Caris Life Sciences filed its first annual 10-K as a public company. Revenue nearly doubled to $812M. The street sees +81% upside. The filing tells a more complicated story.

The ASP Story

The headline number is impressive: $812M revenue, up 97% from $412M. Gross margins expanded from 43% to 66%. Adjusted EBITDA swung $328M positive. Free cash flow turned positive for the first time.

But the revenue bridge table — unusually transparent for a company trying to impress — tells you where the growth actually came from:

• MI Profile ASP increase (solution/payer mix): $311M
• MI Profile volume increase: $52M
• Caris Assure volume + ASP: $54M

78% of molecular profiling revenue growth was ASP, not volume. The FDA approval of MI Cancer Seek in Q4 2024 unlocked dramatically higher Medicare reimbursement rates. This is a one-time step-change. It cannot repeat. Clinical case volume grew 22% — solid but not the doubling story the headline suggests.

2026 will be the first full year comping against elevated ASPs. The growth rate must decelerate sharply. We estimate <30% FY2026 revenue growth (75% probability) versus the 97% just reported.

The DOJ Pattern

In March 2025, the DOJ issued a Civil Investigative Demand under the False Claims Act, focused on the Medicare 14-day rule — which prohibits billing for tests ordered within 14 days before a hospital admission.

This is the same issue CAI settled in 2022 for $2.9M, covering claims through 2018. The new CID covers the post-2018 period — six-plus years of higher-volume testing at higher ASPs. Same rule, same company, second investigation. That's not a one-off compliance lapse. That's a pattern.

The filing notes the investigation focuses on "patients of certain healthcare providers" — suggesting specific referral relationships are under scrutiny. We can't estimate the magnitude (prior settlement was small, but scope was narrow), but the repeat nature is the signal.

The Material Weakness

Identified for FY2024. Still unremediated as of December 31, 2025. The deficiency: insufficient qualified accounting resources for complex transactions. They've hired a new CFO and added staff, but two years without resolution is poor execution for an $812M revenue company.

SOX 404(b) auditor attestation kicks in when EGC status ends — which happens when annual revenue exceeds $1.235B. If CAI hits that in FY2026 (plausible given the ASP base), they face full SOX compliance in their FY2026 10-K while still remediating. The auditor changed from EY to a new firm in the same year the material weakness was disclosed. Coincidence isn't the default assumption.

The Data Moat Is Real — And Consensus

CAI has sequenced 1M+ cases using both whole exome sequencing (WES) and whole transcriptome sequencing (WTS) on every eligible sample. No competitor does both routinely — Foundation Medicine (Roche) and Tempus use targeted panels, which are structurally shallower. This dataset compounds: each case improves the AI/ML algorithms, enables biomarker discovery for biopharma partners, and creates switching costs.

This is a genuine Bustamante-style moat: proprietary data that cannot be synthesized elsewhere.

It's also the consensus bull case. Nine of ten covering analysts rate CAI a buy with a mean target of $33.60 (+81% upside). The moat is visible to everyone. Visible moats aren't alpha.

The Balance Sheet

Post-IPO, CAI holds $796M in cash against a $400M term loan at 10.5% (SOFR + 6.5%), maturing January 2028. Annual interest burden: $42-50M. They can repay — but it costs half their cash reserves. Refinancing at current rates is still expensive. The decision point arrives mid-2027 at the latest.

The $796M looks comfortable until you realize it's the only buffer against DOJ settlement, continued GAAP losses, and new product investment (Caris Detect launch, WGS development). Capital allocation pressure is real.

Sector Context

The diagnostics sector is running hot. Guardant Health +154% 1Y. Exact Sciences +130% 1Y. Illumina +45%. CAI is the clear laggard — down 56% from its post-IPO peak, trading at $18.59 near the 52-week low of $17.15.

This confirms what the regression shows: 90% idiosyncratic variance, -82% annualized alpha. The underperformance is entirely company-specific. The sector isn't the problem. CAI's problems are CAI's problems.

Biopharma data revenue — the long-term monetization narrative — declined 28% to $45M. Management attributes this to contract timing, and the revenue is inherently lumpy. But the direction is wrong when the story needs everything going right.

Insider Activity

Zero open market purchases. The CFO sold $1.65M in December 2025 (conversion + sale, routine post-IPO). Director awards in February 2026 are compensation grants. At $18.59, near the 52-week low, with the street screaming +81% upside — nobody inside the company is buying with their own money.

Absence of buying at post-IPO lows is not bearish by itself (lockup constraints, concentration limits). But it's not bullish either.

Sole Supplier Risk

100% dependent on Illumina for NGS instruments and reagents. No written supply agreements. Illumina can raise prices or discontinue without notice. Any change requires FDA revalidation for approved tests. ILMN is stable today ($124, +45% 1Y), but the concentration risk is structural and unhedgeable.

Factor Decomposition

Six independent thesis factors, each evaluated for edge:

Edge-weighted exposure: 0%. The 90% idio variance is real — returns ARE company-specific — but we have no informational advantage in any of the company-specific drivers. ASP/reimbursement analysis requires dedicated healthcare policy teams. DOJ magnitude assessment requires healthcare fraud expertise.

Conclusion

CAI is a well-run precision oncology company with a genuine data moat that had a transformative 2025. The problem is that 2025's transformation was a one-time ASP reset, not a durable growth inflection. The DOJ investigation is a repeat pattern, not a one-off. The material weakness is a two-year-old problem without resolution. And the bull case is fully consensus — 9/10 analysts already see it.

High idio variance without edge is a coin flip, not alpha. PASS.

The only re-entry trigger worth monitoring: DOJ settlement creates a panic selloff that takes the stock below $15 while the data moat thesis remains intact. Even then, assessing settlement magnitude requires expertise outside our lane.

Evidence