AI Capex Drives Equity Resilience Amid Rising Interest Rates
This article explores the surprising resilience of equities despite rising interest rates, attributing it to higher corporate earnings, particularly those influenced by AI capital expenditures (capex). It delves into the mechanics of how AI capex impacts financial statements for both the companies investing in AI (builders) and those supplying the components (suppliers), and offers investment strategies in this evolving landscape.
Interest Rates and Equity Resilience
The year has been marked by a significant increase in interest rates. The Treasury bond rate climbed from 4.18% to over 5%. Notably, September saw a rapid surge, with the 10-year T-bond rate increasing by 0.54% from 4.75% to 5.29%. This 0.54% increase is statistically significant, placing it in the top 10% of monthly rate changes since 1962.
Despite these rising rates, US equities have shown remarkable resilience. Aggregate market capitalization increased by 12% to 13% over the first nine months of the year. However, this growth has been uneven across sectors and individual stocks.
Sectoral Performance and Market Disparity
Analyzing market performance by sector reveals significant disparities:
- Best Performing Sectors (2026):
- Energy: +39%
- Technology: +24.5%
- Materials: +21%
- Underperforming Sectors:
- Communication Services
- Consumer Discretionary
- Utilities
It's crucial to note that these aggregate market cap figures only include companies publicly listed at the start of the year (December 31, 2025). This avoids skewing data with new public listings, such as SpaceX, which added $2 trillion to the communication services sector's market cap but wasn't available for purchase at zero at the beginning of the year.
Further analysis shows that even within high-performing sectors like technology, the gains are not evenly distributed. While the technology sector was up 24.5%, only 52% of technology stocks increased in value. Across the entire market, approximately 50% of stocks were up, and 50% were down, indicating a highly unbalanced market where individual portfolio performance could vary significantly from the aggregate.
Implied Equity Risk Premium
To understand market sentiment, the implied equity risk premium (ERP) for the S&P 500 is calculated monthly. This is analogous to a bond's yield to maturity, representing the expected return on stocks given their current price and anticipated cash flows (dividends and buybacks, augmented by analyst forecasts for earnings growth).
At the start of October 2026, the S&P 500's implied expected return was 8.99%. Subtracting the risk-free rate (5.29% T-bond rate) yields an ERP of 3.7%. If one considers the US Treasury itself to have some risk (due to its AAA rating loss), the ERP rises to 3.92%. While this ERP is lower than historical averages, it's not drastically so, especially compared to the 2% ERP at the peak of the dot-com bubble in 1999. The total expected return on stocks has increased throughout the year, but the T-bond rate has risen even more, leading to a lower implied ERP.
The Impact of AI Capex on Financial Statements
The core of the discussion revolves around the massive investments in AI capex, estimated to be around $3 trillion currently, with potential to reach $8-10 trillion. This investment has distinct effects on "builders" (companies investing in AI infrastructure) and "suppliers" (companies providing components for AI capex).
AI Builders (e.g., Hyperscalers like Meta, Alphabet, Amazon, Microsoft, Coreweave, Oracle)
- Balance Sheet: AI capex appears as an asset, increasing invested capital. Funding for this capex comes from a mix of debt and equity (e.g., for hyperscalers, roughly one-third debt, two-thirds equity, often from existing cash flows like Alphabet's advertising revenue).
- Income Statement: In the year of investment, earnings typically decrease due to depreciation/amortization of the capex. While some argue companies are amortizing over too many years, this only lessens the immediate earnings hit, not eliminates it. Early in the process, significant earnings from AI capex are not yet materializing.
- Cash Flows: Capex represents a cash outflow. Despite potential future earnings, the immediate effect is a reduction in cash flows.
AI Suppliers (e.g., Nvidia, Taiwan Semiconductor, Broadcom, power companies, data center real estate)
- Revenues: Capex for builders translates directly into revenues for suppliers. A large portion of AI capex goes into chips, networking equipment, and data center infrastructure.
- Income Statement: These revenues flow into net income, especially for companies with high net margins (e.g., Nvidia with 50-60% net margins).
- Cash Flows: High net income translates into higher free cash flow to equity.
Aggregate Market Effects
When considering the market in aggregate (both builders and suppliers):
- Invested Capital: Aggregate invested capital will rise significantly due to AI capex.
- Net Income: Collectively, net income can increase. This is because the high net margins and revenues of suppliers can more than offset the amortization costs incurred by builders, especially if amortization periods are extended.
- Cash Flows: Aggregate cash flows will decrease. The massive spending on capex (trillions of dollars) represents a collective drain on cash flows, with the impact on builders dominating any increases seen at suppliers.
Empirical Evidence from Financial Statements
Analysis of publicly traded US companies reveals:
- Capex Growth: Collective capex is up 36% quarter-over-quarter (Q2 2026 vs. Q2 2025) and 26% over the last 12 months compared to the previous calendar year. Technology, communication services, and consumer discretionary sectors show significant jumps.
- Balance Sheet Changes:
- Book equity: Up 13% (approx. $1.8 trillion)
- Debt: Up 8% (approx. $1.9 trillion)
- Invested Capital: Up 13% (approx. $3.7 trillion) This indicates that debt is funding a slightly larger portion of this capex than equity in aggregate.
- Earnings Growth: Aggregate earnings are up 49% (last 12 months vs. 2025, and Q2 2026 vs. Q2 2025). This supports the idea that supplier earnings are currently outweighing builder amortization. However, similar to market cap, this growth is concentrated: less than half of technology companies, for example, are reporting higher earnings.
- Cash Flow to Shareholders:
- Dividends: Up 5-6% (last 12 months vs. 2025), totaling $857 billion.
- Buybacks: Up 8% (last 12 months vs. 2025), totaling $1.3 trillion. While growing, these rates are significantly lower than the 49% earnings growth. The percentage of earnings returned to shareholders has dropped to about 63%, reflecting that AI capex is absorbing cash flows that might otherwise have gone to shareholders.
Future Scenarios for AI
Two main scenarios for the AI boom are considered:
Scenario 1: The AI Boom is Real
- Earnings: Earnings at AI builders would eventually rise as AI investments generate revenue.
- Market Value: Even with rising earnings, market value might not necessarily increase if the return on invested capital (ROIC) does not exceed the cost of capital. For example, investing $5 trillion and generating $50 billion in income is insufficient if the cost of capital is 8% (requiring $400 billion in earnings).
- Cash Flows: Cash flows at builders should improve as earnings come in and net capex levels off.
- Business Transformation: Builders will become capital-intensive businesses, a significant shift from their previous capital-light, high-growth models.
Scenario 2: The AI Boom Fizzles
- Earnings: Expected long-term earnings for AI builders would not materialize. Accountants would eventually write down the trillions of dollars in capex, leading to significant losses.
- Debt Risk: If AI capex was disproportionately funded by debt, companies could face truncation risk, leading to distress and defaults.
- Market Reaction: The market would penalize these companies for wasted capex and loss of trust in management.
Both scenarios are plausible, and the actual outcome might lie somewhere in between.
Investment Strategies
Given the uncertainty, four investment approaches are outlined:
- Go All-In on AI: Believe AI will cause total and near-term disruption. This involves buying companies benefiting from AI and shorting those that will be disrupted. Caution is advised against excessive leverage, as seen with Leo Ashkanazi's hedge fund.
- Market Consensus: Align with the market's implicit belief that AI will be significant but take longer and be smaller than extreme optimists predict. This is the strategy adopted by investing in broad index funds like the S&P 500.
- Cautious Optimism: Believe AI will succeed but face barriers and be less profitable than currently anticipated. This involves holding existing investments but keeping new cash in reserve, waiting for clearer signals or better valuations.
- Paranoid Pathway (AI Crash):
- Exit Market: Sell all stocks and hold cash, or invest in alternative assets like collectibles or cryptocurrencies.
- Short AI: Stay in the market but short companies benefiting from AI and buy those expected to be disrupted, betting on an AI bust.
The speaker personally adopts a hybrid approach: holding a significant portion of their portfolio in market consensus (including some "Mag Seven" companies involved in AI capex) while keeping new cash in reserve due to lingering uncertainty. This allows for sleeping well at night, even if it means potentially leaving some money on the table.
Finally, for younger investors, it's crucial to consider the impact of AI on human capital – the present value of future wages. If one's profession is susceptible to AI disruption, hedging this risk by investing in less affected areas or acquiring new skills is advisable.
Takeaways
- Despite a 0.54% monthly rise in 10‑year Treasury yields, U.S. equities grew 12‑13% in the first nine months of 2026, showing unexpected resilience.
- AI capital expenditures, now estimated at $3 trillion and potentially reaching $8‑10 trillion, boost earnings for suppliers while initially depressing builder profits through depreciation and cash outflows.
- The implied equity risk premium for the S&P 500 fell to about 3.7% after accounting for higher risk‑free rates, still above the 2% level seen at the dot‑com peak, indicating modestly attractive valuations.
- Aggregate corporate data reveal capex up 36% YoY, invested capital up 13%, and earnings up 49%, yet free cash flow to shareholders lags, falling to roughly 63% of earnings because AI spending drains cash.
- Investors can adopt four strategies—from fully embracing AI to shorting it—while the author favors a hybrid approach that holds core market positions and reserves cash to navigate uncertainty.
Frequently Asked Questions
How does AI capex affect the income statements of builders versus suppliers?
AI capex reduces builders' earnings in the investment year because the expense is recorded as depreciation or amortization, creating an earnings hit, while suppliers record the same spending as revenue, boosting their net income, especially for high‑margin firms like Nvidia. This creates opposite short‑term impacts on profitability.
Why did the implied equity risk premium for the S&P 500 drop to about 3.7% despite higher earnings growth?
The ERP fell because the risk‑free rate rose faster than the expected equity return; with the 10‑year Treasury at 5.29%, the S&P 500’s implied return of 8.99% yields an ERP of roughly 3.7%, lower than historical averages even though earnings have risen sharply. This reflects tighter valuation pressure from higher rates.
Does this page include the full transcript of the video?
Yes, the full transcript for this video is available on this page. Click 'Show transcript' in the sidebar to read it.
Helpful resources related to this video
If you want to practice or explore the concepts discussed in the video, these commonly used tools may help.
Links may be affiliate links. We only include resources that are genuinely relevant to the topic.