CARL — Q2 2026 Financial Report Analysis

Volume is growing fast; the losses are growing too

Carlsmed makes aprevo, a spinal fusion implant designed for one specific patient. The surgeon's scans go through Carlsmed's software, which plans the correction, and the company then 3D-prints an implant shaped to that patient instead of a surgeon picking from a tray of standard sizes. The company listed on Nasdaq in July 2025 and is not yet profitable.

In the second quarter (April–June 2026), revenue rose 56.7% to $18.9 million. According to the 10-Q, that came "primarily" from more lumbar (lower back) and cervical (neck) procedures, "with our average revenue per procedure substantially constant." So the growth came from more surgeries, not higher prices. Costs grew faster, though. Operating expenses rose 66% to $25.6 million, and the net loss widened 55% to $10.5 million. The business is getting bigger, but it isn't yet getting closer to breaking even in dollar terms.

At a glance

• Revenue +56.7% to $18.9M. All of the growth came from more procedures. Price per procedure was flat, so this is real demand, not a price increase.
• Gross margin 76.8%, up from 73.4%. Each implant costs relatively less to make as volume rises. This is the strongest evidence that the model can eventually make money.
• Net loss $10.5M vs $6.8M a year ago. Spending on sales, R&D and public-company overhead is growing faster than revenue. The operating loss was 58% of revenue, compared with 54% a year earlier.

The numbers

Gross margin is the share of revenue left after the direct cost of making the implants. Operating margin is what's left after all running costs (sales, R&D, administration), before interest and tax. Here it's negative because the company spends more than it brings in. Adjusted EBITDA is the company's own profit measure. It leaves out stock-based pay, depreciation and interest; see below for why that matters. Loss per share looks like it improved sharply, but that's only because the Q2 2025 figure was spread across 4.6 million pre-IPO shares, compared with 27.2 million now. The year-over-year change isn't meaningful, so the table marks it n/m.

For the first half of 2026, revenue was $35.1 million (+57.4%) and the net loss was $19.2 million (vs $12.5 million).

Where the money went

All three expense lines grew faster than gross profit, which rose $5.7 million. Together they added $10.2 million.

• Sales and marketing: $11.9M (+51.5%). The largest line. The 10-Q attributes the increase to $1.5M more in personnel costs, $0.9M more in commissions to independent sales agents, $0.8M more in surgical instruments supplied to those reps, and $0.9M more in marketing. Most of this rises with procedure volume, since commissions and instrument sets scale with cases.
• General and administrative: $7.6M (+127.9%). The fastest-growing line, more than doubling. The drivers were $1.8M more in personnel for "compliance and operational requirements as a public company," $1.4M more in professional and legal fees (including intellectual property matters), and a $0.5M larger provision for customer accounts that may not be paid. Much of this is the cost of being a listed company. It should not keep doubling, but it also won't go away.
• Research and development: $6.0M (+45.2%), mainly "higher personnel costs to support product development and surgical planning AI initiatives."

The quarter-on-quarter trend matters here. Revenue rose $2.8 million from Q1 ($16.1M) to Q2. Operating expenses rose $3.9 million ($21.7M to $25.6M). Gross margin also slipped slightly from Q1's 77.1%.

What the headline numbers hide

• Customers are paying more slowly, and more of them may not pay at all. In the first half, Carlsmed collected $31.5 million in cash from customers against $35.1 million of revenue, or about 90%. Accounts receivable (money hospitals owe it) rose from $11.4 million at year-end to $13.8 million. The allowance for accounts it doesn't expect to collect grew faster, from $1.7 million to $2.8 million. That's about 17% of gross receivables, up from about 13%. The Q2 provision for bad debts was $0.7 million, compared with $0.3 million a year ago. This is the clearest warning sign in the filing: some of the revenue being booked may never turn into cash.
• Cash burn is larger than the net loss. Operating activities used $20.4 million of cash in the first half, slightly more than the $19.2 million net loss. That's despite $3.9 million of stock-based pay being a non-cash cost. The rest went into working capital (the receivables above, plus year-end bonuses paid out; accrued compensation fell from $5.8M to $3.9M).
• The adjusted figure flatters by about $1.9 million a quarter. Adjusted EBITDA (-$8.6M) is smaller than the net loss (-$10.5M) mainly because it excludes $2.3 million of stock-based compensation. That's nearly nine times the $0.26 million of a year ago, because equity grants ramped up after the IPO. Stock pay is a real cost to shareholders through dilution, so the GAAP loss is the better guide.
• Interest income is masking some of the operating loss. The net loss was smaller than the operating loss because Carlsmed earned $0.8 million of interest on its IPO cash, more than covering $0.3 million of interest on its bank loan. That cushion shrinks as the cash is spent.
• No one-offs to speak of. The year-ago quarter included a $0.2 million warrant revaluation charge that is gone now. No restructuring or impairment charges appear in either period.

Cash runway

At June 30, Carlsmed had $89.3 million in cash, short-term investments and marketable securities. It also had $15.6 million of term-loan debt at 7.0% and $34.4 million of further borrowing available under its Customers Bank facility. Liquidity fell $7.8 million in Q2 (from $97.1M at March 31), and $20.6 million over the half. At the first-half pace, the cash on hand covers roughly two more years before any borrowing. Management says its resources are "sufficient... for at least the next 12 months." Hitting a revenue milestone in June also pushed the loan's interest-only period out to April 2028, which delays principal repayments.

The catalyst: Medicare's new payment codes from October 1

Hospitals pay for aprevo, then get reimbursed by insurers, and Medicare and Medicare Advantage are about 56% of the mix. So how much Medicare pays a hospital for an aprevo surgery directly shapes demand. CMS's final rule for fiscal 2027, released July 31, 2026, moves all inpatient lumbar fusions using aprevo into three new payment groups (MS-DRGs 523, 524 and 525) from October 1, 2026. Under the old rules, only some aprevo procedures qualified for a higher payment. The 10-Q argues that the new structure removes hospitals' uncertainty about which cases qualify, and pays amounts that "better reflect their incremental costs." Cervical is in a different position. Inpatient Medicare cases can get up to $21,125 of extra new-technology payment, but outpatient cervical cases get no premium over standard implants after CMS declined a pass-through payment in November 2025. Cervical was about 10% of Q2 revenue in only its second full quarter on sale.

Leadership changes since the quarter

Two executive moves were filed after the 10-Q. Jeff Bertolini, previously SVP of operational excellence, became COO on September 8. CFO Leo Greenstein left effective September 25 and was replaced by Richard Heppenstall, formerly CFO of ZimVie. The 8-K states Greenstein's exit was not due to any disagreement over accounting or disclosure. The same release reaffirmed full-year guidance. Changing CFOs about a year after an IPO is worth noting, but the filings give no sign it relates to the numbers.

Outlook

Management raised full-year 2026 revenue guidance to $74–78 million, from $72–77 million in May (and $70–75 million before that). On 2025 revenue of $50.5 million, that's about 50% growth at the midpoint. Some quick arithmetic: the first half brought in $35.1 million, so the guide implies $38.9–42.9 million in the second half. That would be 38–52% above the second half of 2025 (about $28.2 million), slower than the 57% just delivered, even though the new Medicare codes take effect partway through the period. Either the guide is conservative or growth is cooling as the base gets larger. Q3 results, likely in early November, will show which. The corra personalized cervical plate is due to launch commercially in December 2026, and a bilateral lumbar procedure is also slated to launch.

Our read: demand is clearly there. Trained surgeons are up 67%, procedure volume is up roughly 57% at stable prices, and gross margin is improving. But the path to profit depends on operating expenses growing slower than revenue, and this quarter they grew faster. The things to watch in Q3 are whether operating expenses grow less than revenue, and whether the bad-debt allowance stops climbing as a share of receivables.

Takeaway: Carlsmed's revenue growth is real and comes from more surgeries rather than higher prices, with gross margin near 77%. But a rising bad-debt allowance and operating costs growing faster than revenue mean the loss is still widening. Breaking even depends on the October Medicare change bringing in volume without another jump in spending.