INTERLINK ELECTRONICS INC (LINK) stock intrinsic value fair value estimate & owner earnings analysis

Interlink Electronics makes force-sensing resistors, printed electronics, membrane keypads and HMI panels (from plants in Shenzhen, China and Irvine, Scotland) and electrochemical gas sensors (Fremont, California). It is controlled by its chairman and CEO, Steven Bronson, who owns about 77% of the stock. Revenue was $11.9 million in 2025, sold to OEM customers in the United States, Europe and Asia; the top two customers were 18% and 9% of sales. What the business earns for its owners. We measure cash generated by operations after removing working-capital swings, stock compensation and the capital spending needed to keep the business running. On that measure Interlink earned $1.2 to $2.1 million a year in 2015-2017, when it was a smaller, high-margin force-sensor business. Since 2019 the picture has been very different: roughly -$0.38M (2019), +$0.23M (2020), -$0.27M (2021), -$0.71M (2022), +$0.29M (2023), -$1.39M (2024) and -$1.19M (2025), a median of about -$0.38 million a year, or about -$0.02 per share. Acquisitions (SPEC/KWJ in 2022, Calman in 2023 for $4.9 million, Conductive Transfers in 2024) lifted revenue from about $7 million to about $12 million but did not lift cash earnings; in 2024 and 2025 reported operating cash flow looked only mildly negative because inventory and receivables were being run down, which flatters cash flow once and cannot be repeated. Large amortization of acquired intangibles (about $0.7 million a year) is added back in our measure, so the losses above are not an accounting artifact. The window we use is 2019-2025, the business as it has existed since the older, more profitable force-sensor franchise faded; the 2015-2017 years describe a different, smaller business and are excluded. The first half of 2026 was better: revenue up about 13% to $6.8 million, gross margin about 44%, a second-quarter operating profit of $246,000, and underlying cash earnings of roughly +$0.2 million for the half. Most of that half-year cash came from collecting receivables and paying down payables, so it informs the near-term path but does not yet move the earnings level. The October 2025 conversion of the preferred stock into 900,000 common shares ended a $400,000 annual cash dividend; that helps the cash balance but does not change what the operating business earns. The company has no debt and pays no interest. At June 30, 2026 it held $1.8 million of cash. Its obligations are ordinary payables and about $1.0 million of short-lease obligations on its plants and offices, which run through operating cash flow. Base case. The 2026 improvement holds partly: the business runs at about break-even in 2027 and then loses roughly $200,000 a year, about half its 2019-2025 median, as the cost cuts from the Barnsley consolidation and the Shenzhen space reduction stick but mix and pricing pressure from customers without long-term purchase commitments persist. Nothing in the filings names a contract, backlog or capacity that would carry earnings beyond that, so no growth is credited. The business is worth roughly -$1.3 million on that path; adding the $1.8 million of cash leaves about $0.5 million, or about $0.03 per share. Bear case. A downturn like the 2019-2020 one, when revenue fell from $11.2 million (2017) to $6.9 million (2020), combined with loss of a top customer (18% of sales, or the customer that was 41% of receivables at end-2024) and higher China-origin tariff costs already flagged in the risk factors. Cash burn of about $1.2 million, $1.0 million and $0.6 million over three years consumes all of the cash; recovering roughly half of receivables plus inventory less payables and paying out remaining leases leaves about $0.3 million less than nothing. In that case the company appears to owe slightly more than it is worth, about -$0.02 per share, and would need new money from its controlling holder or a sale. Bull case. The levers the company has already shown work: the first-half 2026 gross margin of about 44% and quarterly operating profit hold, the Calman and Conductive Transfers businesses are integrated in a single UK site, and the gas-sensor line keeps its design-win base. Cash earnings of about $0.3 million in 2027 and $0.6 million a year thereafter, still well below the $1.2-2.1 million the smaller business earned in 2015-2017, then fading 5% a year after 2036 since there is no patent or contract we can point to that sustains it. Business value about $7.0 million plus $1.8 million of cash, or about $0.56 per share. Margin-of-safety price. We are not sure of the earnings level at all: the business has had positive cash earnings in only two of the last seven years, and the recent upturn rests on two quarters, one of which was a loss. The worst plausible outcome is that cash runs out within three years and the controlling shareholder recapitalises the company on his own terms, which minority holders would have no power to resist. There is no debt to force a crisis, but there is also no hard asset base: the plants are leased and most of the balance sheet is goodwill and acquired intangibles. If we are wrong about the cost savings and the stickiness of the 2026 margins, the shares are worth roughly nothing; if we are right, they are worth perhaps $0.35. The company pays no dividend, so nothing is paid to owners in the meantime. Price alone cannot justify a margin-of-safety price for a business that has not shown it can cover its costs. If it does prove that - two consecutive audited years of at least $300,000 of positive cash earnings - the base case rises to roughly $0.35 per share with the bear case near zero, and from $0.10 the base case implies a credible path to a 20% annual return over an assumed three-to-five-year holding period even if only part of that proved durable. Once that condition is met, the margin-of-safety price is $0.10. The prior analysis described this business similarly; its figures were built on a different basis and are not directly comparable, but the main change is that the 2024 and 2025 cash earnings were weaker than they first appeared once the one-time release of inventory and receivables is set aside.