The $2 Trillion Question | Tobias Carlisle on SpaceX, the AI Buildout, and the Rotation No One Sees

Historic overvaluation across all metrics

Adviser Perspectives tracks six to seven valuation metrics (Shiller PE, Tobin's Q, market trend deviations) that collectively show the market is at its most expensive level in the dataset, comparable only to the peak of the dot-com bubble.

The AI paradigm vs. mean reversion

Current sky-high multiples are only justified if AI creates a permanent singularity where mega-caps earn supernormal returns forever; if historical mean reversion holds, investors face reduced forward returns and heightened volatility.

Avoid market timing, seek value pockets

Rather than exiting the market entirely due to expensive headlines, investors should hunt for undervalued segments, as bifurcated markets often hide reasonable opportunities beneath the surface.

Early reversal signals flashing

Small caps have recently outperformed the Magnificent 7, while equal-weight S&P 500 (RSP) has begun outperforming market-cap weight, suggesting the decade-long large-growth dominance is showing cracks.

Valuation spreads at extremes

The gap between the most expensive and cheapest stocks sits in the 95th percentile, a level previously seen only at major bottoms like March 2009 and March 2020, indicating potential for significant mean reversion.

Historical cycle positioning

After a 10+ year large-growth bull market (2015-present), the market appears to be transitioning into a value cycle that could persist for 10-15 years, similar to the 2000-2015 period.

Accelerated depreciation cycles

Unlike railroads or fiber optics with 25-year lifespans, AI data center equipment turns over every 5-7 years, requiring massive ongoing capital expenditure just to maintain capacity.

Value capture uncertainty

It's entirely conceivable that all economic value from AI accrues to consumers rather than model creators, particularly as users may settle for cheaper, older-generation models rather than paying premium prices for cutting-edge versions.

Gartner hype cycle risks

While AI may eventually exceed its transformative promises (as the internet did post-2000), the path likely includes a trough of disillusionment and potential crash as investment runs ahead of immediate monetization.

Target deep value segments

Small, micro, and mid-cap value stocks currently trade at discounts to long-term averages and offer superior forward returns even without multiple expansion, simply through earnings reinvestment.

Earnings inflection emerging

After declining since 2022, small and mid-cap earnings appear to be bottoming and rising again, while large-cap tech earnings that drove the market were artificially inflated.

The mean reversion bet

If you believe in mean reversion, the smart bet is concentrated in small and micro value, where compressed multiples offer asymmetric upside as the valuation gap normalizes.