Neostellar Capital Corp. (NSLR) stock intrinsic value fair value estimate & owner earnings analysis

Neostellar Capital Corp. (formerly SuRo Capital; ticker changed from SSSS to NSLR in July 2026) is a publicly traded investment company. It owns minority stakes in technology companies, nearly all of them private, and earns nothing from operations: its value is what those stakes can be turned into, less the debt and the cost of running the company. Operating cash flow mostly records buying and selling holdings, so it does not describe what owners earn, and we value the holdings directly. Our previous estimates were set on a different basis (reported net asset value per share) and are not comparable to these. What it owns. At June 30, 2026 the company held 37 companies (34 private, 3 public) with a reported fair value of $405.9 million, against net assets of $355.9 million, or $13.44 per share on 26.47 million shares. Whoop is 37% of the holdings ($150.8 million); an OpenAI stake held through an ARK fund is 15% ($59.3 million); VAST Data is 8% ($34.0 million); TensorWave, held through a Magnetar vehicle, is 5% ($20.2 million); Blink Health is 4%. At December 31, 2025, ten companies made up 87.5% of net assets and artificial-intelligence infrastructure alone was 30.6% of the portfolio. Net asset value rose from $8.09 at year-end to $14.24 at March 31 as holdings were marked up by $158.7 million after outside financings: OpenAI at an $852 billion valuation, Whoop at $10.1 billion, and a Plaid tender at $8 billion. Most holdings are valued on management's judgment rather than market quotes (Level 3 investments were $221.3 million of $225.5 million at year-end 2025). What changed in the business. On July 15, 2026 the company moved to an outside manager, Neostellar Advisors, owned by company executives and a Magnetar affiliate. It charges 1.75% a year on gross assets plus incentive fees on income above a 7% annual hurdle and on gains from new investments only; existing holdings carry no incentive fee. About $20 million of one-time transition costs were taken in the second quarter and are already in the June net asset value. The company no longer has employees, so the large stock-compensation charge ($11.7 million in the first half) should not recur. In July a Magnetar affiliate lent $20 million at 6.5% due July 2029; it converts to shares only if the company raises at least $230 million of new common stock at or above net asset value, and is repayable at 105% on a change of control. In the second quarter $8.0 million of convertible notes converted into 1,092,504 shares at $7.32; $27.0 million of 6.5% convertible notes remain, convertible at $7.32 into about 3.69 million more shares. $35.8 million of 6.00% notes mature December 30, 2026; the company reports $42.2 million of liquid assets at September 30, enough to repay them but leaving roughly $6 million. What it costs to run. The company has lost money before investment gains every year: $13.4 million in 2023, $14.0 million in 2024 and $16.5 million in 2025. Under the new structure we estimate roughly $14 to $15 million a year, about $0.50 per diluted share (management fee about $7.7 million on roughly $440 million of gross assets, interest about $5.2 million on $82.8 million of notes, other costs about $2.5 million). No dividend has been declared, so nothing is paid to shareholders while holdings are sold. How we built the cases. Each case values the holdings as they are turned into cash, subtracts the costs above and the debt as it falls due, and counts cash received later for less than cash received sooner. We credit no future gains. The filings do not secure any: the portfolio's own marks moved by -$111.6 million in 2022, +$30.5 million in 2023, -$19.0 million in 2024, +$32.1 million in 2025 and +$158.3 million in the first half of 2026, which is not a steady, repeatable process, and no filing dates the exit of any large holding. Management says Whoop's CEO expects an IPO in 2027; we treat that as a statement of intent, not a secured event, and use it only in the bull case. Base case: $9.50 per share. Holdings are held at their June 30 marks and sold evenly over five years (a judgment; nothing in the filings dates the exits). The convertible notes convert, since they are far in the money. Before costs and waiting, holdings of $405.9 million plus $34.3 million of cash (including the Magnetar loan) less $55.8 million of non-convertible notes and $2.2 million of other liabilities leave $382.1 million, or $12.67 per diluted share. Running costs and the wait for cash bring that to $9.50. Each extra year of selling lowers this by about $0.47; each year fewer raises it by about the same. Bull case: $10.50 per share. Marks are the same, but the large holdings are sold over three years, which is what an on-schedule Whoop IPO followed by the other large positions would produce. The same $382.1 million, or $12.67 per share, is reduced by a shorter period of costs. We credit no further markups; the portfolio has repriced upward before, but nothing secured says it will again. Bear case: $4.35 per share. The industry's own record is the guide: from 2021 to 2022 this company's net asset value per share fell 37%, from $11.72 to $7.39, with $111.6 million of unrealized losses in 2022. We apply a 37% decline to every holding, assume exits stall so that selling takes eight years, and keep the full costs. Holdings fall to about $255.7 million; with $34.3 million of cash and $85.0 million owed ($82.8 million of notes and $2.2 million of other liabilities), $205.0 million remains, or $7.74 per share before running costs and the long wait. The convertible notes sit about at their conversion price in this case, so we treat them as repaid in cash in 2029. The first-year cash need is met from cash on hand plus sales. Because Whoop alone is about $5.00 of each diluted share, a decline concentrated in a few late-stage names could be worse than the average 2022 experience; we have not modeled that. What would make this wrong. The company's early, unaudited estimate puts net asset value at September 30 between $12.25 and $12.75 per share, below June's $13.44. It points to realized losses on the exits of Learneo ($5.6 million), Aventine ($2.5 million) and Catona Climate ($1.3 million), partly offset by gains of about $1.7 million; those items and a quarter of running costs do not account for the whole decline, and we could not reconcile the rest until the quarterly report. Our figures start from the filed June marks. If the decline reflects roughly 6% lower marks on holdings, the base case would be about $8.85. If Whoop alone were marked down by half, the base case would be about $7.50. Margin-of-safety price: $5.50. We are not sure of the level: nearly nine-tenths of net assets sit in ten companies, most marks rest on judgment, and net asset value moved 76% in a single quarter on three outside financing events. The worst plausible outcome is a repeat of 2021-22, which the bear case captures at $4.35 after costs; outcomes it does not model include a large loss in one holding, a forced raise at a poor price, or a failure to refinance the Magnetar loan and convertible notes in 2029. The structure would survive the first of these: the December 2026 notes are covered by cash on hand, and the later debt is small against $405.9 million of holdings, but after December liquidity is thin against roughly $14 to $15 million a year of costs, so each year depends on exits such as the $15.4 million of restricted CoreWeave-related shares and distributions like the $6.5 million received in the second quarter. The company pays nothing to shareholders during the holding period, and costs of about $0.50 a share a year come out of the portfolio. The price assumes a four-to-five-year holding period, the period over which the base case turns holdings into cash. From $5.50, the base case implies about 22% a year; if marks are 6% lower it implies about 20%; if Whoop is marked down by half it implies about 14%; and the bear case implies roughly breaking even (about 1% a year) over eight years. That is why the margin-of-safety price is set here: it is the highest price at which the base case clears 20% with room for the unexplained September decline, and the bear case does not imply a loss.