MapLight Therapeutics: Outperform. Near cash, but the drug must repeat its win

At $7.97 a share, MapLight trades at roughly the cash we estimate it holds, so investors pay little for a schizophrenia drug with one moderate Phase 2 win that still needs a second trial. We rate the shares Outperform.

MPLT Outperform, medium conviction: the stock trades near its cash and no raise is forced within 12 months, but the lead drug has one moderate trial win and the cash will be spent on the next trial.

Rating: Outperform Medium conviction
Horizon: 12 months, relative to US-listed biotech stocks
Price basis: $7.97 on Oct 23, 2026, Nasdaq closing price (Stock Analysis quote page)

The call in brief

ZEPHYR met its primary endpoint with a moderate effect, but it is one five-week trial, the once-daily dose missed, and nausea and discontinuations ran above placebo.
At $7.97 the market values MapLight at about our $453.3 million cash estimate, leaving roughly $21 million for the pipeline, though most of that cash is committed to coming trials.
Cash through 2028 means no forced raise in the next 12 months, while the FDA meeting and Phase 3 start can re-rate the stock, or expose it if the second trial disappoints.

How our ratings work

We rate MapLight Therapeutics Outperform, with medium conviction, over the next 12 months. The rating is set against the Nasdaq close of $7.97 on October 23, 2026. MapLight has one schizophrenia trial that met its primary endpoint with a moderate effect, and, on our estimate, about $453 million of cash. At $7.97 the market values the whole company at roughly that cash. In other words, investors pay close to nothing for a Phase 2 result that now needs a second trial to be confirmed. We still think the price gives the drug too little credit, and that the company does not need to raise money in the next 12 months to prove it.

How we got here

One trial met its endpoint. In ZEPHYR, a five-week trial of 307 patients, the twice-daily dose of ML-007C-MA cut the PANSS score by 4.5 points more than placebo, with p=0.015 and an effect size of 0.37. Both the PANSS and severity results met the trial's multiplicity-adjusted threshold at the twice-daily dose. The company designed ZEPHYR as a registrational study, but it is one short trial and a second is still needed.
The market prices the company at about its cash. At $7.97, the 59.6 million shares and pre-funded warrants are worth about $474.7 million. We estimate cash at September 30 of about $453.3 million. The difference, about $21 million, is what the market pays for the pipeline (table below). That treats the cash as free. MapLight spends about $48 million a quarter, so most of the cash is committed to the work the pipeline needs.
There is no forced raise inside our 12 months. MapLight says its cash lasts through 2028. Even if that excludes the confirmatory trial, as we suspect, the company can reach its next decision points without selling stock at today's price.
The fears are real, and only partly priced. The effect size is moderate, the once-daily dose missed, the autism program missed its primary endpoint, Bristol Myers Squibb already sells a drug in the class, and one holder controls half the vote. We think a near-zero value for the pipeline covers these in most outcomes, but not if the second trial fails.
So: the next year brings the FDA meeting that sets the Phase 3 plan and, likely, the trial's start, with no financing overhang to absorb. That is an Outperform. Conviction is medium because of control and because the stock has fallen from $10.86 on September 22 to $7.97, which tells us sellers may not have finished.

The trial: a clear win, with limits

In patients who completed ZEPHYR, the gap was 6.0 points, p=0.002, an effect size of 0.50. The once-daily 330/6 mg dose did not reach significance: a 2.8-point difference, p=0.110. The company reports no serious or drug-related severe adverse events and 2% discontinuation from stomach side effects. The fuller picture is less tidy: 7.1% of patients on the twice-daily dose stopped because of adverse events, against 1.0% on placebo, and 29% had nausea (20% mild plus 9% moderate, against 6% on placebo). A prespecified cognition measure showed an effect size of 0.51, with nominal p=0.041, in 68 patients.

We do not overstate it. An effect size of 0.37 is moderate. The company's deck places it within the range of approved antipsychotics and warns that cross-trial comparisons are uncertain, but in the deck's own chart, where those effect sizes run from about 0.26 to 0.63, 0.37 sits toward the lower end. Across the active arms, 19.9% of patients left the trial, and five weeks says little about long-term safety. The completer figure leaves out dropouts, and the cognition result is a small secondary. MapLight plans one more registration trial, a normal two-trial path, and its design depends on an end-of-Phase 2 meeting with the FDA, per the June quarter 10-Q. Its size and design are not agreed. The second program also missed: ML-004 did not meet its primary endpoint in the 161-patient IRIS trial in autism.

Cash, burn and the 2028 claim

At June 30, 2026, MapLight held $351.3 million in cash, cash equivalents and investments. The August private placement added about $150.0 million gross.

Operating cash burn was $96.0 million in the first half of 2026, about $48.0 million a quarter. If the September quarter matched that, cash at September 30 was about $453.3 million before placement fees. That is our estimate; the third-quarter 10-Q will give the real number.

From June 30, 2026 to the end of 2028 is ten quarters. About $501 million over ten quarters allows about $50 million a quarter, close to today's burn. The company says it plans to increase research spending, and the filings do not say whether the 2028 guidance includes the confirmatory trial. Our base case is that a raise comes in late 2027 or 2028, after the Phase 3 design is set and ideally after it starts. That timing matters: a raise into a started Phase 3 should cost less dilution than one today.

What the price already assumes

	Amount
Common and non-voting common shares, September 1, 2026 (424B3)	55,575,719
Pre-funded warrants from the August placement	3,983,168
Market value at $7.97	about $474.7 million
Our estimate of cash at September 30, 2026	about $453.3 million
Value the market puts on the pipeline	about $21 million
Same, counting 3,935,504 unvested RSUs as shares	about $53 million

The share price has come down in steps: $17.00 at the October 2025 IPO, $11.38 in the August placement, $10.86 on September 22, per the resale prospectus, and $7.97 on October 23. Nothing in the trial data has changed since July. What did change is that the resale prospectus made up to 13,181,055 placement shares free to trade in the open market, and in our view the market has been absorbing them.

One signal we would not read as negative: on October 15, six insiders, including the chief executive, sold shares at prices between $7.755 and $8.3732, per Form 4 filings such as the chief executive's. Each filing says the sale covered tax withholding when restricted stock units vested and "does not represent a discretionary sale".

Control

Catalyst4, Inc. paid $83,400,153.82 for 4,400,000 shares and 2,928,686 pre-funded warrants in the August deal, about 55.6% of it, and now reports 49.9% of the voting common stock. Novo Holdings cut its stake to 4.9% after selling at $12.00 to $15.00 on July 30. A holder that keeps buying is a vote of confidence. It also means outside holders have little say over any future financing, partnership or sale, and the risk that a dominant holder offers to acquire the rest cheaply belongs on any list of what can go wrong.

Bull case, bear case

Bull case (55%). The FDA meeting supports a single confirmatory trial of reasonable size. MapLight starts it in 2027 and keeps tolerability as its edge over the class. The market starts to pay for the pipeline again, and the shares do better than the average biotech stock. VISTA, the Alzheimer's psychosis trial with Fast Track status, adds a second shot in a condition with no FDA-approved medicine.

Bear case (45%). The second trial shows a smaller effect than ZEPHYR's 0.37, or misses. A moderate effect from one short trial gives little cushion, and the once-daily dose already failed. Nausea in 29% and 7.1% stopping for adverse events count against the drug in a crowded class. The FDA asks for a larger or longer trial. Burn rises faster than guided, and MapLight raises equity at a low price, led again by Catalyst4. Competition from Bristol Myers Squibb's Cobenfy and Neurocrine's three candidates makes partners cautious. With no pipeline value left, the shares fall toward the cash that remains after the trial, which is well below $7.97.

We weight the bull case slightly more because the price already sits near cash and the company says it is funded through 2028. A failed second trial would leave no lead drug and a shrinking balance, so the gap between 55% and 45% is small.

Catalysts

10-Q for the quarter ended September 30, 2026. Real cash and burn, against our $453.3 million and $48.0 million.
End-of-Phase 2 meeting with the FDA. The size and design of the confirmatory trial, and whether one trial is enough.
Start of the confirmatory schizophrenia trial, and whether the company says the 2028 runway funds it.
ML-004 in autism. Any partner or decision to stop.
VISTA topline in the second half of 2027, at the far end of our window at best. ClinicalTrials.gov lists December 2027 for primary completion.

Risks to our rating

We would move to Neutral if the FDA asks for a trial that the cash clearly cannot fund, if quarterly burn runs well above $50 million, or if the stock rose more than 25% from our price basis without new information. We would move to Underperform on a discounted raise before the Phase 3 design is set, or a buyout offer from the controlling holder at a low premium.

This is our opinion as of October 28, 2026, set against a share price of $7.97 on October 23, 2026. It is not personal investment advice.

Sources

Every figure comes from the documents above. Chart values come from the company's financial statements as tagged in its SEC filings; derived values are labelled. If you find an error, tell us.