The warning
signs are becoming increasingly difficult to ignore. More than US$2 trillion in
market value is set to disappear from the technology-heavy Nasdaq 100 in
October, which has already declined over 5.8%. Investors are beginning to
question whether the enormous sums being poured into artificial intelligence
(AI) infrastructure can generate adequate returns within a reasonable period.
The problem
extends beyond declining share prices. Big Tech companies are escalating
capital expenditure, building data centres, acquiring expensive computing
equipment and expanding infrastructure at an extraordinary pace. Morgan Stanley
expects AI infrastructure spending to exceed US$1 trillion next year, while
J.P. Morgan projects global AI capital expenditure of US$5.5 trillion by 2030.
What makes
this development particularly worrying is the growing dependence on external
financing. Companies that once financed expansion through substantial operating
cash flows are increasingly approaching bond and equity markets. Meta's
proposed US$97.5 billion fundraising and Microsoft's reported US$25 billion
bond issuance illustrate the scale of financing requirements.
Investors must ask a fundamental question: what
happens if anticipated AI revenues fail to justify these investments?
Data
centres, specialized chips and associated infrastructure involve enormous
upfront costs. Technology also evolves rapidly, creating the possibility that
today's expensive assets could lose economic value before delivering expected
returns.
The
sub-prime mortgage crisis demonstrated how excessive optimism, leverage and
interconnected financial exposures can turn a seemingly contained problem into
a global financial catastrophe.
The AI
investment cycle carries a different risk profile, but its potential
transmission channels are equally concerning.
A sharp correction in technology valuations could
spread into corporate bonds, private credit, banks and institutional investment
portfolios.
There is no
denying AI's transformative potential. However, technological breakthroughs do
not guarantee that every company investing in them will earn attractive
returns. Nor does strong revenue growth automatically justify unlimited capital
expenditure.
My warning to investors is simple: do not become
prisoners of the AI narrative. Examine valuations, debt exposure, cash flows
and realistic returns before committing capital. The danger is not that AI will
fail, but that financial markets may have priced in an exceptionally optimistic
future.
If
expectations collapse, the resulting financial damage could extend far beyond
Silicon Valley. Investors who ignore these warning signs today may discover
tomorrow that the AI bubble was considerably more dangerous than they imagined.
