AI-Bio
A real, regulator-embedded biosimulation moat trading like a broken SaaS roll-up — but the bet only works if the new CEO can re-accelerate software past the +7% Q1 stall before the M&A floor (SLP at 2.5x) is the only thing holding the stock.
Research
The verdict
A real, regulator-embedded biosimulation moat trading like a broken SaaS roll-up — but the bet only works if the new CEO can re-accelerate software past the +7% Q1 stall before the M&A floor (SLP at 2.5x) is the only thing holding the stock.
Certara sells the picks-and-shovels of computational drug development. Its product is biosimulation — computer-aided mathematical simulation of how a drug behaves in the human body — sold as the core technology layer of Model-Informed Drug Development (MIDD), the regulator-endorsed approach to using models instead of (or to de-risk) physical experiments.
The business splits two ways:
Why the moat claim is credible: per the company's internal data, Certara's customers received 90%+ of all novel FDA drug approvals 2014–2025; its software is licensed to 20 global drug regulatory agencies (FDA, EMA, Japan's PMDA, China's NMPA/CDE, UK's MHRA), used by 160,000+ users, and has served 38 of the top 40 biopharma companies by R&D spend and 2,600+ companies/institutions across 70 countries.
Contract structure: software terms are typically 1–3 years; the recurring base is measured by net software retention. No single customer >10% of revenue; top-10 = 24% of FY2025 revenue (27% FY2024) — moderate concentration, improving. Q4 is the seasonal high (renewal-cycle timing). Customers are R&D-budget-dependent — the demand driver is biopharma R&D spend (~$290B/yr industry-wide), which makes Certara a levered play on the health of the biotech funding cycle.
Corporate: Formed 2008 (Tripos + Pharsight), incorporated Delaware 2017, HQ Radnor PA, IPO'd Dec 2020 (NASDAQ: CERT). Auditor RSM US LLP (since 2022). 1,576 employees in 28 countries, 414 with PhD/MD.
Certara is a software+services company, so the "supply chain" is talent, data, and cloud — not physical inputs:
Chokepoints / single-source dependencies: (1) AWS — sole named cloud host; any disruption hits the cloud franchise. (2) PhD labor market — industry has "relatively high employee turnover"; loss of key scientists is a named risk. (3) FDA/regulatory throughput — the 10-K devotes unusual space to FY2025 FDA workforce cuts (Executive Order 14210, ~3,500 FDA positions cut, HHS RIF of ~20,000) as a demand risk: if the FDA slows reviews, customer programs slow, and demand for Certara softens. This is a genuine, currently-live macro chokepoint, not boilerplate.
This lens is concrete: the names are AWS (infra), the FDA/EMA/PMDA/NMPA/MHRA (regulator-users), and the top-40 pharma (buyers).
The moat is real and unusually durable for a sub-$2B company — it is regulatory entrenchment, not just software switching costs:
Bargaining power: strong over small biotech (they can't build this in-house), weaker over top-20 pharma procurement (the 10-K repeatedly flags "pricing pressure… from the customer's procurement department" as total spend rises). Net: Certara needs the top-40 more on any single account than they need Certara, but the aggregate ecosystem lock-in tilts power back to Certara.
The honest moat caveat (developed in Lens 12/13): the moat is strongest in mechanistic/regulatory software and weakest in services and in the new AI layer, where general-purpose LLM providers and well-funded entrants could compress the regulatory-writing and analysis value. Management's "expert-in-the-loop, AI-inside-validated-frameworks" framing is partly a moat narrative and partly a defense against exactly this disruption.
Certara reports a single operating segment with three goodwill reporting units (Certara Data Science Software / Certara Predictive Technologies / Certara Drug Development Services). It does not disclose a clean product-segment EBITDA split or a US/international revenue split — only the software/services revenue cut and a foreign-currency proxy. (segments.csv on disk is empty.)
By product line (revenue, $000):
| Line | FY2025 | FY2024 | YoY | FY2025 mix |
|---|---|---|---|---|
| Software | 183,275 | 155,696 | +18% | 44% |
| Services | 235,563 | 229,452 | +3% | 56% |
| Total | 418,838 | 385,148 | +9% | 100% |
Multi-year total revenue: FY2023 $354.3M → FY2024 $385.1M (+8.7%) → FY2025 $418.8M (+8.7%). Organic FY2025 +6%; acquisitions (Chemaxon) +3%.
The trend that matters is the Q1-2026 reversal:
| Line | Q1-2026 | Q1-2025 | YoY |
|---|---|---|---|
| Software | 49,726 | 46,369 | +7% |
| Services | 57,189 | 59,635 | −4% |
| Total | 106,915 | 106,004 | +1% |
So the story decelerated hard at the top: software growth halved (from +18% FY to +7% in Q1), and services turned negative (the 10-Q attributes the −4% to "lower revenue from a previously acquired business" — i.e., the regulatory-writing unit being run down ahead of divestiture). Mix is shifting toward the higher-margin software, but the absolute growth engine stalled.
Geography: no clean split disclosed; 31% of FY2025 revenue was transacted in foreign currency (28% FY2024) — predominantly GBP, EUR, JPY — the only geographic granularity the filing gives.
This is the print that broke the stock (−17% on the day ). The latest filing on disk is the Q1-2026 10-Q (period ended 2026-03-31, filed 2026-05-11).
The numbers:
Beat/miss: MISS. Adjusted EPS $0.09 vs consensus ~$0.11.
What drove it: (1) Services decline (regulatory-writing run-off). (2) Cost growth ahead of revenue — G&A jumped to $29.4M from $19.7M (+49%), R&D +17% ($12.3M vs $10.5M), driving the operating loss. (3) A $7.2M change in fair value of contingent consideration (non-cash, vs −$0.2M prior year) — this is the bulk of the optical GAAP swing and is added back in adjusted EBITDA. But note: even adjusted EBITDA fell 9% and adjusted EPS fell 36%, so this was a genuinely soft operating quarter, not merely a non-cash optic.
The forward tell — bookings: Q1-2026 bookings $115.3M vs $118.2M, −2% YoY. Bookings are the leading indicator; a YoY decline is the single most concerning data point in the quarter. Net software retention improved to 106.1% (vs 102.4%) — the installed base is fine; it's new business / services that softened.
Guidance / outlook (set at this print, post-divestiture): FY2026 revenue $395–405M (includes ~$18M of stub regulatory-writing revenue; implies ~0–4% growth ex-divestiture), adj. EBITDA margin 30–32%, and adj. diluted EPS $0.35–$0.41 — a guide-down from FY2025's $0.44. (Note a sourcing conflict: one feed cited $0.44–$0.48; the $0.35–$0.41 range is the post-Veristat-divestiture figure carried in the May 11 release and is treated as current — flagged, not silently reconciled.)
Balance-sheet flags (FY2025 year-end): cash $189.4M; total debt ~$298.5M gross ($295.5M term loan); net debt ≈ $104M; deferred revenue $77.8M. Healthy liquidity, modest leverage (~0.8x net-debt/adj.EBITDA).
Market reaction: −17% on the print; the stock hit an all-time low of $4.45 on 2026-05-15, four days after earnings. What the market priced: a growth company that stopped growing, a new CEO with no track record yet, and a guide-down — re-rated from "GARP compounder" toward "show-me / sum-of-parts."
transcripts/ on disk is empty — this lens is ``.
Tone shift over the last several quarters:
Recurring phrases now: "expert-in-the-loop," "MIDD platform," "revenue predictability," "operational excellence," "defined competitive advantage." What they stopped saying: the unqualified "strong demand across software AND services" line — services is now spoken of as a focus/realignment area, not a growth driver. The sentiment arc is confident-compounder → focused-turnaround, which is exactly what a new CEO inheriting a stalled quarter would signal.
Pure-play biosimulation comps are scarce — the cleanest is Simulations Plus, and it is being taken private, which is itself the most important comp datapoint. Multiples are `` with date, or n/a.
| Company | Ticker | Mkt cap (USD) | EV/Sales | EV/EBITDA | P/E | Div yield | 5-yr avg ROE | Note |
|---|---|---|---|---|---|---|---|---|
| Certara | CERT | ~$1.48–1.59B | ~3.8–4.0x | ~14.3x 2026E / 18.5x ttm | n/m (GAAP loss) | 0% | negative (acc. deficit) | biosimulation leader |
| Simulations Plus | SLP | ~$0.33B | ~2.5x | ~18.0x | ~20x | ~0% | n/a | being acquired by Altaris at $18.50/sh, announced Jun 2026 |
| Dassault Systèmes | DSY.PA | n/a | n/a | n/a | n/a | n/a | n/a | named competitor (BIOVIA) |
| Ansys / Mathworks / ICON(NONMEM) | — | n/a | n/a | n/a | n/a | n/a | n/a | named competitors, not pure-play |
for the named competitor set (Mathworks, Dassault Systèmes, Ansys, Simulations Plus, NONMEM/ICON; services peers Metrum Research, qPharmetra, Pharmetheus).
The read-through that matters: SLP — the closest listed pure-play — is being taken private at ~2.5x sales / a premium multiple. Bulls (Leerink, 2026-05-15) argue this "supports a higher multiple" for CERT and establishes a credible M&A / take-private floor, especially with Arsenal Capital already owning ~22% (Lens 9). CERT trades richer than SLP on EV/Sales (~3.8x vs 2.5x) but that reflects CERT's larger scale and software mix; on 2026E EV/EBITDA the ~14x is below CERT's own history and below where a strategic/PE buyer paid for SLP.
Mostly ``; the pattern is unusually clean.
What the market actually reacts to for this name: (1) bookings + software growth rate (the GARP thesis lives or dies on re-acceleration), (2) biotech-funding / FDA-throughput macro (it's a levered R&D-spend play), and (3) M&A / sponsor signals (Arsenal's stake and the SLP deal). It does not trade on quarterly GAAP EPS (chronically near zero from amortization) — it trades on the growth narrative and the strategic/ownership setup.
The single most important governance fact: a brand-new CEO and a PE sponsor with effective control.
Forensic lens — every figure labeled. Findings are hygiene-level, not fraud-level.
Regulatory findings (required sub-section):
Built bottom-up from FY2025 actuals + FY2026 guidance. Output ``; inputs labeled.
Anchors: FY2025 adj. diluted EPS $0.44; FY2026 guide $0.35–$0.41, revenue $395–405M, adj. EBITDA margin 30–32%. The FY2026 decline is mechanical — the Veristat divestiture removes ~$50M revenue / $17M adj. EBITDA (a 34% margin unit), so EPS steps down even though the remaining business is healthier-mix.
The number that actually matters is not EPS — it's the software growth rate. GAAP EPS will hover near zero on amortization regardless; the equity re-rates on whether software re-accelerates from +7% back toward the mid-teens. Watch bookings (turned −2%) as the leading tell.
No forecast.ts create run (watchlist/breadth mode — forecasts are logged only on committed base cases, per skill).
Bull case. Certara owns the regulator-embedded standard for computational drug development — a moat (90%+ of FDA approvals touch it; 20 regulators license it; 20-yr validated Simcyp data) that does not erode in a quarter. The 44% software mix is high-margin, 105–106% net-retention sticky, and grew 18% in FY2025. The new CEO is doing the right things — shedding the low-multiple regulatory-writing drag (Veristat), buying back stock at a trough, and pushing Certara.AI as a platform up-sell. The stock printed an all-time low at $4.45 on a single soft quarter and ~14x 2026E EV/EBITDA — and the nearest pure-play comp (SLP) was just taken private at a premium, with Arsenal already owning 22%. Earnings surprise vector: any bookings re-acceleration or a take-private bid.
Bear case (2–3 permanent-impairment risks). (1) The moat is in mechanistic software; the growth was in services + acquisitions — and both just rolled over (services −4%, bookings −2%, software decel to +7%). If MIDD adoption is maturing and biotech R&D stays funding-constrained, Certara is a ~mid-single-digit grower wearing a software multiple. (2) AI cuts both ways — the regulatory-writing/analysis value (the very unit just sold) is exactly where general-purpose LLMs compress price; Certara.AI is "still in development" and unproven as a revenue engine. (3) Roll-up air — $1.22B goodwill+intangibles (78% of assets) with one impairment already; a sustained de-rating forces more. Multiple too high? vs its own stalled growth, ~3.8x EV/Sales / 14x EBITDA is not cheap if this is a 0–4% grower; it's cheap only if software re-accelerates.
Pre-mortem (18 months out, thesis broke): biotech funding stayed tight + FDA throughput stayed disrupted; software never re-accelerated past high-single-digits; Certara.AI slipped; bookings declined two more quarters; the new CEO's "focus" turned out to mean "shrink to a sale that didn't materialize at a good price"; another goodwill impairment; the stock re-based to ~$5–6 as a no-growth ~12x-EBITDA services-y asset.
Contrarian view (what the market refuses to see): the market is treating CERT as a broken SaaS name, but it's really a regulatory-infrastructure monopoly with a temporarily-stalled growth optic and a motivated 22% PE owner. The most underpriced scenario is not operational re-acceleration — it's a take-private, where the moat + recurring software + the SLP precedent + Arsenal's stake make a bid at a meaningful premium to $9 plausible. The market is pricing the bad quarter; it is under-pricing the ownership setup.
Dismantling the bull case.
A fortress-margin vertical-SaaS monopoly trading at a growth-stock funeral price (~20x forward EPS, near 52-wk lows) because the market is pricing a Salesforce-Agentforce CRM war that threatens the contested ~40% (Commercial) while ignoring the defensible, faster-growing ~60% (R&D/Quality); BULLISH at $153 on a 1–3Y view, but the CRM-migration-to-2030 is a real, watchable execution overhang — not a phantom.
A real, fast-growing oncology-data + diagnostics franchise wrapped in an "AI" narrative it can't yet monetize — own the genomics flywheel, but the round-trip-flavored deals, 30-vote founder, and a CEO famous for cashing out cap the multiple until cash flow turns.
A profitable biosimulation pure-play that the public market broke (Pro-ficiency wrote off ~$72M of a $100M deal, growth fell to single digits) — and is now exiting via a $18.50/sh all-cash Altaris take-private. The fundamental thesis is moot; the only live question is deal-closes-vs-breaks, and this closes (board-unanimous, 16% founder block locked, Q4-2026 close). Merger-arb, not a growth bet.