Finance

Mark Tilbury: S&P 500 Concentration Risk AI Bubble Warning

Marcus van den Berg β€” Financial journalist specializing in markets, central bank policy, and economic trends4 min readUpdated March 31, 2026
Mark Tilbury: S&P 500 Concentration Risk AI Bubble Warning

Key Takeaways

  • β€’Mark Tilbury, a seasoned investor with over 35 years of experience, says he is actively changing how he allocates his money because of AI's growing grip on the S&P 500.
  • β€’In his video 'I'm Changing How I Invest My Money Because of AI,' Tilbury argues that nearly 40% of the S&P 500 is now concentrated in just ten companies, most of them tech giants whose valuations are built on future AI revenue rather than current profits.
  • β€’His concern is structural: passive investing in a market-cap-weighted index no longer means broad diversification.

Ten Companies Are Basically Running the S&P 500 Now

The S&P 500 is a market-cap-weighted index, which means the biggest companies get the biggest slice of every dollar you invest. That has always been true. What has changed is the degree. According to Mark Tilbury, roughly 40% of the entire index is now concentrated in just ten companies, and almost all of them are deeply committed to AI development, carrying the valuations to prove it. In I'm Changing How I Invest My Money Because of AI, Tilbury makes the case that when you buy a passive S&P 500 fund today, you are not buying 500 companies with roughly equal influence over your returns. You are mostly buying a handful of tech giants and hoping the AI revolution goes exactly as planned. That is a very different risk profile than most passive investors signed up for.

The Valuation Problem Nobody Wants to Talk About

Here is the uncomfortable part. The prices on these AI-centric companies are not being justified by what they earn right now. They are being justified by what investors expect AI to earn for them years from now. Tilbury points out that these firms are taking on significant debt to fund AI infrastructure buildout, and the market is essentially pricing in a future where that spending pays off enormously. Meanwhile, passive investment flows keep arriving automatically every month, pushing prices higher regardless of whether the underlying business results are catching up to the projections. It is a self-reinforcing loop, and the question of whether it ends gracefully or badly is one nobody can answer with any confidence. For context on what happens when speculative cycles hit a wall, the history of the dot-com era is instructive β€” and worth keeping in mind as AI valuations continue to stretch.

Our Analysisβ€” Marcus van den Berg, Financial journalist specializing in markets, central bank policy, and economic trends

Our Analysis: Tilbury gets the core right. The S&P 500 is quietly a leveraged AI bet whether you want it to be or not, and most passive investors have no idea they signed up for that.

The 'overlooked zone' framing is the most useful idea here. The companies building picks and shovels for AI adoption, not chasing frontier model supremacy, are where the asymmetric returns likely live. That part deserved twice the airtime.

The gold and cash advice is sound but almost too comfortable. It papers over the harder question of when overvalued AI darlings actually correct, and by how much.

What the video leaves largely unexamined is the feedback loop between index mechanics and valuation inflation. Because the S&P 500 is market-cap-weighted, every dollar flowing into a passive fund disproportionately inflates the companies already sitting at the top. That means the concentration problem Tilbury identifies is not static β€” it compounds automatically with every new contribution from every passive investor, every single month. The index itself has become a price-support mechanism for the very companies whose valuations look hardest to justify on fundamentals alone.

There is also a generational dimension worth flagging. A large portion of passive investors today started building their portfolios after 2010, during a period when concentration in tech delivered spectacular returns. For that cohort, heavy exposure to a handful of dominant technology companies has only ever felt like a feature. The idea that it could be a risk β€” structural, slow-moving, and hard to see until it is not β€” runs against a decade and a half of lived experience. That psychological anchoring is probably the biggest reason the warning Tilbury is raising will not land as hard as it should with the people who most need to hear it.

Finally, the geopolitical layer deserves more attention than it typically gets in these conversations. Several of the companies dominating the S&P 500 are deeply exposed to semiconductor supply chains, data center buildout in politically sensitive regions, and regulatory environments that are shifting fast in both the US and Europe. An AI correction does not have to come from disappointing earnings alone. It could arrive via policy, trade restriction, or a single high-profile infrastructure failure. Diversification strategies that do not account for those tail risks are only solving part of the problem Tilbury is pointing at.

Frequently Asked Questions

What percentage of the S&P 500 is concentrated in AI-related stocks right now?
According to Mark Tilbury, roughly 40% of the S&P 500 is concentrated in just ten companies, nearly all of them heavily exposed to AI development. This figure reflects market-cap weighting, meaning passive investors are far more concentrated in AI bets than most realize. (Note: the exact percentage shifts with market movements, and definitions of which companies qualify as 'AI-related' vary among analysts.)
Is the S&P 500 concentration risk from AI stocks comparable to the dot-com bubble?
The structural parallels are real β€” speculative valuations driven by future revenue expectations rather than current earnings, combined with automatic passive inflows that push prices higher regardless of fundamentals. That said, today's dominant tech companies are far more profitable than most dot-com era firms were, which makes a direct comparison overstated. The risk is genuine but the situation isn't identical, and anyone claiming certainty either way is oversimplifying.
Does switching to an equal-weighted S&P 500 fund actually reduce AI stock concentration risk?
Yes, meaningfully so β€” an equal-weighted S&P 500 fund spreads each dollar across all 500 companies rather than loading up on the largest, which directly reduces your exposure to the top ten AI-heavy giants. The tradeoff is that equal-weighted funds have historically underperformed market-cap-weighted indexes during periods when large-cap tech leads the market, which is precisely the environment we've been in. Whether that tradeoff is worth it depends entirely on how much S&P 500 AI stock concentration risk you believe is currently mispriced.
Why do passive S&P 500 investors keep pushing AI stock valuations higher even if those stocks look overpriced?
Because market-cap-weighted index funds buy more of whatever has already gotten bigger β€” it's a mechanical, valuation-blind process. Every new dollar flowing into a passive S&P 500 fund automatically buys proportionally more of the largest companies, which are currently the AI-heavy tech giants. This creates a self-reinforcing loop where passive inflows inflate valuations further, independent of whether underlying AI revenue projections are actually being met.
What are the practical alternatives for investors who want S&P 500 exposure without the AI bubble risk?
The most straightforward options are equal-weighted S&P 500 funds, sector-diversified ETFs that cap technology weighting, or supplementing a core index fund with allocations to international equities and small-cap funds that carry less AI concentration. Tilbury's broader point β€” which holds up β€” is that 'buying the index' no longer means what it used to, so investors should actively audit what they're actually exposed to rather than assuming passive means diversified.

Based on viewer questions and search trends. These answers reflect our editorial analysis. We may be wrong.

βœ“ Editorially reviewed & refined β€” This article was revised to meet our editorial standards.

Source: Based on a video by Mark Tilbury β€” Watch original video

This article was created by NoTime2Watch's editorial team using AI-assisted research. All content includes substantial original analysis and is reviewed for accuracy before publication.