Tempus (TEM) Q3 2026 Earnings Call Transcript

Tempus AI reported Q3 revenue of $411.5 million, up 23% year over year, with oncology volume up 32% and Data and Applications revenue up 33%. Adjusted EBITDA rose to $25.6 million from $1.5 million in Q3 2025, and the company ended the quarter with $854.4 million in cash and marketable securities. Management cited FY2026 revenue guidance of approximately $1.60 billion and adjusted EBITDA guidance of approximately $65 million; it also said Personalis acquisition regulatory review was cleared, with a shareholder vote scheduled for Nov. 30. Management cautioned that Personalis currently loses money on its tests and that Tempus plans to reinvest roughly two-thirds of incremental gross profit, limiting near-term EBITDA maximization.

Analysis

The key debate is whether TEM’s data platform is becoming a recurring software-like business or remains episodic data monetization packaged as AI. Multi-year Lens and GPU commitments could raise switching costs, but reported bookings are not recognized revenue: verify contract duration, delivery obligations, renewal rates, and cash collection before capitalizing them as a durable growth stream. The BMS/Pathos agreement is useful evidence of demand beyond oncology, not yet proof that the model scales across disease areas. Likewise, the projected 2027 oncology price lift is an earnings catalyst only if approvals, reimbursement, and test mix convert it into realized ASP without slowing volumes.

The other asymmetry is MRD: distribution through TEM could accelerate adoption, but management itself says Personalis tests currently lose money. Broader selling before per-test economics turn positive could enlarge revenue while worsening incremental margins. Natera (NTRA) remains a credible incumbent; TEM’s integration and cross-selling advantages are a hypothesis, not demonstrated share gains. For TEM, the EBITDA milestone may support sentiment, but sustained reinvestment and stock-based compensation make GAAP profitability and cash conversion more relevant than a single adjusted EBITDA quarter.

Near term, the call may reinforce growth expectations; over 1–3 months, watch the Personalis vote and transaction close, then reimbursement and unit economics. Over 6–18 months, the thesis depends on ASP realization, MRD gross-profit contribution, and data-contract conversion. The contrarian risk is that investors price long-duration data economics and MRD scale before either is independently visible.

Key Decisions for Investors

• TEM: Avoid chasing an earnings-driven spike. Consider a staged long on weakness only if the FY outlook holds and subsequent reporting shows data revenue converting from bookings into recognized revenue and cash. Reassess if guidance is cut, bookings fail to convert, or gross-profit growth no longer outpaces operating investment.
• Treat MRD as a catalyst watch, not yet a scale thesis. Track Personalis close and shareholder approval, reimbursement milestones, and gross profit per test; delay underwriting a broad sales-force rollout until unit economics are positive. If expansion drives volume but worsens consolidated margin, the bull case is failing.
• Use NTRA as the competitive monitor rather than an automatic short: compare liquid-biopsy and MRD growth, reimbursement, and clinical adoption. A TEM-over-NTRA relative position is only attractive after evidence of TEM share gains; otherwise the pair risks shorting a capable incumbent on management claims.
• Do not assign full recurring-software value to the BMS/Pathos or other data collaborations yet. Verify revenue-recognition schedules, renewal terms, customer concentration, and cash receipts. For MRNA and MRK, the announced collaboration may support their therapy program, but the call provides no basis to underwrite material economics for either partner.