The perception of a market rally driven solely by a few tech stocks is misleading. The S&P 500 excluding the top 10 companies has seen strong earnings growth and consistent ~15% annual returns for the past three years, indicating broad market health.

Contrary to the AI growth narrative, immense CapEx is transforming 'cap-light' tech giants into capital-intensive businesses. This spending pressures margins, reduces returns on capital, and mirrors historical capital cycles where infrastructure builders rarely reaped the primary rewards.

The extreme market concentration in AI stocks might not end in a tech crash. An alternative is that other sectors like financials, industrials, and energy will "catch up" as they benefit from the massive capital expenditure required to build out AI infrastructure, broadening market performance.

Despite Microsoft's massive AI investments, its stock only grew 4%, while NVIDIA's market cap soared. Investors punished Microsoft's heavy capital expenditure, favoring NVIDIA’s high-margin, fabless "picks and shovels" approach that captured immediate AI profits without the same infrastructure risk.

The market's recent strength is not being driven by the mega-cap MAG7 stocks, which are underperforming. Instead, leadership has rotated to sectors like basic materials, healthcare, industrials, and financials. The breakout in the equal-weight S&P 500 confirms this widening breadth is occurring under the surface.

Beyond the AI-focused headlines, the S&P 500 Equal Weight Index's new highs show market strength is broadening. Capital flowing into formerly lagging areas and strong earnings growth for the median stock suggest a genuine early-cycle economic expansion, not a concentrated tech rally.

MAG7 companies lag the broader tech market because their massive equity and debt issuances fund the AI buildout. This capital flows directly to other tech firms' top lines (e.g., memory, components), boosting their stocks while the MAG7's own shares are diluted by the capital raises.

Market-cap weighting turned the S&P 500 into a momentum fund for megacaps, leading to a decade of outperformance versus its equal-weight counterpart—a historical anomaly. Recent signs of equal-weight taking the lead suggest a potential market regime shift back towards value and smaller companies.

The capital expenditure on AI by a handful of U.S. hyperscalers is projected to hit $600 billion this year alone. This figure is staggering, nearly matching the entire planned 2025 CapEx for every non-technology company combined in the S&P 500.