$25.00
$122.00
bear $25.00
base $79.00
bull $122.00
margin of safety $38.00
market price* $141.76
Duolingo runs the world's most widely used language-learning app. In 2025 it had about 133 million monthly and 52.7 million daily active users, and 12.2 million paying subscribers (up 28% in the year), so roughly 9% of monthly users pay. Subscriptions were $873 million of the $1,038 million of 2025 revenue; the rest is advertising, the Duolingo English Test and in-app purchases. Revenue grew 39% in 2025. About 38% of revenue comes from the United States and 62% from the rest of the world. How we measured earnings. We start from operating cash flow and take out three things: the cash that came from customers prepaying subscriptions and other working-capital swings, the cost of stock-based pay ($137 million in 2025, treated as a real expense), and the part of capital spending needed to keep the business running. On that basis owner earnings were about $182 million in 2025, or $3.76 per diluted share. The prior years were $1.78 (2024), $0.30 (2023), and losses in 2022 and 2021. In 2025 the company also booked a $225 million one-time tax gain, which is not cash and is excluded. Removing the prepaid-subscription cash is deliberately strict: $123 million of 2025 operating cash flow came from deferred revenue. Why we use 2025 as the starting level. Revenue has roughly doubled since 2023 and the business was loss-making before that, so earlier years describe a much smaller and different company. Results for the first half of 2026 support the level: revenue rose 22% to $590 million and operating cash flow rose 22% to $239 million. Two other details point to a slowdown: second-quarter revenue growth was 18%, and the increase in deferred revenue was only $8.9 million in the first half of 2026 against $58.3 million a year earlier, while net income was $76.6 million versus $79.9 million. Growth check. Owner earnings grew about 112% in 2025, and we assume 15% for the first year of the base case, well below that. The 15% is anchored on results already reported for the first half of 2026, not on a hoped-for reacceleration. Base case, about $79 per share. Owner earnings grow 15%, 10%, 8%, 6%, 5%, 4%, 3%, 2% and 1% over the next nine years and are then flat in year ten. We credit this because the subscriber base has grown every year and half of the year's growth is already in the reported figures. It is a slowing path because growth relying on new products, new markets or wider profit margins is not something we can point to today. After year ten we let earnings drift down 3% a year rather than assume they last forever, since a consumer app's staying power beyond a decade is not yet demonstrated. Bear case, about $25 per share. Owner earnings fall about 20% in the first year, then decline 5% a year for a decade and 8% a year after that. The reasons are in the filings: the sharp slowdown in deferred revenue, slower revenue growth, and net income that fell in the first half of 2026 despite higher sales. The company also depends almost entirely on the Apple App Store and Google Play, which take 15-30% of subscription payments and can end distribution agreements on 30 days' notice. Stock-based pay of $137 million a year is a large ongoing cost. Bull case, about $122 per share. The company's own growth engine keeps working: owner earnings rise 22% in year one (matching the first-half 2026 growth already reported), then 18%, 15%, 12%, 10%, 8%, 6%, 5%, 4% and 3%, before a 2% annual drift down. With only about 9% of monthly users paying, there is room for the current record of subscriber growth to continue. What would make this wrong. Growth could stall faster than we assume if engagement or paid conversion falls, which the deferred-revenue slowdown may be an early sign of. Alternatively, growth could stay near 20% for longer than our bull case. We could not fully reconcile why operating cash flow rose 22% in the first half of 2026 while net income fell slightly and the deferred-revenue build slowed; the interim filings do not give enough detail to explain it, so we did not raise the earnings level. No debt is subtracted or cash added: owner earnings are already after interest, and the values above are for the equity directly.