If you've ever tracked Alibaba's earnings reports, you've seen that capex number jump. It’s not a rounding error. In recent years, Alibaba has consistently plowed tens of billions of dollars into infrastructure—data centers, warehouses, undersea cables, and more. Some analysts call it a spending spree; I call it a deliberate long game. Let's peel back the layers.

What Drives Alibaba’s Capital Expenditure?

Alibaba’s capex isn’t random. It’s tied directly to three strategic pillars: cloud dominance, logistics network, and global expansion. I’ve sat through investor calls where management explicitly said, “We will invest aggressively to capture the cloud opportunity.” And they did.

The breakdown is revealing. Based on public filings (and my own number crunching), cloud infrastructure typically eats up about 45-50% of total capex. Logistics (Cainiao) gets another 25-30%. The rest goes to international e-commerce, local services, and technology acquisitions.

Real investor insight: In quarterly calls, Alibaba’s CFO often highlights that capex intensity (capex as % of revenue) hovers around 8-10%. That’s significantly higher than competitors like Tencent (~5%) but still below AWS (~12%). They’re in the middle, but their absolute spend is massive due to revenue scale.

Cloud Computing: The Biggest Piece of the Pie

Data Centers Across the Globe

Alibaba Cloud (now rebranded as Alibaba Cloud Intelligence) is the primary capex driver. They build data centers in regions like Singapore, Germany, the U.S., and of course, China. Each facility costs hundreds of millions. I visited one of their new centers in Zhangbei, China—it’s powered entirely by renewables. The cooling system alone is a marvel.

But it’s not just about building more. It’s about where they build. Alibaba deliberately placed data centers near renewable energy sources (wind and solar) to reduce long-term electricity costs. That’s a capex decision that pays for itself in opex savings.

Undersea Cable Investments

Something less known: Alibaba has invested in multiple subsea cable systems, including the Asia-Africa-Europe-1 cable. This is pure infrastructure capex to reduce latency for international customers. They own capacity, not just lease it. That’s a smart long-term asset.

Logistics: Cainiao Network’s Infrastructure Buildout

Cainiao started as a data platform, but it’s now a physical monster. Alibaba funds robotics-powered sorting centers, automated warehouses, and last-mile hub upgrades. I toured a Cainiao facility in Hangzhou—the autonomous guided vehicles (AGVs) were moving parcels 24/7. The building itself is leased, but the robots and conveyor systems are capex-heavy.

They’re also building overseas: Cainiao has opened eHub facilities in Liege (Belgium), Kuala Lumpur, and Moscow. Each eHub costs $100M+. The goal is to cut cross-border delivery from 5 days to 3 days. That requires serious spending.

How Does Alibaba’s Capex Compare to Peers?

Let’s put numbers where my mouth is. I aggregated data from annual reports to give you a real comparison. Note: figures are normalized for revenue scale.

CompanyCapex (as % of Revenue)Primary Spend Areas
Alibaba8-10%Cloud, logistics, international
Amazon12-14%Cloud (AWS), fulfillment
Microsoft8-10%Cloud (Azure), AI data centers
Google10-12%Cloud, search infrastructure
Tencent5-7%Video, cloud, gaming

Alibaba’s ratio matches Microsoft, but the composition is different. Microsoft spends more on corporate offices and AI hardware; Alibaba spends more on logistics real estate. That’s because Alibaba is also a commerce company.

The Impact of Capex on Alibaba’s Financials

High capex crushes free cash flow in the short term. Alibaba’s free cash flow often dips after a capex surge. But for a growth company, that’s fine. The important metric is return on invested capital (ROIC). Alibaba’s ROIC for cloud has improved from negative to low single digits recently—still not great, but trending up.

Depreciation is the hidden killer. With billions in data center assets, depreciation eats into net income. But management doesn’t worry because they see the long-term revenue stream from cloud contracts (typically 1-3 year commitments).

One thing many miss: Alibaba often uses co-location agreements (leasing space in third-party data centers) to supplement owned capacity. That keeps reported capex lower than actual investment. If you dig into operating lease disclosures, the total commitment is huge.

Looking ahead, I see three capex themes:

1. AI Infrastructure: Alibaba has announced massive orders for AI chips (both NVIDIA and its own Hanguang series). Building AI clusters requires specialized data centers with high-power GPUs. That’s capex-intensive.

2. International Fulfillment: To challenge Amazon globally, Alibaba needs overseas warehouse expansion. Europe and Southeast Asia are hotspots.

3. Green Energy: Alibaba committed to be carbon neutral by 2030. That means investing in solar farms and renewable energy PPAs directly. Again, capex.

I believe capex will remain elevated at $8-12 billion per year for the next 3-5 years. Any slowdown would signal a shift in strategy—perhaps a mature cash cow mode.

Common Misconceptions About Alibaba Capex

Is Alibaba’s capex too high and unsustainable?
Depends on what you compare. For a company with $130B+ revenue, 10% capex is ahead reasonable. It’s higher than Tencent but lower than Amazon. The real risk is if cloud growth slows and capacity gets underutilized. I don’t see that in next 3 years.
Does Alibaba count software development as capex?
No. Software dev is expensed as R&D. Only physical assets like servers, buildings, and networking gear are capitalized. That’s why you see a surge when they build data centers.
How does Alibaba’s capex affect its stock price?
Short-term, high capex depresses free cash flow, and value investors avoid it. But growth investors like the expansion narrative. I’ve seen two stock jumps after major capex announcements—market perception is improving.
What happens if Alibaba cuts capex tomorrow?
It would likely free up cash for buybacks or dividends. But it would also signal that they are stepping back from cloud-as-a-service leadership. That would be a negative signal for long-term growth. Unlikely under current management.

This analysis is based on public filings, earnings transcripts, and site visits. Fact-checked against Alibaba’s annual reports and industry data.