INGN — Inogen, Inc. · 2026-09-07 · Verdict: WATCH · Conviction 3
Price $5.42 (screen close, universe file last built 2026-09-06; no live quotes in this sandbox) · Mkt cap $143.8M (26.54M shares) · EV ~$37.0M · EV/EBIT n/m (FY2025 GAAP operating loss $30.2M) · FCF yield negative in each of the last three years · Net cash $106.8M at 6/30/26, no debt · ADV $1.3M · 39% off the 52-week high Sources read: 10-K 2026-02-27 (Items 1, 1A, 7), 10-Q 2026-08-07, DEF 14A 2026-04-28, 8-Ks of 2026-03-30, 2026-04-06, 2026-05-07, 2026-06-11 and 2026-08-06 with EX-99.1 releases, Form 4s (12 of 22 in trailing 12 months). No call transcript or prepared remarks were filed as an exhibit; the pack file labelled "transcript" is the Q2 press release, so section 5 rests on written guidance and results only.
1. What the business actually does
Inogen makes portable oxygen concentrators for long-term oxygen therapy, mostly COPD patients. It sells three ways: to home medical equipment providers, distributors and resellers (45.1% of 2025 US revenue), direct to consumers for cash (29.5%), and direct to consumers as a rental where Inogen bills Medicare or private insurance itself (25.4%) (10-K 2025, Item 1). It is an accredited DME supplier in all 50 states, which its manufacturing competitors are not, and holds 96 issued patents. Two lines are new: Simeox airway clearance, FDA cleared December 2024, and, via a January 2025 Yuwell collaboration, the Voxi 5 concentrator and Aurora CPAP masks. FY2025 revenue was $348.7M, 39.8% international, Medicare fee-for-service only 9.5% of it (10-K 2025, Item 7).
2. Why it is mispriced — the edge case
The edge is hidden-asset, arithmetically real but thin. Cash, securities and restricted cash were $106.8M at 6/30/26 with no debt outstanding (8-K 2026-08-06, EX-99.1), 74% of the market cap. The residual $37M of enterprise value is what the market pays for $355M to $361M of guided 2026 revenue at a 45% gross margin. Selling are passive holders after a multi-year de-rating and anyone marking the company on GAAP earnings, which were losses of $102.4M, $35.9M and $22.7M in 2023 through 2025 (10-K 2025, Item 7). They are not obviously wrong: against the cash sit $62.2M of purchase obligations due within a year, a $28.4M warranty reserve and $15.8M of lease liabilities. The discount is a real fact and so is the reason for it.
3. Unit economics and growth
The engine does not yet earn. FY2025: gross margin 44.2%, sales and marketing $97.7M (28.0% of revenue), G&A $67.4M, R&D $19.4M, giving a $30.2M operating loss and adjusted EBITDA of $2.7M after adding back $8.0M of stock compensation (10-K 2025, Item 7). Volume grows while price falls: 189,400 systems sold in 2025 against 157,500, up 20.3%, on sales revenue up only 5.9%; H1 2026 repeats it, units up 11.4% on sales revenue up 5.6% (10-Q Q2 2026, MD&A).
The reported 2026 growth is currency: H1 revenue rose 3.2% as reported and fell 0.2% in constant currency, Q2 3.0% versus 0.6% (8-K 2026-08-06, EX-99.1). Underneath, H1 US sales fell 3.4% and US rentals 9.8%, net rental patients down to 45,500 from 50,400. Rental capex fell in step, to $1.8M from $4.9M, so the rental book is harvested rather than fed. The real improvement is gross margin, 45.5% in Q2 against 44.8%, on manufacturing leverage and lower open-market semiconductor premiums.
4. Balance sheet and capital allocation
No debt, $106.8M of cash and securities, $123.3M of net working capital (10-Q Q2 2026, MD&A). Free cash flow was negative in every year of the record: roughly negative $30M in 2023, negative $11M in 2024, negative $21.6M in 2025 counting rental equipment production as capex, improving to negative $6.4M in H1 2026. A $30.0M authorisation approved 20 February 2026 expires 31 December 2027; 1,145,150 shares were retired in H1 for $7.5M, leaving $22.5M, about 15% of the market cap. Shareholders then approved 750,000 more plan shares in June (8-K 2026-06-11) and net share count is down only 1.9% year over year. Insiders own 2.5% as a group and bought nothing on the open market in twelve months; the sole open-market trade was an officer selling 10,938 shares at $6.60 (Form 4, 2026-07-06). Yuwell holds 9.6%, bought at $10.36 in February 2025 (DEF 14A 2026-04-28).
5. Management: what they said versus what they did
On 7 May the company reiterated full-year revenue of $366M to $373M, about 6% growth, and said Q1 positioned it for "growth acceleration and improved profitability in the second half" (8-K 2026-05-07, EX-99.1). On 6 August it cut that to $355M to $361M, guided Q3 merely in line with Q3 2025, and headlined a raise in adjusted EBITDA guidance to about $4.0M (8-K 2026-08-06, EX-99.1). The cut is the larger fact and was not the headline; the $4.0M also excludes $3.7M of H1 stock compensation, $1.1M of restructuring and $0.8M of proxy defence costs (10-Q Q2 2026, MD&A). Governance is in motion: activist Kent Lake Partners signed a cooperation agreement on 6 April 2026 adding director Vafa Jamali, with a 4.99% standstill running only to 11 January 2027 (8-K 2026-04-06, Item 1.01); declassification then failed in June (8-K 2026-06-11). The CEO is the third since 2021, with a new CFO in April and a new COO in July.
6. Valuation
Bear (40%): constant-currency revenue flat to negative, the rental book melts another 10% a year, FX reverses, and $4M of adjusted EBITDA stays below real cash costs, burning $8M to $12M a year. Three years out, about $75M of cash and no operating value, roughly $2.90 a share. Base (45%): international and Simeox offset the US decline, revenue about $380M by 2028 at 46% gross margin, adjusted EBITDA $12M, free cash flow modestly positive. At 8x EV/EBITDA that is $96M of EV plus about $90M net cash, roughly $7.15 a share. Bull (15%): the standstill lapses in January 2027 and the company is sold at 0.8x revenue, $286M of EV plus cash, roughly $14.50 a share. Probability-weighted, about $6.70 against $5.42, roughly 24% upside. Reverse DCF: at a 10% cost of capital and no growth, the $37M enterprise value implies only about $3.7M of sustainable annual free cash flow, near 1% of revenue, from a business that has produced none in three years. A low bar, not yet cleared.
7. Catalysts and timeline
Q3 results in early November, testing whether rental patients stabilise near 45,000 and the buyback accelerates. Simeox H SCOPE China results in H2 2026. The Kent Lake standstill expiring 11 January 2027, reopening a nomination window.
8. Risks and pre-registered kill criteria
- Net rental patients below 43,000 at any quarter-end (45,500 at 6/30/26): the melt is accelerating, not bottoming.
- Cash, securities and restricted cash below $95M at any quarter-end after adding back buyback spend: burn above roughly $12M annualised.
- Constant-currency total revenue negative for two consecutive quarters (H1 2026 was already negative 0.2%).
- The repurchase programme suspended, or more than $10M of the $30M unused at 31 December 2027. Upgrade triggers: rental patients flat sequentially for two quarters, adjusted EBITDA above $3M per quarter, repurchases above $5M per quarter. Unquantified risks: a future Medicare competitive bidding round for oxygen (10-K 2025, Item 1A), and reliance on Yuwell as both supplier and 9.6% holder.
9. Verdict and one-paragraph summary
WATCH, conviction 3. Inogen is a genuine net-cash stub: $106.8M of cash and securities with no debt inside a $143.8M market cap means you pay about $37M, a tenth of revenue, for a business selling $355M to $361M of respiratory equipment at a 45% gross margin. But that business consumes cash rather than producing it, and the 2026 growth is a mirage: H1 revenue rose 3.2% as reported and fell 0.2% in constant currency, US sales fell 3.4%, rental patients fell from 50,400 to 45,500, and in August management cut full-year revenue guidance from $366M-$373M to $355M-$361M while headlining a raise in a $4.0M adjusted EBITDA figure that excludes roughly $7M of annualised stock compensation. Gross margin is improving and the H1 burn narrowed to $6.4M, so the cash is not evaporating; but no insider has bought a share in a year, and only $7.5M of a $30M authorisation has been spent at prices this low. A cheap balance sheet on an engine that has not proven it can earn; the November print shows whether rental patients flatten and whether the buyback runs.
Research for discussion, not investment advice. Positions and sizing are the reader's decision.