AMARIN CORP PLC\UK (AMRN) stock intrinsic value fair value estimate & owner earnings analysis

Amarin sells one drug: VASCEPA (sold as VAZKEPA in Europe), a purified fish-oil medicine (icosapent ethyl) approved to cut cardiovascular events in high-risk patients. In 2020 several of its U.S. patents were declared invalid, and generic versions from Hikma, Dr. Reddy's, Teva, Apotex and others have been on the U.S. market since. The company is now a shrinking single-product business living off its brand loyalty, its foreign licensing partners and a large cash balance. What the business looks like today. Revenue was $213.6 million in 2025, down from $614.1 million in 2020. U.S. product sales were $154.1 million, down 8%, and Amarin's share of U.S. icosapent ethyl prescriptions fell from about 53% to about 47% in a single year, with branded prescriptions down 10%. In June 2025 Amarin licensed VAZKEPA across 59 mainly European countries to Recordati for a $25 million upfront payment, up to $150 million in sales-based milestones and royalties, and cut its European sales force; it expects roughly $70 million a year of savings. European patent protection runs to 2039. Revenue in the first half of 2026 was $87.3 million, against $114.8 million a year earlier. Amarin has no borrowings. At June 30, 2026 it held $314.6 million of cash and short-term investments, plus about $164 million of inventory. What the owners actually earn. We measure owner earnings as operating cash flow, stripped of working-capital swings and charged for stock compensation. Capital spending is close to zero. On that basis Amarin lost money in nine of the last ten years. The only exception was 2021, at about $0.62 per ADS. Losses were about $5.10 per ADS in 2022, $2.94 in 2023, $4.13 in 2024 and $1.90 in 2025. Two adjustments matter. First, in 2023 to 2025 operating cash flow looked far better than these figures because Amarin was turning working capital, mainly surplus inventory, back into cash: about $119 million over three years. That is a real but one-time release, and we count it separately rather than as earning power. Second, interest on the cash pile, about $11 million in 2025, is left out of the business's earnings because the cash itself is counted in full. Excluding interest income, the $36.2 million of 2025 restructuring costs and the $25 million Recordati upfront payment, the business lost about $39 million in 2025. The first half of 2026 ran at about $20 million, an annual pace of roughly $40 million. That is the level we start from: an operating business that, after the restructuring, still consumes cash. Growth credited and left out. We credit no growth. The Recordati milestones and any expansion of royalties depend on future European reimbursement and on Recordati's sales, which have not yet happened. The Canadian and Chinese milestone schedules also depend on partner sales. All of these are left out of the base case. The one thing the filings show is decline: U.S. pricing and share are falling under generic competition, and that is the path every case follows. Base case: about $8.52 per ADS. U.S. revenue keeps eroding. The operating loss, excluding interest, widens to about $42 million in 2027 as revenue falls faster than the remaining costs. It then narrows steadily to zero by 2034, as management shrinks the cost base toward a royalty-and-supply shell. Surplus inventory is turned into about $12 million of cash a year for five years. About $25 million is set aside for the antitrust suits brought by Dr. Reddy's, Hikma, Teva and Apotex and the six class actions over API supply. In February 2026 the court refused to dismiss the Teva and Apotex claims. Nothing is assumed after 2033: no product replaces VASCEPA, and the European royalty tail is not credited. The business itself is worth about negative $136 million. There is no debt. Adding the $314.6 million of cash gives about $179 million, or $8.52 per ADS. Bear case: about negative $2.90 per ADS. U.S. erosion continues at the recent pace of 10% or more a year in branded prescriptions. Management does not cut costs fast enough, and losses of $30 to $50 million a year run through 2035. The antitrust cases end in a $100 million settlement or judgment in 2028; the plaintiffs seek treble damages, so a figure of this size is plausible. Only $20 million of inventory is ever turned into cash. On this path the business destroys about $376 million in present value, more than the $314.6 million of cash. The shares would be worth less than nothing, by about $61 million. There is no debt in this story. What would leave owners with nothing is years of continued losses. Bull case: about $17.69 per ADS. Branded U.S. share stabilizes and the cost cuts take full hold. Losses shrink to zero by 2030, and a small royalty-and-supply business earns about $8 million a year from 2031 to 2039, supported by the European patent. About $100 million of surplus inventory is turned into cash, and litigation costs nothing. The business is worth about $57 million, plus the $314.6 million of cash, for about $372 million in total. The worst plausible outcomes are a large antitrust judgment, or management choosing to spend the cash on an acquisition or new development program rather than return it. Neither is ruled out by anything in the filings. Amarin has had a $50 million buyback authorized since 2024 and had not bought back any shares by the end of 2025. If we are wrong about the pace of decline and losses stay near $45 million a year through 2035, the shares are worth about $0.70 per ADS. If the base case holds but litigation costs $100 million, they are worth about $5.34. Margin of safety price: $3.50 per ADS. We are fairly sure of the earnings level: the business loses money, and has for nine of ten years. The value rests almost entirely on the cash and on how fast it is spent. The $3.50 price is well under a quarter of the cash per ADS, about $14.98. If the base case plays out and the shares come to reflect it over three years, the base case implies a return above 40% a year from that price; the heavy-litigation case implies about 22% a year. In the bear case the loss would be total. That is why the margin-of-safety price is no higher. The business pays no dividend over the holding period, and in the base case each of the next few years uses up roughly $1.40 to $1.50 per ADS of cash, and in the bear case about $1.90 to $2.40 before any litigation payment. Our previous figures were stated per ordinary share, before the 20-to-1 ADS ratio change in April 2025. These figures are per ADS, the unit that trades, so the two sets of numbers are not comparable.