Xeris Biopharma Holdings, Inc. (XERS) stock intrinsic value fair value estimate & owner earnings analysis

Xeris Biopharma is a small US specialty drug company with three marketed products. Recorlev (levoketoconazole) treats Cushing's syndrome in patients for whom surgery was not curative; its 2025 net revenue was about $139 million, more than double 2024 on patient demand, and it is now roughly half the company. Gvoke, a ready-to-use liquid glucagon rescue pen for severe low blood sugar, brought in about $94 million, up 14% mostly on net pricing. Keveyis, for primary periodic paralysis, brought in about $48 million; it lost orphan exclusivity in August 2022 and faces generic dichlorphenamide from Torrent and a branded generic (Ormalvi), yet held within 4% of 2024 as volume offset price cuts. Total 2025 revenue was $291.8 million, and the first half of 2026 was $175.2 million, up 33%. A once-weekly injectable levothyroxine, XP-8121, is heading into a Phase 3 trial in 2026. How we measured earnings. We start from operating cash flow, strip out working-capital swings, and subtract stock-based compensation as a real cost (it was $22.4 million in 2025 and is rising). Capital spending is tiny (under $1 million a year). The $10.8 million a year of amortization on intangibles acquired with Strongbridge in 2021 is added back as a sunk cost; the fact that these product rights wear out is instead carried in each case through the dates exclusivity ends. On that basis the company burned cash every year from 2021 through 2024 while it launched Recorlev, and turned positive only in 2025. Because the business before 2025 was a launch-stage company rather than the one that exists now, we measure from 2025: owner earnings of $15.0 million, or about $37.4 million before $27.1 million of cash interest and excluding interest earned on cash. The first half of 2026 corroborates and improves on that level (operating income of $13.1 million against $1.4 million a year earlier). The 2025 cash flow swing, from about -$24 million to +$37 million before interest, has no meaningful percentage growth rate; our base case assumes a more modest step to about $45 million in 2026, already half visible in reported results, and about $48 million a year thereafter at the mid-2026 run rate. We do not credit further Recorlev patient growth, Gvoke price increases (except in the bull case) or any value from XP-8121: none is yet secured, and the rising research cost of the Phase 3 trial stays in the numbers as a cost. The company has $490.7 million of federal tax loss carryforwards, so we assume only about 5% cash tax until those are used, then 23%. The central question is how long Recorlev stays protected. Orphan exclusivity blocks generic approval until December 30, 2028. After that, protection rests on four Orange Book patents running to March 2040 that cover use of Recorlev in certain patients with persistent or recurrent Cushing's syndrome. Three generic makers have challenged them; Xeris sued Torrent and Somerset in New Jersey on February 26, 2026, which stays generic approval for up to 30 months. No trial date is set. Gvoke's glucagon formulation is covered by composition patents to 2036. Bear case: the generic challengers win or launch as soon as orphan exclusivity lapses, and generic Recorlev arrives in January 2029. Before then Recorlev and Keveyis net prices keep falling (Recorlev net price fell 7% in 2025), so cash flow before interest slips from about $44 million in 2027 to $40 million in 2028. In 2029 it turns to a loss of about $20 million as Recorlev revenue collapses faster than costs can be cut, then recovers only to about $5 million a year from a Gvoke and Keveyis remnant until Gvoke's patents lapse in 2036. This loss coincides with the Hayfin term loan coming due around 2029-2030, so refinancing would be expensive. The business is worth about $93 million against about $99 million owed after cash, so the company would owe slightly more than it is worth, by roughly $6 million, or about -$0.03 a share. Base case: the patent case ends in the kind of settlement common in these disputes, with generic entry around January 2034. Until then cash flow before interest runs about $48 million, easing about $2 million a year as Keveyis erodes. After generic entry it falls to $5-10 million from Gvoke and the Keveyis tail, and fades to nothing a few years after Gvoke's patents expire in 2036. The business is worth about $273 million. Against that it owes $200 million under the Hayfin term loan plus about $26 million for the loan's above-market cost (cash interest near 10-12% versus our 6% yardstick, over roughly 3.5 remaining years). It holds about $97 million of cash beyond operating needs (from $121.7 million at June 30, 2026) and would receive about $44 million if outstanding options were exercised. Net debt of about $85 million leaves about $188 million, or $0.93 a share. Bull case: the patents are upheld and Recorlev stays protected until March 2040. Gvoke's demonstrated pricing power offsets Keveyis erosion, holding cash flow before interest near $52 million a year through 2035 and about $50 million to 2039, before fading after generic entry in 2040. The business is worth about $459 million, net debt is about $79 million, and equity is about $380 million, or $1.89 a share. Per-share figures use about 201 million fully diluted shares. That total counts 173.6 million shares outstanding, 7.1 million options, about 9.3 million restricted units and the 2028 convertible notes as converted: $33.6 million at 8% converting at about $3.06 a share into roughly 11 million shares. The notes' reported fair value of $89.8 million at year-end 2025 shows conversion is far more likely than repayment. Shareholders' equity on the balance sheet was negative $9.7 million at June 30, 2026. The margin-of-safety price is $0.40 a share. With the bear case near zero and the outcome resting largely on one patent case and one refinancing, a wide margin of safety is required. What would make this wrong: a clear court win on the 2040 patents would move value toward the bull case, and continued rapid Recorlev patient growth would lift every case; a successful XP-8121 program is not valued at all. On the downside, an early generic launch while the term loan is outstanding leaves shareholders little. Not reconciled from the filings available to us: the second quarter of 2026 included a $30.8 million loss on debt extinguishment, and the full terms of the underlying debt change could not be read. The debt balances reported at June 30, 2026 ($200 million term loan principal now due in the fourth year out, $33.6 million convertible) were used as given. Some first-half 2026 working-capital lines were also estimated from partial data, which affects only the confirmation of the 2026 run rate, not the 2025 base.