Four major U.S. technology enterprises have jointly forecast capital expenditures soaring to approximately $650 billion (RM2.57 trillion) by 2026.
This staggering sum is allocated for the establishment of new data centers and the requisite array of equipment, encompassing artificial intelligence (AI) chips, networking infrastructure, and backup generators.
The planned expenditures, attributed to Alphabet Inc, Amazon.com Inc, Meta Platforms Inc, and Microsoft Corp, reflect an unprecedented surge as these companies vie for supremacy in the burgeoning AI tools market.
According to Bloomberg data, each organization’s projected spending for this year would achieve a historic peak in capital disbursements for any single corporation over the past decade.
Finding a parallel to such ambitious spending forecasts entails looking back to the telecommunications bubble of the 1990s, and possibly even the extensive railway network expansions of the 19th century or post-WWII federal infrastructure investments, not to mention the New Deal-era relief initiatives.
This substantial increase—an estimated 60% rise compared to the previous year—signals yet another acceleration in the global data center construction wave.
The rush to develop these expansive facilities, housing rows of operational servers powered by costly processors, has strained energy supplies, heightened concerns about inflated prices for other users, and sparked disputes between developers and communities wary of competing demands for power and water resources.
Additionally, the concentration of construction expenditures by a select few wealthy corporations, which already represent an increasing segment of U.S. economic activity, may distort overarching economic data.
According to Gil Luria, an analyst at DA Davidson, the four companies perceive the competition for AI computational power as a potential winner-takes-all scenario. “None of them is willing to lose,” he emphasized.
Last week, Meta announced that its capital expenditures for the year could rise to as high as $135 billion—a staggering increase of approximately 87%.
On the same day, Microsoft reported a 66% jump in its second-quarter capital spending, surpassing estimates, and analysts predict it will allocate nearly $105 billion for the fiscal year concluding in June. This news catalyzed the second-largest single-day market value decline for any stock.
Alphabet, which began its journey in a garage south of San Francisco in 1998, startled investors on Wednesday by revealing a capital spending forecast that not only exceeded analyst predictions but also surpassed a substantial portion of U.S. industrial expenditures, with plans to invest up to $185 billion.
Amazon followed suit on Thursday with a projected $200 billion in capex for 2026, also causing its shares to plummet during extended trading.
In stark contrast, the largest U.S.-based automakers, construction equipment manufacturers, railroads, defense contractors, wireless carriers, parcel delivery companies, along with Exxon Mobil Corp, Intel Corp, Walmart Inc, and various spinoffs of General Electric—21 firms in total—are anticipated to allocate a combined $180 billion in 2026, as estimated by Bloomberg.
Each tech titan has crafted a distinct strategy to recoup its investments, but all share a fundamental belief: that AI tools, like OpenAI’s ChatGPT and competitors capable of emulating human reasoning, will play an increasingly pivotal role in both professional and personal spheres.
The development of sophisticated software models driving this transformation is a remarkably capital-intensive endeavor, necessitating the assembly of thousands of chips, each costing tens of thousands of dollars.
Thus, the astronomical expenses accumulate. There is also the underlying assumption that these investments will yield exponentially greater future revenues.
These outlays are reshaping companies that, not long ago, possessed a relatively modest physical presence, even as their digital services reached billions globally.
Throughout most of their histories, both Meta and Alphabet (Google’s parent) recorded their plush corporate campuses and office spaces as significant elements of their tangible assets, with most expenditures directed at salaries and stock options for engineers and sales personnel.
That paradigm is shifting. Last year, Meta’s spending on capital projects eclipsed its research and development costs—representing engineers’ salaries—for the first time in six years.
By year-end, the parent company of Facebook and Instagram held $176 billion in property and equipment, roughly five times the amount recorded at the close of 2019.
As these figures escalate, uncertainty persists regarding the capability of these corporations to fulfill their ambitious aspirations.
With the ramp-up in data center construction, they are grappling for limited resources, including electricians, cement trucks, and chips from Nvidia Corp, produced in Taiwan Semiconductor Manufacturing Co (TSMC) facilities. “Bottlenecks are imminent and even ongoing,” cautioned Luria.
An additional concern revolves around the financial viability of these expenditures. Meta and Google, primarily profiting from digital advertising, alongside Amazon—the leading online retailer and cloud services provider—and Microsoft, the largest business software vendor, each hold substantial market positions and liquid assets.
Their readiness to invest significant amounts of cash into a future dominated by AI will test both these reserves and the patience of investors.
As Tomasz Tunguz, an investor at Theory Ventures and former Google employee, noted, “These companies were once cash-generating machines. Now, suddenly, they require that cash, and more, leading them to borrow.”
Tunguz, who previously compared the AI surge to historical investment panics in a blog post last year, remarked, “Such frenzies do not always end positively. However, during their ascent, they serve as substantial catalysts for the economy.”
Notably, investor enthusiasm for tech giants’ stocks, which surged over the past year, is now tempered in light of the burgeoning capital expenditures.
In certain cases, shares have been sold despite steady performance in core business areas—ranging from online advertising, web search, and e-commerce to productivity software—exceeding revenue expectations.
“What’s causing the trepidation? Undoubtedly, the analysts’ narrative and their rhetoric regarding the speed at which AI will alter business landscapes,” remarked Steve Lucas, CEO of Boomi, a company that assists other firms in integrating their data and software.
“I wouldn’t contest the potential of AI,” he emphasized. “What I would challenge is the timeline, as well as the economic implications.”
Source link: Theedgemalaysia.com.
