ZNTL — Zentalis Pharmaceuticals

Company context · prepared 2026-08-12 · company sweep 2026-08-12 · USD · framework v5.6.1

How to read this. Written for a reader who has not studied pharmacology or finance. General terms are defined once in framework/04-glossary.md. Tags: [VERIFIED — source] / [UNVERIFIED] / [WEB ESTIMATE — source, date].

Program analyses for this company are the sub-folders listed in C.2.

C.0 Company-tier coverage — CLEARED

C.1 What the company is

Zentalis Pharmaceuticals is a clinical-stage cancer-drug company based in San Diego. It has no product revenue and has never sold a medicine [VERIFIED — BPIQ fetch_company_info 2026-08-12].

It is a single-asset company in everything but name. The data feed returns twelve program rows and eight of them are the same molecule, azenosertib (also called ZN-c3), an oral once-daily tablet that blocks a protein called WEE1. Five of the twelve rows are marked failed and discontinued [VERIFIED — BPIQ fetch_company_drugs 2026-08-12]. Exactly one row carries a disclosed catalyst: the Phase 2 DENALI readout of azenosertib in Cyclin E1-positive platinum-resistant ovarian cancer.

That one catalyst moved since the last version of this document. On 2026-08-06, with its second-quarter results, the company moved the DENALI Part 2 topline from “year-end 2026” — guidance it had repeated unchanged seven times over fifteen months — to “1H 2027”, “to allow for data maturation post full enrollment”. The same release disclosed an FDA Type D meeting on the accelerated-approval strategy [VERIFIED — Q2 2026 results release, 2026-08-06]. The shares fell roughly 16% over the following three sessions (see C.7).

Azenosertib is licensed in rather than owned outright. Zentalis’s subsidiary Zeno Pharmaceuticals licensed the underlying intellectual property from Recurium IP Holdings in December 2014, and owes Recurium milestone payments and a royalty on any sales [VERIFIED — Zentalis Form 10-K, Recurium Agreement description; see the program document B.3b].

The company is building a commercial organisation ahead of a result it does not yet have. In May 2026 it appointed Sarah Kelly as Senior Vice President of Commercial Strategy and added Shannon Campbell to the board [VERIFIED — company release 2026-05-27]. In January 2025 it had cut about 40% of its staff to fund this same programme [VERIFIED — Fierce Biotech 2025-01-29].

The share price this document is written against is $4.16, read on 2026-08-12. Every figure below is measured against that.

C.2 Pipeline — every program

Twelve rows are listed because twelve rows were returned. Five are discontinued programmes. That is the point of listing them: this company has shut down its entire non-WEE1 pipeline — an oral oestrogen-receptor degrader (ZN-c5), a BCL-2 blocker (ZN-d5), an EGFR blocker (ZN-e4) and a Pfizer-partnered colorectal combination — and one live catalyst therefore carries essentially all of the equity.

Drug names in this feed are dirty — azenosertib appears with and without a trailing space and with inconsistent suffixes. Programs are keyed on the integer bpiq_drug_id throughout, never on the name.

C.3 Financial position

The runway-versus-catalyst picture deteriorated, and it is the most important change on this page. The previous version of this document said the company was funded roughly eleven months past a year-end 2026 readout. The readout has since moved to H1 2027 and a quarter of cash has been spent. Cash now covers the guided window — into late 2027 against a readout no later than 2027-06-30 — but the margin past the readout has shrunk to a few months, and completing a 420-patient Phase 3 and preparing a launch cost more than what remains. Projected cash at the likeliest readout (2027-03-31, on the reported burn) is about $62.9M, and at the window’s latest edge (2027-06-30) about $25.7M [UNVERIFIED — modelled; flat reported burn, no financing]. A raise or partnership before or at the readout is now a live possibility rather than a remote one.

Dilution history.

Is a raise likely before the next catalyst? More likely than before, and no longer easy to rule out. The previous version answered “no” on the strength of cash covering the readout roughly twice over. That arithmetic is gone: the readout moved six months further out while a quarter of cash burned. The company can still reach the readout without raising, and the readout remains the obvious moment to raise or partner on better terms, but an S-3 shelf is on file and the incentive to de-risk the balance sheet into a binary event has grown. [UNVERIFIED — judgement, built on the verified cash, burn and filing facts above]

The December 2025 block, read carefully. Matrix Capital Management, a 10% owner, disposed of 7,500,000 shares at $1.33 on 2025-12-15, sixteen days before Walters Group acquired 6,459,973 shares at $1.20. Matrix had bought 4,760,000 shares at $22.66 in June 2023. The most straightforward reading is that one large holder capitulated near the bottom and another took the block off its hands, which would make the Walters purchase a change of ownership rather than dilution. That reading is consistent with the sizes, the prices and the dates, but it was not confirmed in this sweep — no 13D/G or offering document was read. [VERIFIED — the three transactions; UNVERIFIED — the connection between them]

C.4 Valuation frame

The shares gave back a fifth of their run in one week, on the timeline and not on the science. The price cache’s last committed bar is 2026-08-05, the day before the second-quarter release, closing at $4.95. The connector’s 2026-08-12 quote is $4.16 with a previous close of $4.11, so the stock fell roughly 16% across the three sessions after the topline moved from year-end 2026 to H1 2027 [VERIFIED — data/prices/ZNTL.json last bar; BPIQ quote 2026-08-12]. Before that drop the shares had roughly quadrupled from the December 2025 low, most of it after the April 2026 dose selection.

At $123.6M of recomputed enterprise value the market is paying roughly a quarter of a single year of the base-case peak sales modelled in the program document. That is not a price that assumes the readout works, but it remains well above the price that assumed it would not.

C.5 Ownership and flow

Insider flow: no conviction buying anywhere near the current price. Across 250 transactions back to January 2022, only twelve are acquisitions carrying a real purchase price, and the most recent open-market insider buys are $1.40 (director Scott Myers, 2025-04-30) and $2.28 (Chief Medical Officer Ingmar Bruns, 2025-02-06) against a $4.16 spot. The cluster is unchanged: five directors and officers bought between $1.40 and $2.28 in the January–April 2025 window, near the bottom, and nobody has bought since. Everything in 2026 is either a grant — 1.34M options awarded 2026-01-08, 57,100 restricted units to each of five directors on 2026-06-16 — or a small disposal to cover tax on vesting, at $2.39 to $2.55 in February 2026. This corrects the previous version of this document, which dated the Bruns purchase to 2026-02-06; the feed dates it 2025-02-06, and the February 2026 rows are tax-withholding disposals. [VERIFIED — BPIQ fetch_company_insider_transactions, 250 rows read in full]

Concentration, from the 13F feed — with the filing-quarter labels corrected. A 13F is filed up to 45 days after the quarter it reports, and this connector labels each block by filing quarter, so every block’s holdings are one quarter older than its label says (02-connectors.md § Field-level bugs). Confirmed on ZNTL exactly: the “2026Q2” block’s value-per-share is $2.34, the 2026-03-31 close; “2026Q1” gives $1.35, the 2025-12-31 close; “2025Q4” gives $1.51, the 2025-09-30 close [VERIFIED — value ÷ shares matched against data/prices/ZNTL.json quarter-end closes].

Read correctly: five funds held 7,097,632 shares at 2026-03-31, about 9.9% of the filed count — 5AM Venture Management (3,947,913), Opaleye (1,412,656), Decheng (1,323,327), Stonepine (269,100) and ADAR1 (144,636). The repositioning happened in the first quarter of 2026: between 2025-12-31 and 2026-03-31 the group went from 4,183,677 to 7,097,632 shares, driven by 5AM Venture taking 490,000 to 3,947,913 and Opaleye adding, while Tang Capital’s 1,550,000-share exit shows one quarter earlier, in Q4 2025. So the dedicated biotech money positioned itself before the April dose selection — that conclusion survives the label correction — but there is no 13F visibility past 2026-03-31 yet: the previous version’s statement that the group “held flat in 2026Q2” was the label bug speaking, not a fact about the June quarter. Filings holding 2026-06-30 positions land around 2026-08-14. [VERIFIED — BPIQ fetch_company_hedge_fund_holdings, labels shifted back one quarter]

Two 13G filings corroborate the picture: Squadron reported a 3,676,900-share stake on 2026-03-03, and a new Schedule 13G was filed on 2026-07-30 whose holder this sweep did not identify [VERIFIED — BPIQ press feed; EDGAR submissions; the 13G contents were not read].

No dominant specialist fund anchors this stock. The largest disclosed 13F position is 5.5% of the shares. Ownership is spread across a large institutional base and a meaningful short position, which is the setup that produces violent two-sided moves on news.

C.6 Options chain and its readability

Plain takeaway: the chain cannot price this event, and it no longer even spans it. The last listed expiry is 2027-01-15, which sits at the very open of the guided H1 2027 readout window — a topline landing in the likeliest part of the window, the second quarter of 2027, is beyond every listed contract. On top of that, strikes are $2.50 apart on a $4.16 share, and the near-spot contracts barely trade.

Four contracts print the BPIQ implied-volatility floor artifact of exactly 0.01488 (deep in-the-money $0.50 and $1.00 calls across the October, December and January expiries) and are treated as null throughout. The ceiling artifact does not appear on this chain today.

C.7 Reaction to past catalysts

This stock moves 35% to 50% on news that matters, in both directions, and there is still no clean precedent for a pivotal efficacy topline. The two large clean up-moves are a dose selection (+46%) and a regulatory hold being lifted (+36%); the newest read is a −16% de-rating on a six-month timeline slip with no new science in it. The one event that resembles the coming readout — the January 2025 Part 1b topline — is irrecoverably confounded by a 40% layoff landing in the same release. C.7 therefore establishes the magnitude the market is capable of and constrains the direction of the coming move almost not at all.

C.8 Company data-quality flags

• max_52_week_position and market_cap both reproduce the documented previous-close bug. The connector reports 0.50505; substituting the previous close of $4.11 recovers exactly that figure, while the last price of $4.16 gives 0.5138. Market capitalisation of $294.6M divided by the previous close gives 71.67M shares, matching the EDGAR-filed 71,671,914 to 0.01%; dividing by the last price would imply a share count no filing supports. C.4 uses the filed count directly — since v5.6.0 that derivation is a sweep row, not a hand task.
• Enterprise value implies about $30M of debt that this sweep could not identify. Recomputed market capitalisation of $298.2M less cash of $174.6M is $123.6M, against the connector’s $153.5M. No debt facility appears in 24 months of press releases or the EDGAR form list. The gap is plausibly operating-lease liabilities or a stale figure. Flagged, not resolved. It was $40.7M at the previous sweep, so the unexplained residual also moves.
• The hedge-fund filing-quarter labelling bug (v5.6.1) is confirmed on ZNTL exactly. Every block’s value-per-share matches the prior quarter-end close in the committed price cache: “2026Q2” → $2.34 = 2026-03-31; “2026Q1” → $1.35 = 2025-12-31; “2025Q4” → $1.51 = 2025-09-30. All C.5 statements shift the labels back one quarter, and the previous version’s “flat in 2026Q2” claim is corrected — no 13F visibility past 2026-03-31 exists yet.
• The documented 1000× hedge-fund value bug is present again (Boxer Capital 2023Q4: $9,027,000,000 on 450,000 shares).
• Options implied volatility prints the 0.01488 floor artifact on four deep in-the-money calls; all treated as null. The 9.99512 ceiling artifact seen on 2026-08-04 does not appear today.
• The unexplained block of deep in-the-money call open interest shrank but persists. The 2027-01-15 $2.50 call shows 1,230 contracts open against 308 at $5.00, on no volume — down from 4,302 on 2026-08-04. Whether that was a close-out or a data correction is not resolvable from this feed. Flagged, not interpreted.
• short_float units remain unconfirmed (11.76% reported). The FINRA count over the EDGAR share count — 7.78% — is the figure C.5 uses, per the v5.6.0 sweep rows.
• The catalyst placeholder moved with the disclosure this time. BPIQ stores catalyst_date 2027-06-30 for row 17802, the documented period-end synthesis of the newly disclosed “H1 2027”. Unlike the 2026-08-04 flag — when the placeholder moved while the text did not — this is the text itself moving: the first real slip on this row.
• Insider-transaction dating corrected. The previous version placed the last open-market insider buy (Bruns, $2.28) in February 2026; the feed dates it 2025-02-06, and the February 2026 rows at $2.39–$2.55 are disposals. See C.5.
• Drug names in the pipeline feed are dirty, with trailing spaces and inconsistent suffixes on the same molecule. Programs are keyed on the integer bpiq_drug_id throughout.
• The committed price cache ends at 2026-08-05, the session before the slip was announced, so the −16% reaction is derived from the cache’s last bar against the connector’s 2026-08-12 quote rather than from daily bars. The cache is fetched on its own cadence outside this analysis (02-connectors.md § The gate); its next refresh will carry the reaction in full.