Sword Health

Price discovery arrives on someone else's schedule. Either the IPO opens the books on gross margin, retention and clinician cost and the multiple resets toward the listed comp — which on the last stamped revenue pair implies nearer $2-2.5B than $4B before revenue growth is credited — or an adverse verdict on 14 Sep 2026 hands away ~5% of the equity mid-process. Meanwhile Hinge compounds at +53% on 86% gross margin with $99.6M of quarterly FCF, buying adjacencies in cash while Sword pays in stock priced off a mark that has never met a bid.

The mark holds roughly where it is through the next negotiated round — plausibly the ~$50M at ~$5B the CEO forecast — because nothing forces price discovery: the company generates cash, does not need capital, and controls the timing. The gap between the ~16.7x headline and the comp's ~7-9x closes by revenue growing into the mark rather than the mark moving.

Sword lists in 2028 on audited profitability — four pillars cross-selling into a renewed employer base plus a statutory-reimbursed German line, at a gross margin approaching the ~86% Hinge now prints. On Hinge's ~7-9x revenue, a base compounding from ~$240M into the high hundreds of millions carries $4B comfortably and clears it at IPO, into the receptive tape canon expects once a pure-play lists.

• Cash-flow positive at ~$240M run-rate (17 Jun 2025) — scarce in digital health, and why the IPO is deferred by choice
• Four clinical pillars on one AI care specialist and one enrolment, so each condition is a cross-sell, not a new sales motion
• Outcome pricing with third-party claims validation ($3,177 saved per engaged member, 3.2x ROI) — the evidence buyers now demand
• Germany's DiGA statutory reimbursement acquired with Kaia (>70M lives) — payer-funded revenue independent of US employer budgets
• 90+ US/EU patents across 53 inventions, incl. US 12,573,494 on the dual-signal reinforcement-learning architecture

• The $4B was set by a $40M round (~1% dilution) and is ~13.7 months stale — canon's break-point for the private aggregate
• No disclosed or audited economics — no gross margin, no clinician cost per engaged member, no retention curve, no cash balance
• COGS carries state-licensed clinician labour AND a physical device, so software margin is unreachable by scaling code
• Second to Hinge Health on every disclosed metric — not the revenue leader in its own category
• ~5% of the equity (~$200M at the mark) goes before a jury on 14 Sep 2026 after two failed motions to dismiss

• Pulse (11 Mar 2026, sold on outcomes) attacks the largest employer spend pool of the four pillars
• DiGA is a template: statutory reimbursement plus RCT evidence repeats into other single-payer systems
• Falling inference cost plus Phoenix taking more of the session converts clinician COGS into margin expansion
• Consolidator role in a fragmenting market — Kaia showed Sword will buy distribution and reimbursement
• A 2028 IPO on audited profitability lets Sword list on earnings, and canon notes the first listed pure-play re-rates the cohort

• Hinge Health as an armed public competitor — GAAP-profitable, $99.6M quarterly FCF, buying adjacencies in cash ($105M for Cylinder Health) while Sword pays partly in stock
• Point-solution rationalisation: employers cutting vendor counts, and winner-take-most stranding the runner-up
• Health plans and carriers bundling MSK and behavioural into the medical plan, removing the benefit line entirely
• Outcome pricing cuts both ways: missing contracted thresholds withholds revenue on cost already incurred
• Marking risk into a listing — a $4B built on ~1% dilution rounds has never met a clearing bid