Long-Term Thesis
Long-Term Thesis — Dingdong (Cayman) Limited
1. Long-Term Thesis in One Page
The honest 5-to-10-year thesis is that there isn't a conventional one — DDL has agreed to sell the only asset that could compound. The 5-Feb-2026 sale of the entire China operating business to Meituan for up to US$717M (¥4,992M) plus a binding 27-March-2026 AGM resolution committing ≥90% of proceeds to buybacks and/or dividends converts what was a fresh-grocery operator into a Cayman holding company with a cash claim, a small loss-making overseas stub, and a founder-controlled capital-allocation discretion bounded by a 5-year non-compete. The long-term case "works" only in one of two narrow ways: (a) the 90% return mandate is executed at or above ¥21.50/ADS net realized cash within 12 months of close and the residual is wound down or returned on a similar discipline, producing an annualized return that competes with the risk-free rate on a probability-weighted basis; or (b) the deal breaks, the standalone YRD frontline-warehouse moat widens against platform pressure (which it has not done in the last four years), and a re-rated standalone DDL compounds toward ¥30/ADS over five years. Base case sits closer to (a). The compounder version is dead.
Thesis strength
Durability
Reinvestment runway
Evidence confidence
This tab does not underwrite a compounder. Five-to-ten-year owner-value creation requires either (i) a closed deal + disciplined capital return + minimal cash leakage, in which case the equity becomes a short-duration cash-distribution claim, or (ii) a broken deal + structural standalone improvement, which the evidence (gross-margin compression 30.9%→29.2%, AOV ¥72.1→¥70.1, third-place Shanghai share against scale-flywheel rivals) does not support. The single biggest decision is whether Chairman Liang — who controls 68.6% of votes — executes the 90% mandate on the timing the AGM resolution implies, or interprets it loosely.
2. The 5-to-10-Year Underwriting Map
The drivers below are the things that must be true over the next 5–10 years for DDL equity to earn an acceptable return for an underwriter who owns it today. Most cluster in years 1–3 because the deal-event structure foreshortens the investment horizon; the durable-business drivers (#5–#6) only matter if the deal breaks.
The driver that matters most is #2 — execution of the 90% capital-return mandate at a specific per-ADS amount within 12 months of close. Driver #1 (the deal closing) is structurally bounded by SAMR and is partly out of management's hands; driver #2 is fully within board discretion and is the single largest determinant of whether minority ADS holders capture the ¥21.50/ADS implied realised cash that the bull case prices. Drivers #5 and #6 are tail scenarios — they only matter if the deal breaks or HoldCo retains far more than 10% of proceeds, and neither is the base case.
3. Compounding Path
There is no compounding path in the conventional sense. Owner-value creation is best modelled as a probability-weighted cash-distribution stream over 1-3 years, not a multi-year ROIC × reinvestment-rate calculation. The table below traces what a long-term holder actually receives across a 5-year horizon under three scenarios; the chart that follows shows the historical revenue, FCF and operating-margin trajectory that informs the deal-break ("Bear") scenario.
The probability-weighted 5-year total return is approximately ¥20.30/ADS against today's ~¥17.30/ADS — an annualized return of roughly 3-4%, modestly below a comparable Hong Kong dollar / RMB risk-free rate over the same horizon. The asymmetric thinking is that the bear case (–10% IRR) is real but capped by the going-concern floor, while the tail case (founder reroute) is the genuinely structural downside. A long-term holder is being paid Treasury-like returns to underwrite an equity-like tail.
The underlying business shape — revenue plateau at ¥24B, FCF margin of 1.5%, operating margin of 0.5% — is what supports the deal-break floor at ~¥10/ADS. A 5-10 year holder of the standalone business with no acquirer would be underwriting roughly ¥360M of FCF per year at a 0.16x P/S multiple. The business is profitable but not compounding. Reinvestment runway is structurally capped at <1% of revenue capex — there is no margin engine to compound on the existing base.
4. Durability and Moat Tests
Four tests separate the durable thesis from the contingent one — two competitive, two financial. Each carries a 1-3 year resolution window because the deal-event collapses what would otherwise be 5-10 year tests into the closing horizon.
Test #3 is the only financial test where the long-term thesis is genuinely active. Tests #1 and #2 either become irrelevant post-close (the moat passes to Meituan with the asset) or describe a tail-scenario that contradicts management's own strategic answer. Test #4 is the closest thing to a true 5-10 year durability question for the residual HoldCo.
5. Management and Capital Allocation Over a Cycle
Founder Changlin Liang's track record over a single 9-year cycle is the relevant evidence base, and the cycle has just closed. Three pieces of that record matter for the long-term thesis: (1) he absorbed ¥10B+ of cumulative pre-IPO and post-IPO losses, then executed a credible turnaround that produced 14 consecutive non-GAAP profitable quarters and a sustained net-cash position by FY25YE (¥1.5B); (2) he agreed to sell the China business rather than try to compete with subsidy-funded platform rivals — a strategically rational decision but one that shifted the long-term thesis from "operating compounder" to "capital-return claim"; and (3) he retained 68.6% voting control via Class B super-shares while stepping into the Chairman role, leaving the principal capital-allocation discretion concentrated in his hands during exactly the window when minority holders need that discretion exercised cleanly.
The transition mechanics support a runoff posture, not a redeployment posture. New CEO Song Wang (effective 4-Mar-2026) is a 17-year retail-finance veteran (Lianhua Supermarket CFO, Ele.me finance director, head of finance at Hema Fresh) — finance-trained, not strategy-trained. The CTO seat is vacant. Cash compensation across all executives totals ¥22.8M annually (~US$3.3M) — modest for an NYSE-listed ¥24B-revenue operator and not engineered around non-GAAP EPS or aggressive growth targets. Equity-incentive grants to the new CEO are at zero-strike prices (¥0 strike), which biases him toward returning rather than deploying capital. The audit committee remains independent.
The pattern is consistent: Liang has made hard, rational, sometimes painful capital-allocation decisions throughout the cycle, and the new CEO carries no specific reason to deviate from that pattern in the post-close runoff. The thing that should worry a long-term holder is not the historical record but the structural asymmetry that remains: Class B super-voting shares give Liang discretion over mechanism, and there is no contractual sunset on that discretion. The 90% AGM resolution binds magnitude but not method. That is the cycle-end question that matters.
6. Failure Modes
The severe failure mode is #1 — founder discretion under Class B. Every other failure mode (SAMR escalation, overseas burn, withholding, deal break, cash hoarding) is either bounded by the AGM resolution or by external regulatory mechanics. Failure mode #1 is bounded only by Liang's interpretation of his fiduciary duty to minority ADS holders. The next 12-18 months of mechanism disclosure are the test.
7. What To Watch Over Years, Not Just Quarters
The long-term thesis changes most if Chairman Liang discloses, within 90 days of SAMR clearance, a specific per-ADS buyback or dividend amount that implies ≥¥21.50/ADS net realised cash within 12 months of close — that single signal converts an ambiguous Cayman-shell capital-return claim into a defined cash-distribution security, and at today's ¥17.30/ADS would price a 24% gain in the validating direction or a 30%+ decline in the refuting one.