 2026-09-25 (10Y 5.17%). S8, S10 and S11 are on watch. S3 is a High row here, S8 and S10 are Med, and S11 is Low.

Model value change vs Base, by scenario
S2 Slow bear / recession−$10.78S3 Rapid rate shock−$8.47S8 Stagflation−$7.57S4 Slow rate grind−$5.03S1 Fast equity crash−$4.93S6 Recession-driven cuts−$3.34S7a Credit liquidity shock−$2.96S7b Slow default cycle−$2.42S11 Energy supply shock−$0.26S9a Dollar spike−$0.16S12 Mega-cap/AI derating$0.00S9b Dollar slide+$0.16S10 Melt-up+$5.92S5 Soft-landing cuts+$10.40

What would change the call

Upgrades if:

Downgrades if:

Watch items

Sources