The market is too optimistic about how quickly cloud growth will translate into earnings after AI-development costs. Even with substantially higher cloud profits and narrower AI losses, my twelve-month target is $92.41 per ADS.
Key points
- 22% cloud margin is needed in FY28 to support today's price under my valuation assumptions.
- RMB13.9 billion of AI-lab losses exceeded cloud profits in the June 2026 quarter.
- $92.41 per ADS is my twelve-month target despite substantial earnings recovery.
The call
The market expects too much profit from cloud growth. At my chosen earnings multiple, today's price requires a FY28 cloud operating margin, profit as a share of sales, of approximately 22%. That assumes commerce holds up and AI-development losses narrow. June's cloud margin was approximately 12%.
Where June-quarter operating profits went (RMB millions)Alibaba August 20, 2026 release; adjusted EBITA excludes employee equity costs and certain charges.
I expect recovery, but not enough. My FY28 commerce profit estimate is RMB162 billion, with cloud contributing RMB45 billion. AI labs and other operations absorb RMB52 billion. Cloud capacity can sell quickly while developing models remains expensive; management's server-payback framework does not establish returns after those development costs.
Valuation
| Driver | My estimate |
|---|
| FY28 underlying earnings | RMB105.8 billion |
| FY28 diluted ADS equivalents | 2.53 billion |
| Forecast currency conversion | RMB6.785 per US dollar |
| FY28 earnings per ADS | $6.16 |
| Forward earnings multiple | 15 times |
| Twelve-month target | $92.41 per ADS |
Underlying earnings deduct employee equity costs and tax but exclude exceptional investment gains and impairments. Fifteen times recognizes cloud expansion while allowing for cash consumption and Chinese ownership and policy risks. The downside warrants a sell, although faster cloud-margin expansion creates meaningful upside risk.
The cloud margin needed to support today's share priceOur FY28 sensitivity, October 8, 2026; all other forecasts fixed; 15 times underlying earnings.
Risks
- Cloud profits could beat my forecast: watch whether revenue growth comes with sustained margin expansion rather than heavier spending.
- Commerce could recover faster: watch merchant advertising revenue excluding the subsidy-accounting change, alongside delivery losses.
- Financing and currency could change the outcome: watch new share issuance, unrestricted liquidity and renminbi movements; higher borrowing costs could compound cash consumption.
What would change my mind
I would reconsider if cloud margins exceed 20% for two consecutive reported quarters while AI-lab losses fall below RMB8 billion per quarter and group cash flow after infrastructure spending turns positive.