Nektar Therapeutics NKTR

Key data

The business

Nektar is now, in practice, a one-drug company. That drug is rezpegaldesleukin, or rezpeg, an engineered interleukin-2 molecule built to expand regulatory T cells, the cells that act as the immune system's brakes. Turn the brakes up and inflammation from an overactive immune response is meant to fade. The lead target is moderate-to-severe atopic dermatitis, the chronic eczema that Sanofi and Regeneron's Dupixent turned into a multibillion-dollar market. In March 2026 the Phase 2b REZOLVE-AD trial read out positive and the stock tripled off its $33.40 low. In July 2026 Nektar started the two registrational Phase 3 trials, ZENITH AD-1 and ZENITH AD-2, in patients twelve and older. The mechanism is well described, but no primary-literature citation (a DOI, PubMed id, or trial NCT number) could be pulled this run, so the science here rests on the company's 10-Q, not on a paper I read. Everything else, the old oncology programs and the bempeg wreckage, is history. What is left besides rezpeg is a shrinking stream of legacy royalties. The moat, if the durability data hold, is the conjugate chemistry and the claim that patients stay clear after they stop dosing, which Dupixent cannot say.

The numbers

The last five quarters show a steady loss and an eroding top line.

Q4 2025 is derived (full year less nine months) and its revenue almost certainly carried a one-time item; the underlying royalty line is drifting from $11.2M to $10.1M. The net loss holds near $40M a quarter. The shrinking per-share loss, from -$2.95 to -$1.23, is not progress, it is the denominator: shares went from about 14 million to 33 million in a year. Each quarter's loss came in smaller than the Street looked for (-$1.23 against -$1.97 in Q2 2026), but a smaller loss spread over a bigger share count is a mirage.

The 2024 and 2025 per-share figures reflect a 1-for-15 reverse split and are not comparable to the earlier years. Revenue halved in 2025 as royalties rolled off, and the cumulative net loss since 2021 is about $1.45B.

The whole company is a race between cash and a readout. Working capital is $668M, burn ran $209M in 2025, so roughly three years of runway into 2029. But two global Phase 3 trials cost more than the mid-stage work that got here, burn will rise, and a raise before the data is the central read, which means the 33 million shares become more still. Strip out the cash and the market is paying about $1.5B for rezpeg. Here is what the market has not settled: it prices rezpeg as probably-works-in-eczema, one drug, one indication. The honest read is binary. If ZENITH AD confirms the Phase 2b durability, rezpeg is a Dupixent challenger that reads across to other immune-driven diseases, worth multiples of today; if it misses, the floor is the cash, somewhere around $15 to $20 a share. The single print that settles it is ZENITH AD-1 topline, expected in 2028.

Management

Howard Robin has run Nektar since 2007, which means he presided over the 2022 bempegaldesleukin collapse, when a melanoma drug partnered with Bristol Myers in a deal reported at up to $1.8B failed across trials and took roughly 90% of the stock with it. That is the scar tissue here: this company has turned a promising immune-modulator into a wipeout before. Management ran a 1-for-15 reverse split in 2025 to hold its listing, and funds itself by selling stock, diluting holders about 134% in a year. Insiders sold $2.7M over twelve months across twenty sales with zero buys, the largest from Robin and chief scientist Jonathan Zalevsky; the sums are small against a $2.1B cap and plan status is not disclosed, so read it as routine rather than a signal. Whether the pay matches the record is your call; that record shows a team that can advance science and destroy capital in the same decade.

How it fails or surprises you

Phase 3 misses the bar. Dupixent clears roughly 60% of patients to EASI-75 at four months. Rezpeg's Phase 2b hit, but Phase 3 in atopic dermatitis is where good mid-stage drugs die. If ZENITH AD-1 or AD-2 misses EASI-75 or IGA 0/1 at week 16 in 2028, the $1.5B pipeline premium deflates to cash, a 60%-plus fall from $74.

A raise before the readout. Runway is about three years, but two global Phase 3 programs plus adolescents cost more than the trials that got here, and the royalty line now covers almost none of it. A secondary priced before topline data, the central read, adds to a share count already up 134% in a year and caps the upside even if the drug works.

Durability reads across (right tail). The Phase 2b claim is not just clearance but clearance that holds after patients stop dosing. If Phase 3 confirms that, rezpeg is differentiated from Dupixent and the same Treg mechanism plausibly extends to alopecia areata, lupus, and other autoimmune targets. The market pays for one indication today; a platform would be worth multiples of $2.1B.

Closing thoughts

This is a setup where one print settles it. Nektar is a $2.1B option on a single week-16 data cut, and nearly everything between now and 2028, financing, enrollment, a Phase 2b in another skin disease, moves the stock without answering the question. When ZENITH AD-1 and AD-2 read out, rezpeg is either a Dupixent-class drug with a durability edge or it is not, with little middle ground: a clean EASI-75 and IGA win with response maintained off-drug makes the platform, while a miss on either endpoint, or a fade in durability, leaves a cash-box worth $15 to $20 a share against a $74 price. The left tail is the deeper one, a 60%-plus drop, and the Phase 2b hit is the only thing standing between here and it; that raised the odds, it did not remove the risk. Call the distribution roughly even with fat tails on both ends, weighted by how much you trust one mid-stage trial to repeat at scale.

The bet is still that rezpeg clears eczema in Phase 3 the way it did in Phase 2b, and clears it durably, the thing Dupixent cannot claim. What breaks it is a week-16 endpoint miss, or a raise that dilutes you before the data land. The two numbers that tell you first: the EASI-75 response rate against Dupixent's roughly 60% bar, and the share of responders who stay clear after stopping the drug. Everything else is waiting.