Patience required in AI development

Transcription

Markets are questioning whether leading tech and cloud firms can make good on their AI investments.

1) Markets question AI

Capital spending by these firms has increased as they race to develop AI.

Some recent research from the investment community questions whether AI revenues alone will catch up with this capital.

We think it will take some time for big tech companies’ revenues to reflect their AI capital spending. AI development will take years – not quarters – to complete.

2) General demand for AI

It’s important to distinguish between some of the companies that are spending big on AI – and the overall macro investment that will likely be in AI. It has room to grow over time.

Nvidia’s Q2 revenue doubles from last year on demand for its chips shows AI capex it is considerable and continuous.

We’ll monitor top AI companies’ revenue growth and adoption trends to gauge whether AI investments are paying off.

3) Our roadmap

Our three-phase roadmap helps us assess the potential economic and market impacts of artificial intelligence.

We’re seeing the first phase unfold now as big tech firms race to invest in data centers.

Early AI winners include big spenders and chipmakers. Beyond this, we see opportunities in firms that provide key inputs such as energy, utilities, materials and real estate.

We remain overweight on AI, but we see signs to change our view.

Investors are debating whether the future revenues of top tech and cloud computing firms could justify investing billions of dollars in artificial intelligence (AI). We believe it is essential to distinguish between individual companies and the wider economy when assessing impact. We are overweight on AI and see winners along the AI ​​supply chain. However, we are watching for indicators to change our view, including stagnant revenue growth or slow adoption of AI.

Big expenses
Major Capital Expenditures by Technology and Cloud Companies, 2007-2024

In recent months, market buzz about the benefits of artificial intelligence has turned to worries that companies investing heavily in it may not see the benefits as quickly. We have assessed this sudden change in sentiment with our portfolio managers. Total capital spending by top tech and cloud players has increased in recent years, particularly in power-hungry data centers, as they race to develop AI. See diagram. Earlier investments ultimately led to an increase in revenue, helping to achieve a return on investment, data from LSEG shows. However, some recent research has questioned whether AI revenues alone will ultimately justify this wave of capital spending on it. When evaluating AI capital by individual companies, investors should consider whether they are making the best use of their balance sheets and capital. But for the economy as a whole, we judge investment in AI by the major revenues that AI could generate across all sectors.

Investments in AI could be compared to capital expenditures for earlier technological innovations such as cloud computing. However, shareholders may not see further investment in artificial intelligence as the best use of corporate balance sheets. We are seeing a disconnect between the short-term lens of some investors and the long-term visions of technology and cloud service providers. This divergence has sparked jitters among investors – but we think patience is called for. Some big AI spenders have allocated investments to building new data centers and exponentially multiplying AI processing power. Such plans take years – not quarters – to complete. So it may take some time for revenues to fully realize AI capital expenditures. Some tech companies are already reporting growing revenue from launching AI-related products.

Large scale AI demand

We see room for general AI capital to drive some of the transformational waves driven by mega-forces or structural forces shaping returns. Nvidia’s second-quarter revenue doubled from a year ago, showing that AI investment is substantial and continuing. Nvidia’s ( NVDA , NVDA:CA ) results highlight how AI development is expanding: More than half of its AI revenue came from non-tech sectors. We track a few indicators to gauge our bullish view on the AI ​​theme. First, we look for signs of stagnant revenue growth at top AI companies, increasing the importance of each earnings season. Second, we assess changes in the still low adoption of AI beyond the technology sector. Third, we look at any slowdown in US growth that could cause big tech companies to cut spending.

We use our three-phase roadmap to track the economic and market impact of AI. The first phase – construction – is now underway as big tech companies race to invest in data centers. Early winners in this phase include big spenders and chipmakers. We also see opportunities in firms that provide key inputs such as energy, utilities and real estate. In the second phase, we believe AI adoption will expand into sectors beyond technology, such as healthcare and finance. This could lead to a third phase of broad productivity gains, but the size and impact are uncertain. Our future research will explore this in more detail.

Our conclusion

Investors debate the implications of the AI ​​capex boom. Some investors have reduced their positions in the technology in recent months, implying room to rebuild holdings. We remain overweight on AI, but eye cues to change our view.

Market background

US stocks are back near all-time highs. AI trading regained its footing after volatility in July and early August, even as Nvidia shares stumbled from profit-taking after its Q2 earnings beat expectations. The expansion in corporate earnings growth in the second quarter, along with upbeat jobless claims and last week’s GDP data, show that U.S. economic growth is holding up — a positive for risk assets. We think it is unlikely that the Federal Reserve will cut rates as drastically as the markets are pricing them in.

US payrolls for August are in the spotlight this week. At the Jackson Hole symposium, Federal Reserve Chairman Jerome Powell noted that the central bank is now more focused on any easing of the labor market. A further rise in the unemployment rate helped fuel recession fears last month. However, this was caused by an increase in immigration that increased the labor supply, not layoffs. We expect this week’s data to show signs of sustained job gains, easing lingering growth concerns.