MongoDB, Inc. (MDB) stock intrinsic value fair value estimate & owner earnings analysis

MongoDB sells a document database used by developers to build applications. Most revenue now comes from Atlas, its cloud-hosted service billed largely on consumption, with the rest from subscriptions to its self-managed software. Revenue was $2.46 billion in the year ended January 2026, up 23%, and the six months to July 2026 came in at $1.46 billion, up about 28%. Contracted future revenue not yet recognized was $1.47 billion at January 2026, about half of it due within a year. Existing customers spent about 21% more at year end than a year earlier (net ARR expansion rate 121%), and that is the main source of growth we credit. How we measure earnings. We start from operating cash flow, remove swings in receivables, deferred revenue and commissions, and then subtract stock-based compensation as a real cost, because it is paid in shares that dilute owners. That cost was $550 million last year, 22% of revenue. Capital spending is tiny (about $5 million). On this basis MongoDB has never earned positive owner earnings in the years we can measure: about -$3.52 per share in the year ended January 2021, -$4.40, -$4.85, -$2.79, -$2.44, and -$0.69 in the year ended January 2026. The trend is clearly improving: the loss narrowed by roughly $126 million last year alone. Cash flow before stock compensation reached $495 million. Interest earned on roughly $2 billion or more of cash and investments is part of that cash flow, so it flatters the operating picture somewhat. Why the earnings level moves. Full-year results alone would put value at or below zero. But the two most recent quarters (both profitable on a net income basis, $4.4 million and $40.9 million) and first-half operating cash flow of $343.5 million against stock compensation of $286.8 million show a level of roughly +$50 million of owner earnings for six months, from a seasonally weaker half (the same half a year earlier was about -$90 million). We therefore use a level of about 3% of revenue, in line with what has actually been delivered, and do not assume it rises in the base case. Base case, about $26 per share. Revenue grows 25% this year (first half already 28%), then fades through 18%, 15%, 13%, 11%, 9%, 8%, 6%, 5% and 4% by the tenth year. We credit this because it is the continuing output of existing-customer expansion above 120%, which held even through the slowdown of 2023 and 2024, and because contracted backlog covers part of it. Owner earnings stay at about 3.2% of revenue, growing from roughly $100 million to about $230 million, and then hold flat for a further ten years because the database is deeply embedded in customers' applications; after that we assume nothing. Year-one growth of owner earnings assumed is well above the last full year's (-$0.69 per share, itself an improvement), and the reason is the reported first-half profit improvement, not a forecast. We left out any margin expansion, new products such as AI-related workloads, and share gains from rivals, because none is secured. Bear case, about $1 per share. Atlas consumption slows the way it did in the year ended January 2024, when deferred revenue fell $82 million and growth dropped sharply; cloud providers with competing database services (Amazon, Microsoft) pressure pricing and growth fades to low single digits after this year. Stock compensation stays near a quarter of revenue, since $1.0 billion of unrecognized awards is already committed, so owner earnings return to about breakeven and the interest income on cash gradually shrinks. Only this year's first-half profit is left. The company has no significant debt (its convertible notes were converted into shares), so leverage is not the risk here; profitability is. Bull case, about $120 per share. Growth stays near the recent 28-30% pace for a couple of years (24%, 21%, 18%, 15%...), and the operating leverage MongoDB has shown for five straight years continues: owner earnings margin rises about three points a year to roughly 14%, with a decade of flat earnings afterward. This depends on stock compensation not growing as fast as revenue, which is not guaranteed. What would make this wrong. The value is highly sensitive to stock compensation: each point of revenue it consumes is roughly $30 million a year, and our base case value moves by several dollars per share for each point. If second-half results show owner earnings slipping back below zero, our base case is too high. Conversely, sustained margin gains would make even the bull case look modest. Two things could not be fully reconciled and are handled by judgement: the treatment of lease payments (left in operating cash flow) and how much of the interest income on cash will persist as cash is used for buybacks (the company repurchased $400 million of stock in the year ended January 2026). Because the range of outcomes is wide and depends on a margin that has not yet been sustained for a full year, the margin-of-safety price is set at roughly half of the base value and assumes confirmation that a full year of positive owner earnings is delivered.