Three years have passed since OpenAI ignited a surge of enthusiasm in artificial intelligence through the launch of ChatGPT. Despite the ongoing influx of investments, there is increasing skepticism about the sustainability of this momentum.
The recent selloff of Nvidia's shares, Oracle's downturn after revealing escalating AI-related costs, and growing pessimism around companies linked to OpenAI are signs of skepticism. As 2026 approaches, investors are contemplating whether to reduce their AI exposure in anticipation of an impending bubble or to double down to leverage this transformative technology.
"We've reached the juncture where reality meets expectations," said Jim Morrow, CEO of Callodine Capital Management. "It's been an impressive narrative, but we're now seeing if the investments will yield fruitful returns."
The concern around AI investments revolves around its applications, the significant development expenses, and consumer willingness to pay for these services. These elements are critical for the stock market's future.
The S&P 500's $30 trillion bull run over three years has largely been fueled by major tech firms like Alphabet and Microsoft, along with companies benefiting from AI infrastructure investments, such as Nvidia and Broadcom, and energy providers like Constellation Energy. A halt in their growth would impact equity indexes.
"These stocks don't adjust due to declining growth rates," remarked Sameer Bhasin, principal at Value Point Capital. "They adjust when growth rates cease to accelerate further."
Nevertheless, optimism persists. The tech titans responsible for significant AI spending have abundant resources and are committed to continued investment. Moreover, AI service developers like Alphabet's Google are progressing with new models, maintaining the debate.
Key trends to watch as this narrative unfolds include:
Access to Capital
OpenAI plans to spend $1.4 trillion in the coming years. However, despite being the world's most valuable startup as of October, it's generating far less revenue than its operational expenses. It's projected to burn $115 billion through 2029 before turning profitable in 2030, The Information noted in September.
Fundraising has not been an obstacle so far, with $40 billion already secured from SoftBank Group and others earlier this year. Nvidia committed to a $100 billion investment in September, raising concerns about circular financing within the AI sector.
If investors hesitate to provide further capital, OpenAI could face challenges, affecting surrounding companies like CoreWeave, a computing services provider.
"A huge amount of capital — now in the trillions — is concentrated on a few themes and names. The slightest hint of short-term issues or overstretched valuations will trigger simultaneous withdrawals," commented Eric Clark, portfolio manager at the Rational Dynamic Brands Fund.
Other companies also rely on external funding for their AI initiatives. Oracle's stock soared amid bookings for cloud services, yet the construction of data centers entails significant cash needs, addressed through substantial bond issuance. Debt introduces pressure, as bondholders demand timely cash payments, unlike equity investors who benefit mainly from stock price appreciation.
Oracle's shares plunged Thursday following higher-than-expected capital expenditures reported in its fiscal Q2, coupled with cloud sales growth missing analyst estimates. Additionally, delays in some OpenAI-related data center projects sent Oracle shares lower, affecting other AI-infrastructure-related stocks. Meanwhile, a measure of Oracle's credit risk hit a high not seen since 2009.
An Oracle spokesperson stated their confidence in meeting obligations and expansion plans.
"Credit analysts tend to be more prudent than equity analysts, focusing on getting their money back," said Kim Forrest, CIO at Bokeh Capital Partners.
Big Tech Spending
Forecasts indicate Alphabet, Microsoft, Amazon, and Meta Platforms will collectively spend over $400 billion on capital expenditures in the next year, mostly on data centers. Though these firms are gaining AI-related revenue from cloud computing and advertising, it's dwarfed by their costs.
"Declining growth projections or deceleration could drive market concerns," said Michael O’Rourke, chief market strategist at Jonestrading.
The Magnificent Seven tech giants, including Apple Inc., Nvidia, and Tesla Inc., anticipate 18% earnings growth in 2026, the slowest in four years, according to Bloomberg Intelligence data, slightly surpassing the S&P 500.
Rising depreciation expenses from data center investments are concerning. Alphabet, Microsoft, and Meta incurred around $10 billion in depreciation in Q4 2023, escalating to nearly $22 billion in the September-ended quarter, with an expectation of about $30 billion by next year.
This trend could restrict buybacks and dividends, which return cash to shareholders. By 2026, Meta and Microsoft might have negative free cash flow, while Alphabet is expected to break even, considering shareholder returns, according to Bloomberg Intelligence.
A major concern is the strategic shift it implies. Big Tech previously thrived on rapid revenue growth at low costs, leading to significant free cash flow. Their AI plans alter this dynamic.
"If we leverage up to build out AI hoping to monetize it, multiples will contract," stated Jonestrading's O’Rourke. "If it doesn't materialize, this pivot proves to be a severe misstep."
Rational Exuberance
Although Big Tech's valuations are elevated, they're not as extreme as past market euphorias. While comparisons with the dot-com bust arise, AI gains pale in comparison with those of internet development. The Nasdaq 100 Index, tech-heavy, trades at 26 times projected profits, per Bloomberg data. This contrasts sharply with over-80-times projections during the dot-com bubble peak.
Dot-com era valuations were excessively high due to soaring stock prices, yet companies were smaller and less profitable, if profitable at all.
"These aren’t dot-com multiples," said Tony DeSpirito, BlackRock's global CIO and portfolio manager of fundamental equities. "Pockets of speculation or irrational exuberance exist, but they're not significantly present in AI, especially within the Mag 7."
Palantir Technologies Inc., trading at over 180 times estimated profits, and Snowflake Inc., nearly 140 times, exemplify AI stocks with lofty valuations. However, Nvidia, Alphabet, and Microsoft remain below 30 times, subdued amid the prevailing euphoria.
This leaves investors at a crossroads. Risks are tangible despite continued AI investments, yet many companies aren't panic-priced currently. Which path AI trades take next is uncertain.
"This consensus thinking will likely crack," predicted Bhasin of Value Point. "While a collapse akin to 2000 is unlikely, a rotation is anticipated."
Comments