PIXELWORKS, INC (PXLW) stock intrinsic value fair value estimate & owner earnings analysis

Pixelworks spent two decades building and selling display-processor chips used in smartphones, projectors, and other consumer electronics. That chip business is gone: on January 6, 2026, the company completed the sale of its semiconductor subsidiary (which made the mobile and home/enterprise chip products) to a Chinese buyer for approximately $51 million in net cash. This was a fundamental change in what the company is, not an ordinary year of results, so we've valued Pixelworks as it exists today rather than as an average of its chip-making past. What remains is a much smaller company centered on TrueCut Motion, a technology for controlling motion blur and judder that filmmakers and display makers can license to keep a movie or show looking the way its creator intended across different screens and frame rates. This business generated about $0.7 million of revenue in both 2024 and 2025, at attractive (85%) gross margins, but the licensing/certification model has not yet been adopted broadly, and running it alongside the costs of being a public company produced an operating loss of roughly $8-12 million over the last two years. The company has been cutting costs steadily (three restructuring rounds since mid-2024), which narrowed the loss meaningfully from 2024 to 2025, but revenue itself has not grown. The dominant asset on the balance sheet today is cash: roughly $53 million, against almost no debt or other liabilities, on about 6.3 million shares outstanding (reflecting last year's 1-for-12 reverse stock split, done to stay listed on Nasdaq, which has since worked). That is about $8 of cash for every share. The central question for value is not whether the historic chip business was worth something -- it no longer exists -- but how much of that $8 per share of cash will be spent keeping the TrueCut licensing effort alive, and whether that effort ever becomes self-sustaining. Our base case assumes TrueCut Motion gains only slow, incremental licensing traction and cost discipline continues, so losses narrow gradually over the next decade without ever clearly turning profitable. On that path, a large share of today's cash gets spent funding the business before it can support itself, leaving modest value beyond the cash itself. Our bear case assumes the licensing business never gains real commercial traction; spending continues near recent levels for several more years until the cash cushion is substantially used up, after which the company scales back sharply. Because there is no debt, the downside is bounded by the cash on hand and the pace at which it can be spent -- there isn't a way for shareholders to end up owing money -- so the bear and base outcomes land fairly close together. Our bull case assumes TrueCut Motion is adopted as a real standard by a handful of major display makers or studios within a few years, the business becomes durably profitable at a modest scale, and the cash is preserved rather than spent funding losses; this case is worth several times the others because it is the only one where the company becomes something more than its cash balance. On this basis we estimate the shares are worth roughly $1.35 in our bear case, $1.60 in our base case, and $18.80 in our bull case -- a wide range that reflects a business with a firm cash floor but genuine, binary uncertainty about whether its one remaining product ever finds a market. The margin-of-safety price is around $1.10, and would move up toward the base-case estimate if there's concrete evidence the licensing business is gaining real customers, or if management commits to returning cash it doesn't need rather than spending it in pursuit of an unproven product. A few things worth noting plainly: the cash-flow statement does not separately break out how much cash the continuing licensing business used versus the now-sold chip business in past years, so our estimate of the licensing business's recent cash burn is built from its reported revenue, costs, and net loss rather than read directly off the cash flow statement -- we consider this reliable but it is a step removed from the ideal source. We also did not assume any value from the roughly $50 million of accumulated tax losses the company has built up over its unprofitable history, which could have some value in a future transaction but which we have not tried to price. Finally, because the current form of this company has only about two years of separately reported results, our estimates carry more model uncertainty than a business with a long, stable operating history would, and we've built in extra margin of safety in the margin-of-safety price for that reason.