AU · rank #9 · 2026-07-24
GH
NASDAQ · $20.67B (USD)
The exact numbers the algorithm saw.
| Composite score Z-score blend of all factors below. | 1.572 |
|---|---|
| EPS revisions (30d) Change in consensus EPS estimate over the past 30 days ($/share). | +$0.10 |
| Net revisions (30d) Number of analysts revising upward minus downward. | 2 |
| Insider signal (90d) $78,456 across 1 unique buyers. | 3.000 |
| Analyst upside Consensus target $171.08 | +13.9% |
| Momentum acceleration 3-month minus 12-month momentum. | +11.39% |
| Analyst rating 5 = Strong Buy, 1 = Strong Sell. | 4.70 |
| Gross margin TTM | +64.9% |
| 52w high | $174.08 |
| Market cap (USD) | $20.67B |
The AI research card
Independent qualitative review of each pick before the order is placed. On the rare day the research service is unavailable, the paper book trades on the quant ranking alone and no card appears here.
Summary
Guardant Health is a liquid biopsy and blood-based cancer screening company generating ~$1B+ trailing revenue across oncology diagnostics (Guardant360), Shield colorectal screening, and biopharma/data services.
Rationale
The analyst_upside signal is reinforced by 10 target raises in 30 days, a 4.70/5 average rating from 18 analysts, accelerating EPS revisions across 7/30/60/90-day windows, and a Q1 2026 revenue beat of ~$22M on 48% YoY growth — all consistent with a genuine fundamental re-rating rather than pure sentiment drift.
Material risks
- 1Reimbursement and guideline dependency is the primary thesis-breaker — Shield's >600% screening growth is fragile if ACS colorectal guideline inclusion stalls or payers tighten coverage, which would collapse the screening revenue ramp that is driving the bulk of analyst target raises.
- 2Earnings are reported 6 days after the as-of date (2026-07-30), creating binary event risk that could rapidly invalidate the current analyst consensus if Q2 results or updated guidance disappoint, particularly given the $185–$195M guided FCF burn and the 73% target price dispersion signaling wide analyst disagreement on fair value.
AI verdict council
Each pick is reviewed independently by 3 models before any order. 2 of 3 voted to proceed.