top of page

What is a Hyperscaler?

Writer: Oliver McDermott
Oliver McDermott
4 days ago
3 min read

AI has become one of the biggest drivers of investment in financial markets, with technology giants committing hundreds of billions of dollars to developing the infrastructure needed to support its growth.


However, following the release of Q3 earnings and a recent sell-off in technology stocks, attention is shifting from the excitement surrounding AI to the pressures associated with its rapid expansion.


What Are Hyperscalers?


Hyperscalers are companies that operate enormous computing networks and data centres, allowing businesses to access AI services on a massive scale. Companies such as Amazon, Microsoft, Alphabet, Meta and Oracle are investing heavily in infrastructure, purchasing advanced semiconductors and developing the energy capacity needed to operate them.


These investments are known as capital expenditure (CapEx), which refers to money spent on long-term assets expected to generate future economic benefits. CapEx involves significant upfront costs, with companies expecting to recover their investment over several years.


However, AI technology is developing so quickly that expensive equipment could become outdated before companies have fully recovered what they spent on it.


What Have Q3 Earnings Revealed?


Amazon reported that its cloud computing division, Amazon Web Services (AWS), experienced revenue growth of 39.3% year-on-year. However, the company also expects capital expenditure to reach approximately $220 billion in 2026.


Similarly, Alphabet expects to spend between $195 billion and $205 billion in 2026, with further increases anticipated next year.


While these figures reflect the growing demand for AI services, the cost of expanding infrastructure is placing pressure on free cash flow. This is the cash a company has remaining after paying for its operations and capital investments.


Alphabet's Q3 capital expenditure of $51.4 billion contributed to negative free cash flow of $9.1 billion during the quarter.


This highlights an important distinction between profitability and cash. A company can report strong earnings while spending so heavily on infrastructure that it has little/negative cash available.


Is There an AI Bubble?


The speed and scale of AI investment have raised concerns that the industry could be experiencing a bubble, where valuations disconnect from the underlying financial performance of companies.


One significant concern is the growing reliance on debt. Rather than funding expansion entirely through existing cash, hyperscalers are increasingly issuing corporate bonds to finance new projects.


Goldman Sachs estimates that around 35% of hyperscaler capital expenditure could be debt-financed in 2027, with total debt issuance potentially reaching $420 billion.


Borrowing allows companies to expand more quickly without using all their available cash. However, it also creates fixed obligations, as interest must be paid and debt eventually repaid regardless of how successfully the investment performs.


This becomes especially important when bond yields rise. Companies issuing new debt may have to offer higher interest rates to attract investors, increasing their financing costs and putting further pressure on cash flow.


Credit markets are already showing signs of caution. Credit default swaps (CDS), which provide protection against a borrower failing to repay its debt, have become more expensive for several hyperscalers. In October, Oracle's five-year CDS spread reached a record 288 basis points, indicating that investors were demanding more compensation for taking on its credit risk.


These concerns have also contributed to volatility in technology stocks, as investors reassess the high valuations attached to companies leading the AI expansion. 

However, today's largest hyperscalers already operate highly profitable businesses, so the 'hype' is also very warranted.


Looking Ahead


Investment in AI infrastructure is expected to continue growing throughout 2027, as hyperscalers compete to expand.


For investors, the focus will increasingly be on how efficiently companies use this infrastructure, whether demand continues to grow and how comfortably they can manage their borrowing.


Strong earnings and cash generation could help support further expansion. However, any slowdown in demand, combined with rising financing costs, could make investors wary, and push prices down.


Ultimately, AI has already transformed investment across AI and Tech. The next challenge for hyperscalers will be turning this unprecedented spending into sustainable financial growth.

Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
bottom of page