The Most Obvious AI Company Nobody's Talking About

Article argues Alibaba could shift from a China e-commerce story to an AI infrastructure platform, anchored by its open-source Qwen models and Alibaba Cloud’s ecosystem strategy. It cites Alibaba Cloud revenue growth of 34% in the latest fiscal year ended March 31 and triple-digit growth in AI-related product revenue for the 11th consecutive quarter. Net message: investors may be underappreciating cloud’s higher-recurring-revenue, scaling margins profile, though it flags ongoing challenges from domestic commerce stabilization, high AI capex, and intense competition.

Analysis

The market still prices BABA like a consumer internet name, which means it is applying the wrong margin model and the wrong terminal multiple. If AI-related workloads keep pulling enterprise demand into cloud, the earnings mix shifts from low-visibility retail traffic to higher-recurring infrastructure usage, and that is what can drive a rerating rather than raw headline growth. The important second-order effect is that every developer adoption win in the model layer creates downstream demand for storage, compute, security, and tooling, which can lift utilization and operating leverage even before AI monetization is obvious.

The relative winners are BABA’s cloud stack and, selectively, China’s enterprise software/infrastructure ecosystem; the likely losers are the more consumer-exposed China internet names where capital and management attention stay trapped in subsidy wars. JD and PDD remain the cleaner expressions of domestic spending pressure, so a cloud-led BABA rerating would widen the valuation gap between infrastructure and retail internet. For global AI bellwethers like NVDA, the read-through is mixed: China AI spend is supportive at the margin, but export controls and local chip substitution cap the quality of that demand.

The contrarian risk is that investors may be over-assigning AWS-like economics to a China cloud stack that could face weaker pricing power, heavier capex, and more policy interference. Open-source distribution is a top-of-funnel strategy, not proof of durable monetization; if enterprise adoption is slow or chip access tightens, the flywheel stalls and cloud revenue can still disappoint on ROIC. Near-term, the thesis is tested by the next 1-2 earnings prints; over 6-18 months, what matters is whether cloud growth stays above 25-30% while capex intensity normalizes.

Key Decisions for Investors

• Long BABA on any 1-2 week post-headline pullback; this is a 6-12 month rerating trade, with upside if cloud remains the profit-growth engine and downside limited by current market skepticism. Falsify on a material slowdown in cloud growth or evidence that AI spend is capex-heavy but not accretive.
• Pair trade: long BABA / short JD over the next 3-6 months to isolate AI/cloud optionality versus pure retail exposure. Best case is multiple divergence as BABA is reclassified toward infrastructure; invalidated if China consumer stimulus sharply improves retail fundamentals.
• Alternative relative value: long BABA / short PDD as the cleaner expression of 'cloud over commerce.' This works if the market starts rewarding recurring enterprise revenue over subsidy-driven retail share gains; exit if PDD's growth reaccelerates while BABA cloud margins flatten.
• If options liquidity is sufficient, express via a 9-12 month BABA call spread rather than outright stock to cap policy/geopolitical downside while retaining rerating convexity. This is attractive only if you want to own the narrative shift without full beta exposure.
• Set a watch item, not a trade, on the next cloud disclosure: cloud growth, AI product growth, capex, and FCF conversion. If AI revenue slows materially or capex/depreciation outruns gross profit for 2 consecutive quarters, the 'AI infrastructure platform' story should be de-risked.