PBYI — Puma Biotechnology, Inc. · 2026-09-10 · Verdict: PASS · Conviction 3
Price $9.05 (2026-09-10 screen row close; no live price used) · Mkt cap $467M · EV $373M · EV/normalised EBIT 12.6x · FCF yield 8.8% · Net cash $93.9M · ADV $2.8M Sources read: 10-K 2026-02-26 (Items 1, 1A, 7), 10-Q 2026-08-06, DEF 14A 2026-04-28, 8-Ks 2026-05-07, 2026-06-16, 2026-08-06, Form 4s (12 filings, 19 rows). No transcript was filed; the Q2 press release is the closest SEC substitute.
Desk stats - Revenue trend. FY2025 revenue -0.9% to $228.4M: product +4.6% to $204.1M on 5.5% more bottles plus price, royalty -31% to $24.3M on lower China sub-licensee sales (10-K, Item 7). Q2 2026 +7.8% to $56.5M: product +8.9% on 12.3% more domestic bottles and higher price, partly given back in Medicaid deductions; royalty -9% (10-Q, Item 2). - Normalised after-tax operating profit. LTM to 30 June 2026 GAAP operating income $30.5M (FY2025 $37.3M, less H1 2025 $10.6M, plus H1 2026 $3.8M). One adjustment: deduct the $0.9M reversal of a previously recorded AstraZeneca legal accrual credited to SG&A in Q2 2026 (10-Q, Item 2). No impairments, discontinued operations, restructuring or one-time tax items. Normalised EBIT $29.6M, taxed at 25% = $22.2M (GAAP equivalent $22.9M). - EV / normalised after-tax profit: 16.8x. EV $373.3M = 51.62M shares at $9.05 less $36.5M cash and $57.4M marketable securities, zero debt since the Athyrium note was repaid in full on 4 May 2026 (8-K 2026-08-06). On management's own 2026 guide the same EV is about 28x. - Leverage. Net cash $93.9M, 2.3x normalised EBITDA of $40.0M (EBIT plus $10.4M LTM D&A). No debt; $8.1M of lease obligations. - Growth sustainable? Organic and cash-backed at the revenue line, not the profit line: 2026 revenue is guided flat at the midpoint ($227.5M vs $228.4M) while guided net income of $17-20M is 40% below 2025's $31.1M, because R&D rises from $62.1M toward roughly $88M on alisertib. - What the screen got wrong. It counted only $36.5M of cash and missed $57.4M of securities, overstating EV by $57M in the company's favour; and it charged CY2025's $37.3M of operating income against today's price when the guide implies about $17M for 2026.
1. What the business actually does
Puma sells one drug. NERLYNX (neratinib) is an oral kinase inhibitor approved in the US for extended adjuvant treatment of early-stage HER2-positive breast cancer after trastuzumab, and with capecitabine in third-line-plus metastatic disease (10-K, Item 1). It is sold domestically through about 38 sales specialists to a handful of specialty pharmacies and abroad through royalty-paying sub-licensees. The second asset, alisertib, an Aurora Kinase A inhibitor, was in-licensed from Takeda in 2022 for $7.0M up front plus up to $287.3M of milestones and sits in two Phase II trials. Neratinib is itself licensed from Pfizer at a low-to-mid-teens royalty.
2. Why it is mispriced — the edge case
There is none, which caps the verdict. Nothing forces anyone to sell: the stock is up 49.3% in six months and 90.6% over twelve, and sits 6.6% below its 52-week high on $2.8M of daily volume. The cheapness is a stale-period artifact, not a dislocation, and the market found the story first: 2026 net income guidance went $10-13M in February, $16-19M in May, $17-20M in August.
3. Unit economics and growth
The franchise is good. On the 2026 guide, about $227.5M of revenue carries roughly $50M of cost of sales and $72M of SG&A, so NERLYNX earns near $105M before research. Cash taxes are near zero: the deferred tax valuation allowance was $324.8M at 31 December 2025 (10-K, Item 1A) and current tax expense was $0.4M in H1 2026.
Three things consume it. R&D of roughly $88M in 2026, against $62.1M in 2025 and $54.9M in 2024, almost all alisertib, takes about 84% of pre-research profit. Gross-to-net erosion is steady: reserves for variable consideration were 19.5% of product revenue in 2024, 24.3% in 2025, guided 26.5-27.5% for 2026, which is why Q2's 12.3% bottle growth became 8.9% of revenue. Royalties are in structural decline, $35.3M to $24.3M to a guided $19-22M, and the 10-Q says the China royalty rate faces a contractual reduction on a generic-share threshold management believes "could be reached... in late 2026 or in 2027."
The clock matters more. The US composition-of-matter patent, after a five-year Hatch-Waxman extension, expires 29 December 2030; polymorph patents expire 2028 (10-K, Item 1). This is a wasting asset with about 4.3 years of protected life, and management's own competition disclosure names DESTINY-Breast05, where trastuzumab deruxtecan reduced recurrence or death in high-risk early-stage HER2-positive breast cancer, plus CompassHER2 RD for tucatinib, both aimed at the setting that produces most of NERLYNX's US revenue.
Alisertib is early. ALISCA-Breast1's stated primary objective is "to determine the optimal alisertib dose level"; ALISCA-Lung1 enrols up to 60 patients, after which management "anticipate[s] meeting with the FDA to explore the potential for an accelerated approval pathway." The striking numbers in the file, hazard ratios of 0.29 and 0.395, are retrospective biomarker subgroups of Takeda-era trials whose intent-to-treat endpoints missed.
4. Balance sheet and capital allocation
$93.9M of cash and securities, zero debt, equity $138.5M. No dividend, no buyback authorisation; every dollar of franchise cash goes into alisertib, and basic weighted-average shares rose 2.5% year over year. Alan Auerbach beneficially owns 19.6%, including 7,046,751 shares outright and a warrant over 2,116,250 shares at $16.00 expiring 4 October 2026, deeply out of the money (DEF 14A). Insiders sold 141,414 shares for $1,084,258 over twelve months with zero open-market buys: directors at about $7.20 on 12 June 2026, Auerbach, the CFO and Mr Hunt at $8.263 on 6 July 2026 (Form 4s).
5. Management: what they said vs what they did
Guidance has been raised twice and met so far; Auerbach said in May the aim was "carefully managing our resources to achieve positive net income for this year" (8-K 2026-05-07) and H1 delivered $4.4M. But the year rests on Q4: after Q3 guidance of $2-3.5M, the full-year $17-20M implies roughly $11M in the final quarter, consistent with 2025's seasonality (H1 $8.8M, H2 $22.3M) but unproven. Shareholders are less indulgent: on 11 June 2026 they voted down the Auerbach warrant extension, 18,234,150 for against 19,596,238 against, and say-on-pay support fell to about 79% of votes cast from roughly 95% a year earlier (8-K 2026-06-16; DEF 14A).
6. Valuation
Base (50%). Product revenue holds near $205-210M to 2028 then fades into the cliff; royalty decays to about $12M; R&D runs $90M two more years then falls to $40M absent a registrational path; cash taxes near zero on the NOLs. Discounted at 10% the franchise is worth about $155M, plus $93.9M net cash: $4.85. Bear (30%). US demand rolls over as it did in six of the eight years before 2025, gross-to-net passes 28%, Enhertu takes adjuvant share, the China step-down lands. Revenue $180M by 2028 against $80M of R&D leaves break-even: $2.75. Bull (20%). Q4 2026 ALISCA-Breast1 data selects a dose with a clear biomarker effect, the FDA agrees an accelerated path in small cell lung cancer, and alisertib is partnered or re-rated: $17. Probability-weighted $6.66, about 26% below price. Reverse DCF. At $9.05 the $373M enterprise value, against roughly $100M of cumulative after-tax owner earnings NERLYNX can produce before 29 December 2030 at the current research run-rate, puts about three-quarters of enterprise value, some $4.25 a share, on a Phase II asset whose current trial exists to choose a dose.
7. Catalysts and timeline
4 October 2026: the $16.00 warrant expires, removing 2.1M shares of overhang. Early November: Q3 tests the guide and the implied $11M Q4. Q4 2026: updated ALISCA-Breast1 data, the only event that can validate most of the enterprise value. Late 2026 or 2027: the China royalty step-down. February 2027: first FY2027 guidance shows whether research spending rises again for a Phase III.
8. Pre-registered kill criteria
- FY2026 net income lands within or above the $17-20M guide and initial FY2027 guidance puts R&D below $90M.
- The board authorises and executes at least $50M of buyback (there has never been one).
- Q4 2026 ALISCA-Breast1 data shows a confirmed objective response rate above 30% in a pre-specified biomarker subgroup at a selected dose, with a registrational path discussed with the FDA.
- Confirming the pass: domestic bottle growth negative for two consecutive quarters, gross-to-net above 28%, or FY2027 royalty guidance below $15M.
9. Verdict and one-paragraph summary
PASS, conviction 3. Puma is a debt-free single-product company with $93.9M of net cash whose franchise earns about $105M a year before research, and the screen is wrong in both directions: it missed $57.4M of marketable securities, which helps, and it charged CY2025's $37.3M of operating income against today's price when the twice-raised 2026 guide implies about $17M, which hurts far more, making the real multiple roughly 28x guided after-tax operating profit rather than 15x. The one reason to pass is that the enterprise value is not really a claim on that franchise: the US patent expires 29 December 2030, management's own filing names two competitor trials aimed at the setting that produces most of the revenue, royalties are falling with a China step-down flagged for late 2026 or 2027, and roughly 84% of pre-research profit is spent on alisertib, a Phase II asset whose current trial exists to pick a dose and whose best numbers are retrospective subgroups of trials that missed. At $9.05 you pay three-quarters of enterprise value for that bet, with no forced seller, the stock 6.6% off its high after a 49% run, zero insider buying against $1.08M of sales, a rejected CEO warrant extension and say-on-pay support down from 95% to 79%. Probability-weighted value of $6.66 is 26% below price.
Research for discussion, not investment advice. Positions and sizing are the reader's decision.