Alibaba (BABA): The $160 Stock Selling for $120
E-commerce, cloud, and Ant — taken apart piece by piece, none of the math justifies the current price. So what is the market actually afraid of?
The Framework
It’s impossible to value Alibaba as a single company. The stock is four very different businesses bolted together, plus a pile of non-operating assets. What’s left are three pieces that matter: China e-commerce, Alibaba Cloud, and the non-operating asset bridge. AIDC (international e-commerce) and All Others (Freshippo, Amap, Cainiao, Qwen) contribute just ~10% of equity value and don’t require much precision.
China E-Commerce: The Normalized Earnings Question
The China e-commerce segment is by far the largest contributor to Alibaba’s value. It is also the easiest of the three components to understand on a conceptual level. How the segment is valued is straightforward: something on the order of 8x to 10x normalized operating cash flow. The difficulty lies entirely in the numerator.
The headline EBITA numbers for FY2025 and FY2026 tell a clear story: RMB 193 billion and RMB 108 billion. Adjusted EBITA declined by nearly 50% year-over-year. The decline was driven primarily by (i) subsidies and fulfillment costs for instant retail (Taobao Flash Sale + Ele.me), (ii) user experience improvements, and (iii) competitive spending targeted against Pinduoduo and Douyin.
The question, then, is how much of the RMB 86 billion year-over-year drop in EBITA is permanent versus temporary. At a bare minimum, the fulfillment layer supporting Taobao’s instant retail business — riders, warehouses, and everything else that goes into the physical infrastructure required to deliver orders within 30 minutes — is here to stay. It is extremely unlikely that Taobao reverts back to being a pure marketplace. It is a “platform plus fulfillment” hybrid now, and that cost structure is not going away.
Less clear — and likely to normalize — is subsidy intensity, competitive marketing spending, and one-off costs associated with building out Taobao’s new systems. Successfully defending market share against competitors like Pinduoduo and Douyin will require subsidy spending in the near-term, but those costs can come down as growth moderates.
Incrementally generous judgements on each line suggest roughly RMB 30–35 billion can be added back, bringing normalized EBITA to roughly RMB 140 billion, which implies normalized OCF of ~RMB 120–125 billion. At a multiple of 9x (which only implies modest annual growth), e-commerce is worth about RMB 1.1–1.25 trillion.
The single largest unknown is unit economics on instant retail. The company still does not disclose absolute losses, per-order economics, subsidy rates, or fulfillment costs. Every analyst working on Alibaba right now is making a guess. This alone accounts for ~$20/ADS of the stock’s trading range.
Alibaba Cloud: A Wide-Range DCF
Alibaba Cloud simply can’t be valued on current earnings — EBITA margin is only 9%, artificially low given how aggressively Alibaba is investing in this business today. Nor does applying a simple price-to-sales multiple work — the difference between 3x and 6x PS is a 100% valuation swing, and which multiple you pick depends entirely on your growth and margin assumptions. Better to build a two-stage FCFF DCF at a uniform 10% discount rate and make every assumption explicit.
Two variables dominate the outcome: how fast can Alibaba Cloud grow, and what’s the achievable terminal FCFF margin?
On growth: Full-year revenue grew 34% in FY2026, with quarterly growth accelerating to 38% in Q4 alone. AI-driven services revenue grew triple-digit percentages for 11 consecutive quarters and still accounts for just ~30% of Alibaba Cloud’s total revenue. Google Cloud grew revenue 48% last year. Its top-line run rate is ~$70 billion, compared to Alibaba Cloud’s ~$23 billion. Tack on China’s unique structural advantages on the supply side — every measure of annual new power generation capacity adds up to 8x greater than the US, the “East Data West Computing” national infrastructure program has already attracted north of RMB 1 trillion in investment and isn’t showing signs of slowing down, and there is no equivalent to the United States’ 5-to-7-year grid interconnection queue bottleneck.
On margins: a unit economics cross-check reveals that bare-metal gross margins are nearly identical between Chinese and US cloud providers (approximately 74%), consistent with AWS GPU instance rental implied margins. The gap between Alibaba Cloud’s 9% EBITA margin and AWS’s 35% is almost entirely attributable to temporary factors — lower share of high-margin PaaS/SaaS in the revenue mix, R&D costs spread over a smaller revenue base, and a heavy capex cycle inflating depreciation. These factors converge as revenue scales and the product mix shifts. Google Cloud went from ~0% to 24% operating margin in just two years, providing a reference path. A reasonable competition discount (China’s more fragmented market structure, ByteDance’s aggressive pricing, telecom operators cutting into government/enterprise accounts) is roughly 5–10 percentage points, implying terminal operating margins of 22–28% and FCFF margins of 15–20%.
In the conservative scenario (growth decelerating from 28% to 14%, terminal FCFF margin of 12%), the cloud is worth roughly RMB 500 billion. This price implies “Alibaba Cloud is a low-growth IaaS utility with zero AI premium” — requiring AI demand to peak and price wars to persist indefinitely, a joint probability of roughly 15–20%. In the optimistic scenario (growth sustaining at 40–50% for two to three years, terminal FCFF margin of 17–20%), the cloud is worth RMB 1.0–1.5 trillion. The midpoint sits around RMB 750 billion to 1 trillion.
The sheer width of this range (RMB 500 billion to 1.5 trillion, a 3x spread) is itself a statement about how low conviction on cloud valuation remains. The swing is approximately $31/ADS.
The Non-Operating Bridge: Not the Cash Floor You Think It Is
Alibaba has about RMB 640 billion of non-operating assets on its balance sheet at book value. That sounds like a cushy buffer but each of these pieces is problematic.
Approximately RMB 260 billion of net cash is tangible, but converting cash trapped in onshore Chinese subsidiaries into cash at the Cayman holding company incurs withholding taxes (about 5% through the Hong Kong conduit) and FX risk due to foreign exchange controls. More acutely, there’s the ever-present risk of management funneling cash into poor capital allocation decisions — Alibaba just announced a RMB 50 billion instant retail subsidy program while simultaneously announcing a three-year, RMB 380 billion AI capex plan, and the firm has extensive history of low-ROI investments (Ele.me: RMB 200 billion invested over 7 years to achieve 30% market share).
The largest equity-method investment is a 33% stake in Ant Group. Ant posted RMB 38 billion of net income in 2024, but is best thought of as a leveraged financial holding company wrapped in a tech company. As such, it should be valued on price-to-book rather than price-to-earnings. Chinese bank stocks trade around 0.5–0.8x P/B. Far more importantly, Alibaba doesn’t control Ant — the firm can’t dictate dividend policy, capital allocation, or broader strategy.
Other investments (listed equities plus private holdings) total approximately RMB 240 billion at book value. Private investments are carried at cost, given the lack of market prices.
After applying conservative haircuts (net cash at 0.75x, Ant on a P/B basis, private investments at 0.5x, subtracting out contingent liabilities), the bridge is worth roughly RMB 350 billion — about 55 cents on the book-value dollar.
Putting It Together
Summing the three components: the conservative case (e-commerce at 9x normalized OCF plus cloud at RMB 500 billion plus bridge at RMB 350 billion) comes in around $143/ADS. The optimistic case (cloud at RMB 1 trillion) comes in around $174/ADS. The midpoint is ~$160.
AIDC and All Others add approximately $13–14/ADS on a probability-weighted basis, but don’t have great precision and are not drivers of the thesis.
A Utility Stock in Growth-Stock Clothing
Alibaba’s return distribution is strange. The upside is muted (expected value roughly 30–35% above current price). The downside, however, is extremely tightly constrained.
We estimated the price where “one thing goes wrong” — either e-commerce profitability doesn’t recover or cloud doesn’t earn an AI premium, but not both — at approximately $129. The price where “two things go wrong at once” is around $108. These two risks have low correlation: e-commerce profitability is being compressed by the food-delivery war with Meituan and JD, while cloud profitability is being compressed by the price war with ByteDance and Huawei — completely different markets, completely different competitors, completely different economics.
At current prices of roughly $120, the market is pricing somewhere close to “two things going wrong at once.” This means that buying Alibaba near current levels only requires that e-commerce and cloud don’t suffer downturns at the same time for you to make a positive return.
This “limited upside, even more limited downside” distribution is remarkably similar to that of a utility stock. And yet Alibaba’s near-month implied volatility is running at 44–46%, while actual utilities (Duke Energy, Southern Company) trade at 15–18% implied volatility.
If our fundamental analysis is correct — that Alibaba’s true outcome distribution is closer to that of a narrow-range, low-downside utility — then the options market is systematically mispricing Alibaba’s volatility. As a seller of volatility (e.g., covered calls), you earn structural excess returns: the market pays you a growth-stock volatility premium, but you bear utility-stock actual volatility. The edge isn’t in the direction of the stock. It is in the price of the insurance.
Disclaimer: This article is intended for informational and educational purposes only. It should not be viewed as investment advice, a recommendation, or a solicitation to buy or sell any security. The author may hold, acquire, or dispose of positions in the securities mentioned at any time without notice. All estimates, projections, and valuations are based on publicly available information and the author’s own analysis, which may be incomplete, inaccurate, or outdated. Past performance does not guarantee future results. Investing involves risk, including the possible loss of principal. You should consult a qualified financial advisor before making any investment decision. Do your own work.