Cogstate (ASX: CGS) — Investment Research Report

3-Year / 5-Year Price Scenarios

Base EPS: $0.09 — TTM EPS (AUD); company reports EPS in USD (~$0.06 FY25 statutory)

0. Information Richness & AI Limitations

Grade B (moderate). A ~A$430M micro-cap healthcare-tech stock with thin broker coverage but disciplined ASX disclosure. Grade-B trap: AI tends to 'reasonably infer' its way across lumpy contract-revenue gaps, producing false certainty. Two must-know caveats: (1) Dual-currency — Cogstate reports in USD but trades in AUD on the ASX (AUD/USD ~0.65); never compare growth/valuation without unifying the currency. (2) Lumpy revenue — clinical-trial revenue is recognized on contract milestones and swings sharply half-to-half; do not annualize a single period. (3) Customer concentration — revenue depends on a few large biopharma sponsors running Alzheimer's trials.

1. Data & Cross-Validation

Price A$2.53; 169.88M shares; market cap A$429.8M (verified 0.00%); PE 27.5x, forward 21x; EPS TTM A$0.09; dividend yield 0.79%; 52-week A$1.55–3.01; analyst target ~A$3.00 (thin coverage).

5-year trend (USD millions): Revenue 32.7 → 45.0 → 40.5 → 43.4 → 53.1 (FY21–FY25, ~12.9% CAGR but non-linear); net income 5.2 → 7.5 → 3.6 → 5.5 → 10.1; operating margin 11.8% → 24.3% → 5.7% → 14.2% → 24.8%. FY2023 was the trough (trial gaps + front-loaded cost); FY24–25 rebounded strongly.

FY2025 (company, USD): Group revenue US$53.1M (+22%); clinical-trial new-contract sales US$41.3M (+53%); Healthcare segment −37%; net profit US$10.1M (+86%); NPBT US$13.9M (+96%); EBITDA US$16.0M (margin 30.1%, +8.7pt); gross margin 61%; contracted future revenue US$90.0M (−11.4%). Watch the divergence: new-contract sales +53% vs contracted future revenue −11.4% — a timing mismatch worth tracking.

2. Business Essence — Duan Yongping

One line: Cogstate is the 'pick-and-shovel' seller for Alzheimer's / CNS drug R&D — it doesn't make drugs; it provides scientifically validated digital cognitive-assessment tools and services for pharma clinical trials, billed per contract.

Model: between SaaS and professional services. The core asset is regulator-validated cognitive test batteries (e.g., Cogstate Brief Battery) and a data platform. Once written into a trial protocol, switching cost is very high (changing the tool invalidates data comparability).

Revenue mix: (1) Clinical Trials (growth engine, ~85%+ of revenue) — sold to biopharma sponsors; (2) Healthcare / Direct-to-Consumer (Cognigram etc., currently shrinking) — a potential second curve not yet delivered.

Economics: 61% gross margin, capital-light, healthy cash flow, has begun paying a (small, 0.79%) dividend. Operating leverage is real — FY25 net margin jumped from ~12.6% to ~19.1% as revenue grew.

Duan's verdict: a good business — high-margin toll collector with strong switching costs — but lumpy revenue, concentrated customers, and a ceiling set by total CNS-trial volume.

3. Moat — Buffett

Switching costs ★★★★☆ (a test written into a protocol can't be swapped mid-trial without breaking data comparability); brand/scientific credibility ★★★★☆ (30 years of peer-reviewed validation, regulatory acceptance); scale/data network ★★★☆☆ (one of the largest standardized cognitive databases — more trials → stronger standard, but not a pure network effect); pricing power ★★★☆☆ (some premium, but large sponsors bargain hard); patent/tech ★★★☆☆ (IP exists but the moat is more 'regulatory validation + trust' than raw tech).

Trend: As Alzheimer's drugs (Leqembi/lecanemab, Kisunla/donanemab) win approval, more pharma enter CNS → more demand for validated cognitive endpoints → Cogstate benefits; the moat widened over the past 5 years. Buffett question: likely still there in 10 years if Alzheimer's R&D persists. Threats: (1) industry downturn; (2) AI-native cognitive assessment bypassing traditional batteries; (3) large sponsors building in-house.

4. Reverse Thinking & Risks — Munger

Historical analogy: a 'CRO/pick-and-shovel' logic (like early ICON, Medidata) — great compounders when the field is right, but violently volatile at small-cap stage; one big client or one therapy-cycle reversal hits hard. Munger question / where I'd be wrong: anchoring on FY25's pretty '+86% net profit' and mistaking a one-off operating-leverage release for a sustainable norm, then chasing a high PE. Why smart investors don't buy: too small (hard to build a position), poor revenue predictability, not cheap.

5. Management — Duan Yongping + Buffett

CEO: Brad O'Connor (former CFO turned CEO) — steady, financially disciplined. Key decision: in April 2024, amended and partly reclaimed the Eisai global license — taking back the cognitive-test commercialization rights previously ceded to Eisai, regaining control of Healthcare/Cognigram monetization. A 'reclaim moat control' capital-allocation move reflecting management's view of its own IP value.

Capital allocation: capital-light, initiated a dividend, no aggressive M&A — consistent with 'right people don't waste money.' Risk: execution is strong, but the Healthcare second curve has yet to deliver, suggesting consumer/clinic-side commercialization was underestimated.

Duan question: if the CEO retired, would it stay competitive? — Yes, likely: the core assets are the battery IP and client relationships, not one individual, though key-person risk remains for a small company.

6. Industry & Civilizational Trend — Li Lu

Civilizational-scale trend: global ageing + Alzheimer's moving from 'untreatable' into the disease-modifying-therapy (DMT) era (lecanemab, donanemab approved) — a clear long-cycle structural wave. Cogstate's position: it sits at the 'measurement layer' — whichever drug ultimately wins, objective cognitive assessment is a must-have. It doesn't bet on a single drug; it bets on the activity level of the entire field — the classic pick-and-shovel advantage. TAM: as more CNS indications (Alzheimer's, Huntington's, schizophrenia, etc.) enter trials, addressable trial volume expands; a future consumer 'early-screening/diagnostic' layer (Cognigram) could lift the ceiling, but the market is still small today. Li Lu question: more likely a 'niche hidden champion' — won't become a giant, but with a chance to be the long-term standard provider in its vertical if Alzheimer's therapies keep scaling.

7. Valuation & Scenarios — Buffett + Duan

PE(TTM) 27.5x / forward 21x. Reverse read: the market prices ~18–20% mid-term EPS growth. FY25's actual +86% net profit far exceeds this, but that was a low-base + operating-leverage one-off, not linearly extrapolatable. On EV/Revenue: market cap A$430M ≈ US$280M against US$53M revenue ≈ 5.3x P/S — not expensive for a 60%+ gross-margin, 20%+ growth software-type med-tech, but deserves a discount for lumpy revenue.

Three scenarios (base EPS A$0.09, tool-verified):

3-year (to ~2029): Bull 30% growth / 32x → A$6.30 (+150%, 25%); Base 18% / 23x → A$3.40 (+34%, 50%); Bear 2% / 14x → A$1.30 (−47%, 25%). Prob-weighted ≈ A$3.60 (+42%). 5-year (to ~2031): Bull 26% / 30x → A$8.60 (+239%, 25%); Base 16% / 22x → A$4.20 (+64%, 50%); Bear 2% / 12x → A$1.20 (−53%, 25%). Prob-weighted ≈ A$4.55 (+80%, ~12.4%/yr).

Target prices are capital-appreciation only; the 0.79% dividend adds negligibly. Interpretation: positive expected value (3yr +42%, 5yr +80%) but very high variance — the bear case halves the stock over 5 years. A classic 'positive-EV, high-volatility' small-cap growth stock: suits a small position, staged entry, long hold — not a heavy or short-term bet.

Duan question ('hold 5 years if the market closed?'): yes for a small position (business is in the right field), but not overweight (lumpy revenue + not-cheap valuation + micro-cap volatility). A great track, good business — but not yet a great price. Margin of safety must come from buying cheap: below A$2.20 (forward PE <18x) offers enough cushion.

8. Decision Memo

Decision: No position: watch / small starter — enter in stages below A$2.20, don't chase. Hold: hold as long as clinical-trial new-contract sales keep growing and the thesis holds. Sell signals: new-contract sales fall two consecutive halves / Alzheimer's pipeline mass failure / PE >40x with growth unable to match. Add signals: pullback below A$2.20 with contracted revenue still growing / Healthcare-Cognigram turns positive.

One-line conclusion: a hidden champion at the measurement layer of the Alzheimer's drug wave — great niche, real moat, FY25 operating leverage delivered — but tiny-cap, lumpy, and not cheap. 3-year neutral A$3.40 (+34%); 5-year neutral A$4.20 (+64%); prob-weighted 5yr A$4.55 (~12.4%/yr) with wide dispersion. Small position, staged entry, long hold; not for the risk-averse or the position-concentrator.

Four-Master Commentary

Buffett: "I like a shovel-selling business — you don't have to guess which pharma wins, only that the field keeps mining. But the price has to be right; 27x demands it keeps sprinting, so I'd wait for a wider margin of safety."

Munger: "Invert it — the big mistake is treating one year's profit surge as the norm. For a company with revenue this bumpy, keep the position light and don't be fooled by pretty numbers."

Duan Yongping: "Right track, right business — but not the right price. Good companies still need good prices; I'd only get interested after a pullback. What's the hurry?"

Li Lu: "From no drug to a drug for Alzheimer's is a civilizational shift. A hidden champion at the measurement layer belongs on the watchlist — view it over 20 years, but bet with a small position."