ASMB — Q2 2026 Financial Report Analysis

Overview

Assembly Biosciences' second-quarter net loss narrowed to $3.9 million ($0.20 per share) from $10.2 million ($1.33) a year earlier. The improvement is mostly accounting rather than operations. Collaboration revenue from Gilead rose 39% to $13.4 million, but $5.1 million of that was a one-time "cumulative catch-up" adjustment after the company revised its estimate of the work still left under the Gilead agreement. The per-share figure also fell partly because the share count is about 2.5 times larger than a year ago. The more important news for the company's future is the balance sheet: a $115.0 million stock offering in May lifted cash and investments to $320.4 million, which management says funds operations "into late 2028", and into 2029 if Gilead pays a $75.0 million extension fee due in the fourth quarter of 2026.

At a glance

• $320.4M cash and marketable securities at June 30, up from $248.1M at year-end. Against a six-month operating cash burn of $37.6M, that is the company's main protection against running out of money before its next trials report.
• $5.1M of the $13.4M revenue was a one-time catch-up. Without it, the quarter's net loss would have been about $9.0M, close to last year's $10.2M. The filing says the adjustment cut the per-share loss by $0.27.
• Two Phase 2 starts are guided for Q4 2026: Gilead's study of the genital-herpes drug 1179, and Assembly's own study of 6250 in hepatitis delta. A third (6250 in two cholestatic liver diseases) is guided for Q1 2027.

What the company does and where the pipeline stands

Assembly is a clinical-stage biotech: it has no approved products and earns its revenue from a research partnership with Gilead Sciences, which also owns about 25.1% of its stock. Under the October 2023 agreement, Gilead can take an exclusive license to any Assembly program by paying an opt-in fee of $45–125 million per program. Assembly then becomes eligible for up to $330 million in milestones plus royalties.

Two decisions this year changed the shape of the pipeline:

1. Gilead dropped one of its two herpes drugs. The licensed herpes program now depends on 1179 alone. Assembly's upside there comes from milestones and royalties, or from a 40% US cost-and-profit share that it can choose instead. The company says it will decide on the profit share only after Gilead sends its commercial cost estimates.
2. 6250 is now Assembly's main wholly owned asset. It is being developed for both HDV and cholestatic liver diseases (conditions where bile backs up in the liver and damages it). 6250 blocks NTCP, the liver-cell protein that both the hepatitis B/D viruses and bile acids use to get into cells. Bulevirtide, the only approved HDV drug, works the same way, but it has to be injected every day. 6250 is a once-daily pill: its Phase 1a showed a half-life of about four days.

Key figures

n/m = not meaningful (a percentage change between two losses).

Where the R&D money went (external program spending)

R&D (research and development) is the company's main cost. The 10-Q breaks out "external" spending, meaning payments to outside labs, trial operators and manufacturers, by program:

Management attributes the $1.2 million drop in total R&D to "lower external program expenses, particularly for 5366, reflecting the completion of clinical trials and lower manufacturing and toxicology study costs." Part of that saving was offset by spending on early-stage research and by higher staff costs from "increased headcount, annual salary adjustments and stock-based compensation." The quarter's spending is low because one set of trials has finished and the next has not started. Both 6250 Phase 2 studies are still in preparation, so R&D should rise again from Q4 2026. Management says it expects operating expenses "to increase over the coming years."

Takeaway: The narrower loss is mostly a one-off. Strip out the $5.1M revenue catch-up and the loss is roughly $9.0M, about where it was a year ago. What really changed this quarter is the funding: $320.4M of cash, plus a Gilead fee expected in Q4, gives Assembly money to run both 6250 Phase 2 programs into 2028–2029 without needing to raise again immediately. That cushion was bought by increasing the share count by about 25% in six months.

What the headline numbers hide

• Revenue is mostly non-cash. Gilead paid most of its money upfront, so Assembly books it as "deferred revenue" (cash received for work not yet done) and moves it into revenue as the work gets done. Deferred revenue fell from $36.9M to $16.3M in the first half, and that $20.6M reduction appears in the cash-flow statement as an adjustment. This is why the six-month operating cash outflow was $37.6M while the reported net loss was only $12.9M. The cash burn is the better guide to how fast the money is going. The company also lists the transaction price, the total Gilead payments being spread across the service period, at $54.4M.
• The $5.1M catch-up is a one-off. The filing says it came from "revisions made during the three months ended June 30, 2026 to estimated costs to complete the remaining performance obligation." When a company expects less remaining work, more of the fixed deferred amount counts as already earned. This does not mean the Gilead relationship produced more money. It also means less revenue is left to recognise in later quarters.
• Interest income is helping. Interest and other income nearly tripled to $2.5M because of "a larger portfolio balance" from the August 2025 and May 2026 financings and Gilead's December 2025 payment. This covers about 40% of the quarter's operating loss. It will shrink as the cash is spent.
• Per-share figures are flattered by dilution. Weighted shares rose from 7.66M to 18.85M. Even with the same dollar loss, the loss per share would have fallen by more than half. Shares outstanding went from 15.86M at December 31 to 19.85M at June 30, and to 20.36M by August 7. The May offering sold 3.92M shares plus pre-funded warrants at $26.50. Gilead bought 471,698 shares ($12.5M) in it. In July, holders exercised 510,205 Class A warrants at $21.60, bringing in about $11.0M.
• More warrant dilution is possible. Gilead holds 1.15M Class B warrants at $21.60 that can be exercised only between November 15 and December 31, 2026. They terminate automatically if Assembly announces at least $75.0M of non-dilutive collaboration capital before November 15. The 10-Q does not say whether the Gilead extension fee counts toward that. Another 1.15M Class A warrants and 179,500 warrants from 2024 at $17.00 are also outstanding.
• Stock-based compensation is rising. It was $2.4M in the quarter versus $0.8M a year ago, driven by performance-based restricted stock units granted in mid-2025. This non-cash cost explains most of the G&A increase and part of the R&D increase.
• No going-concern warning. The 10-Q contains no doubt about the company's ability to keep operating. Management states that its cash will last "into late 2028", and adds that "we may utilize our available capital resources sooner than we currently expect."

Outlook

Management's stated timeline (from the 10-Q):

• Q4 2026: Phase 2 start for 6250 in HDV. Additional Phase 1 pharmacology studies of 6250 to support labeling. Gilead's $75.0M extension fee on the third anniversary of the collaboration.
• By end of 2026: Gilead expects to start a Phase 2 of 1179 in recurrent genital herpes.
• Q1 2027: Phase 2 start for 6250 in PBC/PSC.
• Pending: Assembly's decision on the 40% US profit share for the herpes program, after Gilead's commercial cost estimates arrive. A partner search for 4334.

Our read. The next 12 months are mostly about starting trials, not reading out results. Neither 6250 Phase 2 study will have data in that window, so news on the stock is likely to come from the Gilead extension fee, whether the Class B warrants are exercised or terminated, and the profit-share decision. The profit-share choice is a real financial trade-off. Opting in would mean paying 40% of US development costs for 1179 in exchange for 40% of US profits, instead of milestones and royalties. With operating cash burn running at about $75M a year on the first-half pace ($37.6M in six months), that commitment could noticeably shorten the late-2028 runway, so how management frames it is worth watching.

On the science, 6250's Phase 1a looked clean: all side effects were Grade 1–2, there were no serious adverse events, and bile-acid levels rose as expected, which shows the drug is reaching its target. One cohort, at the highest single dose (25 mg), showed off-target activity on other liver transporters and a Grade 2 liver-enzyme elevation. That is one reason the multiple-dose cohorts were kept at 1 mg or below. Whether those low doses suppress the virus in HDV patients is the question the Phase 2 has to answer. The financial statements are now strong enough that this trial, not funding, is the main risk.