BIVI — FY2026 Annual Financial Report Analysis

BioVie spent more to finish two mid-stage trials, got a positive Parkinson's readout, and ended the year with about six months of cash

BioVie is a clinical-stage drug developer: it has no products on the market and has never had revenue, so its annual report is mainly about three things: how much it spent, what the spending bought in trial results, and how long its remaining cash lasts. In fiscal 2026 (the year ended June 30, 2026), the net loss widened 26% to $22.1 million. The cause was research and development (R&D), which rose 51% to $14.0 million as the company ran its Phase 2b SUNRISE-PD trial in early Parkinson's disease and its Phase 2 ADDRESS-LC trial in long COVID at the same time. Cash fell from $17.5 million to $9.0 million, and the auditor's report again includes a going-concern warning: a formal statement that there is "substantial doubt" the company can keep operating for the next year unless it raises more money.

At a glance

• $9.0 million cash vs. $19.0 million used by operations in the year. At last year's pace, that cash covers roughly 5–6 months from June 30, which means into around December 2026, so another share sale is very likely.
• SUNRISE-PD met its primary endpoint. In 57 patients, bezisterim reduced a composite blood measure of inflammation (−0.28 vs. +0.19 on placebo, nominal p=0.0018). Those were blood biomarkers, not the motor-symptom endpoints regulators approve drugs on.
• The per-share loss fell from $12.12 to $3.05, but only because there are more shares. The weighted share count rose almost fivefold after the August 2025 offering. The dollar loss grew.

The numbers

Year-over-year percentage changes for net loss and loss per share are not meaningful as "growth" figures when both years are losses, so they're described in words above rather than as a percentage in the summary figures. There is no operating margin to report: with no revenue, there is nothing to take a margin of.

Where the money went

Almost all of the $4.6 million increase in operating expenses was R&D. The 10-K splits direct clinical-study costs by program:

Two points stand out:

• The Parkinson's trial is the program paying for itself. Its cost more than doubled as the trial completed enrollment and the last patient was treated (May 2026). The company paid for it from its own cash.
• The long COVID trial was almost entirely grant-funded. BioVie won a $13.1 million clinical-trial grant in April 2024. By June 30, 2026 it had spent about $12.9 million on the study, and it had been reimbursed $12.9 million as of early August 2026. The reported R&D line is shown after subtracting reimbursements, so gross R&D was about $21.6 million (vs. $14.6 million in FY2025). The grant is close to used up, so it won't offset future costs much.

General and administrative costs were flat at $8.5 million. Higher stock-based pay for executives and directors was offset by about $0.6 million less in consulting fees.

Trial results: what was actually shown

SUNRISE-PD (Parkinson's, reported August 2026). 57 patients with early Parkinson's who had not yet started levodopa (the standard drug) took bezisterim 20 mg twice daily or placebo for 12 weeks. According to the 10-K:

• The primary endpoint, a composite of six blood-cell ratios used as markers of inflammation, improved on bezisterim (−0.28) and worsened on placebo (+0.19), nominal p=0.0018.
• A 15-item composite of motor and non-motor clinical measures (EPNIC-15, built from parts of the standard UPDRS Parkinson's rating scale plus a sleep scale) moved −0.04 on drug vs. +0.18 on placebo, nominal p=0.0006.
• Adverse events occurred in 39.3% of bezisterim patients vs. 51.7% on placebo, with no serious adverse events.

The company's own caveats matter here. The p-values are "nominal and unadjusted for multiplicity". That means no statistical correction was made for testing many endpoints at once, which raises the chance that some results are false positives. The FDA has not validated these biomarkers as stand-ins for clinical benefit in Parkinson's. And the trial was small and short. BioVie says it will use the data to design a "potentially pivotal" Phase 3 trial. That trial would be far larger and more expensive than anything the current balance sheet can fund.

ADDRESS-LC (long COVID). Enrollment finished in May 2026. In the 10-K, filed August 13, the company still expected topline results in "late summer 2026". As of this writing (early October), we found no results announcement among the company's EDGAR filings since then, so this readout is late compared with that guidance.

Liver (BIV201) and Alzheimer's. The company is "finalizing the protocol design" for a Phase 3 trial of BIV201 in cirrhosis with ascites (fluid build-up in the abdomen) and spent almost nothing on it this year ($18,000). No new Alzheimer's trial is under way. The last Phase 3 (topline November 2023) was left underpowered after the company excluded patients from 15 sites over protocol violations.

What the headline numbers hide

• Flat cash burn partly reflects unpaid bills, not lower spending. The net loss grew by $4.5 million, yet cash used in operations was the same $19.0 million both years. Part of the gap is non-cash: stock-based compensation rose to $2.9 million from $2.5 million. The larger part is timing. Accounts payable and accrued expenses rose by $1.5 million this year after falling by $1.4 million last year. That $1.5 million of bills still has to be paid in cash.
• Runway is short. Cash of $9.0 million plus a $2.2 million grant receivable gives about $11 million of liquid resources against a $19.0 million annual cash burn. Working capital (current assets minus current liabilities) was $9.8 million. With the PD trial closed out and the long COVID trial winding down, direct trial spending should fall for a while. Starting any Phase 3 would push it far above current levels.
• Dilution is the funding plan. Shares outstanding went from 1.91 million to 7.54 million in one year. Almost all of that came from the August 2025 offering of units at $2.00 each, which raised $10.5 million net. At June 30, 2026, warrants for 8.28 million shares (including 0.38 million pre-funded warrants, with many exercisable at $2.50) and options for 2.79 million shares were outstanding. Together that is more than the 7.54 million shares in issue. On September 11, 2026, BioVie signed an "at-the-market" sales agreement with A.G.P./Alliance Global Partners to sell up to $6.46 million of stock directly into the market over time. The company had also filed an S-1 registration statement (amended through mid-August) for a separate offering.
• The 10-K contains an unedited drafting note. The liquidity section's offering heading reads "Registered Direct Offerings [TO BE UPDATED FOR SUBSEQUENT CLOSE OF CAP RAISE]", and the section below it describes only the August 2025 deal. The filing reports 7,542,638 shares outstanding on August 10, 2026, the same as at June 30, so no new shares had been issued by then. It's a small issue, but a placeholder left in an audited annual filing doesn't inspire confidence in the review process.
• Small one-offs. A $43,544 "deemed dividend" from a warrant price reset adds to the loss attributable to shareholders, down from $369,465 last year. Interest expense nearly disappeared ($18,000 vs. $333,000) because the company repaid its note in December 2024. Interest income fell to $0.6 million from $0.9 million because less cash was invested. The last of the intangible-asset amortization ($0.18 million) ran off this year, and that line is now zero.
• No adjusted figures. BioVie reports GAAP results only, so there is no gap to reconcile between "adjusted" and reported earnings.

Takeaway: The year delivered what BioVie paid for: a positive (if biomarker-based and statistically unadjusted) Parkinson's readout. But the company ended with $9.0 million of cash against a $19.0 million annual burn, plus a going-concern warning. Its next major step, a Phase 3 trial, can only be funded by issuing many more shares relative to the 7.5 million now outstanding. For a shareholder, the bigger near-term risk is how much each share gets diluted, not the science.

What to watch next

• ADDRESS-LC topline. Guided for late summer 2026 and not yet reported in the company's filings as of early October. A clear result would strengthen the case for bezisterim working across several diseases. A null result would leave Parkinson's as the only active program, resting on one 57-patient trial.
• Phase 3 design and FDA feedback in Parkinson's. Watch whether the FDA accepts clinical endpoints that fit what SUNRISE-PD showed. The trial's strongest numbers were biomarkers the FDA hasn't validated.
• Fiscal Q1 2027 10-Q (quarter to September 30, 2026), due around mid-November. This will show how much the at-the-market facility raised, the new share count, and whether quarterly burn fell as the two Phase 2 trials wound down.

Our view: BioVie has scientific momentum but very little financial room. Unless it signs a partnership that pays cash up front, it will probably keep funding itself with small share sales at low valuations, which dilutes existing holders more with each raise. The long COVID readout is the event most likely to change that.