Weekly Investment Update (11/07/2025)

U.S. equities retreated moderately during the week. Several developments dampened investor enthusiasm, including weaker labor market data and renewed concerns about artificial intelligence-related capital expenditures (capex).

The October report from the job placement firm Challenger, Gray, and Christmas showed job-cut announcements surged to over 153K, a 180% increase from the previous month. However, it is worth noting that monthly jobs reports can be both volatile and contradictory. For example, the ADP jobs report showed a gain of 42K jobs in October, a strong increase from September. Nevertheless, the aggregate data seem to suggest that the U.S. labor market is continuing to soften, reinforcing our view that the Federal Reserve will cut interest rates in December despite last week’s FOMC statement being interpreted as being rather hawkish.

Regarding AI capex, notably, OpenAI CFO Sarah Friar stated Tuesday that the company would welcome government investment to help finance its data-center buildout. Although CEO Sam Altman later clarified that the maker of ChatGPT does not need government help, recently there has been more investor attention on the rise in debt financing among the hyperscalers. Our investment team continues to monitor how the debt situations of these companies evolve, but it is important to note that current leverage levels are still low, with aggregate debt among Amazon, Microsoft, Alphabet, and Meta making up approximately 2% of their combined market cap.

Continued AI Capex Spending Plans Draw Scrutiny Over Capital Discipline

What is happening: The third-quarter earnings season revealed an acceleration in capital spending among the world’s largest cloud providers, with Meta, Alphabet, Amazon, Microsoft, and Oracle now expected to invest nearly $400 billion in AI-related projects in 2025. This figure is projected to rise to $540 billion in 2026 and $615 billion in 2027 — totaling $1.5 trillion over three years, or 17% of all U.S. corporate capex in 2021, before the AI boom. In the most recent quarter alone, Google, Meta, and Microsoft spent close to $80 billion on AI infrastructure. Microsoft’s quarterly capex surged 74% year-over-year to $35 billion, with plans to double its data-center footprint and spend up to $140 billion next year. CEO Satya Nadella described the initiative as building “planet-scale” infrastructure.

To finance this spending surge, the tech giants are increasingly relying on debt rather than profits. U.S. technology firms have issued over $200 billion in corporate bonds this year, reversing a decade-long trend of self-funding. Meta recently announced plans to sell $30 billion of bonds — the largest investment-grade bond offering of the year — and has additionally raised $27 billion in private debt from institutional lenders including Pimco and Apollo to fund its 5GW Hyperion data center in Louisiana. Oracle sold $18 billion in bonds in September, underlining a broader shift toward debt markets to support the AI buildout.

Why it matters: Recent debt-financed spending plans mark a new phase in the AI buildout, inviting comparisons to the dot-com era, though with important differences. Unlike the speculative startups of the early 2000s, today’s AI investments are being driven by profitable, well-established firms with strong balance sheets. Still, the pace and scale of this spending have raised concerns about overreach and overly optimistic expectations, especially as much of the capital is being committed before clear revenue streams have materialized.

Investor responses reflect increasing selectivity. Alphabet’s shares rose 3% after the company raised its 2025 capital expenditure forecast by $8 billion to $93 billion, signaling market confidence in its ability to generate returns from AI. In contrast, Meta’s shares fell 12%, erasing $240 billion in market value, after it indicated AI-related spending could exceed $100 billion next year. Meta stands out among the hyperscalers as the only firm increasing investment faster than its free cash flow, raising questions about capital discipline and return on investment.

One important difference from the dot-com era is how financial risk is being distributed. By turning to bond markets, tech giants are starting to share the funding burden with a broader base of investors, particularly institutional debt holders. This helps ease pressure on equity valuations but also embeds AI-related risk more deeply across financial markets. At Bessemer, we continue to engage with the AI theme selectively while remaining mindful of the growing trend of debt financing and are adjusting positions accordingly. Our portfolios maintain a focus on valuation discipline, as seen with the addition to Alphabet earlier in the year when it was trading at a discount to the broader market.

Overall Consumer Strength Masks a Deepening Divide 

What is happening: The U.S. economy has displayed resilience this year despite several headwinds. While much of the focus has been on AI and technology-related investment driving growth, consumers remain a fundamental part of the story. Consumption accounts for roughly 70% of U.S. GDP, continuing to serve as the main engine of economic growth. Despite signs of labor market softness, overall spending has remained resilient, though it continues to show bifurcation between higher- and lower-income consumers. Notably, the University of Michigan’s Consumer Sentiment Index fell to its lowest level on record today. However, consumers in the top third of stock ownership reported an 11% surge in sentiment, underscoring the growing divide across income and wealth segments.

Third-quarter earnings from major U.S. banks, including JPMorgan and Bank of America, indicated that consumer spending and credit trends remain stable. At the same time, many consumer companies emphasized a widening divergence between higher- and lower- income customers. McDonald’s CEO Chris Kempczinski said customer traffic for lower-income customers is down double digits, while Chipotle CEO Scott Boatwright noted a meaningful pullback from customers in their late 20s to 30s. The commentary aligns with data this week showing the share of U.S. consumer debt delinquency rose in the third quarter to the highest level since 2020, with transitions into serious delinquency the highest among borrowers in their 20s to 30s.

Why it matters: Despite signs of stress among lower-income cohorts, the high-end consumer remains the primary driver of demand, now accounting for roughly half of all consumption. This dynamic underpins our view that consumer spending will remain resilient but bifurcated, with premium spending offsetting softness elsewhere. In portfolios, we continue to emphasize exposure to luxury and high-end discretionary categories, recently adding LVMH to capture both the durability of affluent spending and early signs of recovery in the broader luxury market.

That said, looking ahead, there are also several tailwinds that we expect will support the low-end consumer. Fiscal measures are estimated to provide upwards of $150 billion in incremental consumer aid. This will come through the 2026 tax-refund season and include expanded deductions, no tax on tips, and other provisions. Additionally, the Federal Reserve remains focused on preempting further labor market weakness and is likely to continue easing interest rates, in our view, supporting consumers through lower borrowing costs and improved access to credit.