KOSS CORP (KOSS) stock intrinsic value fair value estimate & owner earnings analysis

Koss Corporation is a Milwaukee headphone maker founded in 1953. Over 90% of its products are made by contract manufacturers in China and Taiwan. Sales for fiscal 2026 (the year to June 30, 2026) were $13.0M. About 83% came from the United States, where direct-to-consumer online sales grew 20.6% and now make up nearly 30% of revenue. The rest was exported, mostly to European distributors, and those export sales fell 41% last year. Sales have fallen from $24.1M in fiscal 2017 to $17.7M in fiscal 2022 and to roughly $12-13M a year since fiscal 2023. Selling, general and administrative costs have stayed at about $6-7M a year throughout, so the operating business now loses money every year. How we measured it: owner earnings are operating cash flow with working-capital swings removed, minus stock compensation and the capital spending needed to keep the business running. At June 30, 2026 Koss held $19.8M in cash and U.S. Treasury bills, which earn it roughly $0.9M a year in interest. Because we count that cash at full value, we take the interest out of the operating figures so it is not counted twice. On that basis the headphone business consumed $2.17M of cash in fiscal 2024, $2.07M in fiscal 2025 and $1.28M in fiscal 2026. The fiscal 2026 figure was helped by a one-time tariff refund of about $753K; without it, the loss was about $2.03M. We anchor on the median of these three years, a loss of about $2.07M, or $0.22 per share. We use only these years because the business shrank to its current size in fiscal 2023. Fiscal 2023 itself is excluded because it included a one-time $33M patent licence payment, against which large legal fees were charged. The $250K of licence income in fiscal 2026 was offset by matching legal costs, and the company calls these proceeds non-recurring, so we give patent licensing no ongoing value. What shareholders own besides the business: the $19.8M of cash and Treasuries, of which we set aside $1.5M as working cash. The company also holds company-owned life insurance with a $6.9M cash surrender value. Against this, it owes a $2.46M deferred-compensation obligation to a current senior officer, payable for his life and then his spouse's, plus $22K of finance leases. There is no bank debt. The net comes to about $22.7M, or $2.40 per share. What the shares are worth is that sum minus what the business is expected to lose before the losses stop. Bear case, $0.25 per share: two things that helped fiscal 2026 do not recur, a large custom order for the education market and the tariff refund, and export sales stay weak. Losses then run at about $2.5M a year. Nothing in the record shows management cutting costs as sales fell, so the losses continue until the company's resources are used up, in about 12 years. In this case the business is worth about -$20.4M, the company's net resources are $22.7M, and the difference leaves shareholders about $2.4M. Base case, $0.79 per share: losses continue at the recent median of about $2.07M a year for ten years, by which time most of the surplus cash has been spent. That loss is close to fiscal 2026's underlying loss of $2.03M once the tariff refund is excluded, because the improvement reported last year came almost entirely from that refund. After the ten years we assume the company runs at breakeven, as it roughly did from fiscal 2017 to 2022 when it had little spare cash. We credit no profits beyond that point because nothing in the record supports them. In this case the business is worth about -$15.3M against $22.7M of net resources, leaving about $7.4M. Bull case, $1.40 per share: losses hold at the fiscal 2026 reported level of about $1.28M a year, the best of the last three years. Two things support this: the Section 122 tariffs on China-made goods expired on July 24, 2026, and direct-to-consumer sales are growing. After ten years the business reaches breakeven. In this case the business is worth about -$9.4M against $22.7M of net resources, leaving about $13.3M. No case credits sales growth or wider margins. The small sales gains in fiscal 2025 and 2026 came from one-off custom orders and mix rather than a repeatable process. All three cases discount at 6%. The cash itself earns somewhat less than that, so the estimates slightly flatter it. Since our previous analysis: that analysis counted only about $1.9M of cash and missed the $16.8M of Treasury bills. It also treated the business as roughly breakeven because Treasury interest was mixed into operating results. Now that the two are separated, we credit the full cash balance and charge the full operating losses. This raises our bear and base estimates and slightly lowers our bull estimate. What would make this wrong: on the downside, losses could grow faster, or cash could be spent on acquisitions or further litigation. On the upside, management could cut overhead to fit the smaller sales base, sell the company, or return cash to shareholders. None of these has been announced. The CEO controls about 40.6% of the shares, including shares held in a voting trust, so outside shareholders have little say over how the cash is used.