Jarden Wealth Weekly: Eye-popping amounts of money in the AI race

Larry Ellison’s company Oracle surprised the market with a bullish outlook driven by AI demand for its data centres.
Felix Fok, CFA, is director of global equities at Jarden Wealth.
OPINION
The tech sector accounts for 31% of the US market, and with the launch of ChatGPT has led global equities higher since the fourth quarter of 2022.
Yet amid the highly anticipated jumbo US Federal Reserve rate cut and more volatile markets since mid-July, the performance of shares in the tech sector has been mixed.
In the three months to September 23, the broader global share market index, MSCI All-Country World (up 5.0%), outperformed the US tech-centric Nasdaq Composite (up 1.7%), a reversal in the trend that has otherwise favoured the higher growth Nasdaq this year.
Within technology, semiconductors, led by Nvidia with a staggering market capitalisation of US$2.85 trillion ($4.56t), fared worse.
What might be going on here? Some plausible reasons are:
Meanwhile, the outperformance by the utilities and real estate sectors supports this narrative as these rely on financing and would see relief from lower borrowing costs.
The weak earnings outlook of global industrial companies Mercedes-Benz Group and FedEx, as well as the moderation of oil prices, suggest weakness in the global economy.
These market dynamics are relevant to the tech sector performance because tech companies are exposed to both the economic cycle and structural trends, and the shares are typically more volatile than the market.
However, we shouldn’t let the tech sector’s current market performance distract from notable progress in the innovation-driven sector that will underpin long-term investments and financial performance.
OpenAI, the company behind ChatGPT, is raising US$6.5 billion ($10.5b) in new funding at a private market valuation of US$150b, up 74% since only February this year.
The capital raise comes as the company announced an update to its gen-AI model, GPT-o1 (codename “Strawberry”).
Earlier gen-AI models could understand and output language well but lacked reasoning skills.
They were viewed as a less grounded form of internet search due to variability in their answers and made-up assumptions.
The latest version, the firm claims, “performs similarly to PhD students” across a range of tasks such as physics, maths and coding.
Like internet search engines, gen-AI models improve with usage, and the model that receives the most queries and feedback accrues a competitive advantage, hence the race.
When ChatGPT first launched, subscribers reached 100 million within the first two months. The number then fell back as early subscribers churned off. Still, subsequent updates have improved capabilities substantially, and the professional cohort is increasingly embedding gen-AI tools into everyday work.
OpenAI disclosed in August it has 200 million weekly active users, double the number from last November.
Apple also became a partner of OpenAI this year and will integrate its iOS services closely with ChatGPT.
Investors are understandably mindful of the hype cycle. Innovations tend to get overvalued in the short term as expectations run ahead of business fundamentals.
The current AI era is no exception; good stories can become crowded, and OpenAI, the leading AI firm, is likely substantially loss-making.
However, where public market valuation multiples for tech leaders in the dot-com day were 50 to 100 times price-to-earnings (PE), Nvidia currently trades on 32 times, similar to Microsoft and Amazon.com.
And many large tech companies investing in AI and data centres do have substantial cash flows from existing profitable business segments, and the market is rewarding them.
A case in point is Oracle, the business software and cloud computing company.
This month, the company surprised the market with a bullish medium-term sales outlook driven by AI demand for its data centres.
Founder Larry Ellison predicts a few large technology companies will battle it out for AI model “technical supremacy” in the next several years, spending US$100b or more each.
Vested interest aside, the direction of more spending is clear, as other large tech companies have already laid out investment plans worth tens of billions a year each.
Tech companies hold dear an axiom: “Only the paranoid survive.”
To this end, boards and management teams will err on investing as the cost of missing out on the next big thing is perceived to be critical – stand still, and you are going backwards.
While this strategic rationale is powerful, it doesn’t guarantee a return on investment. But it does provide a level of durability in terms of technology roadmap and spending, an example being Meta Platforms’ Reality Labs experiment in virtual reality headsets, which had losses of US$16b last year.
In the long run, profits will matter, but the market is, for now, rewarding AI investments as many see the potential and progress.
Two years after the ChatGPT moment, our adeptness at using gen-AI models has improved and their capabilities have progressed remarkably. With a better understanding of context and reasoning, gen-AI models are helping solve more problems, increasing their value proposition for users.
While the future may not just be about tech companies, it will be about those who make the most of tech.
Jarden Securities Ltd is an NZX firm. A financial advice disclosure statement is available free of charge at jarden.co.nz/our-services/wealth-management/financial-advice-provider-disclosure-statement/.
Full disclaimer available at jarden.co.nz/wealth-sales-and-research-disclaimer.