Ben Inker of GMO argues the market is in an AI-driven bubble, but not a simple repeat of 2000. His core point is that today’s bubble is easier to navigate than the 2000 or 2007–08 episodes because investors can still own non-U.S. equities, value, small caps, and other risk assets without having to abandon risk entirely; the harder part is that the bubble may also be an earnings bubble, not just a valuation bubble. He also warns that AI capex, circular financing, and a wave of new supply/issuance could pressure returns over the next year.
Watch on YouTube ›Get the market thesis, key claims, assets, contradictions, and follow-up questions from any financial video — then unlock a version personalized to your portfolio, watchlist, and favorite speakers.
This episode is a long-form interview with Ben Inker of GMO focused on bubbles, expected returns, AI capex, and private equity. The central thesis is that today looks like a bubble in U.S. stocks, especially AI-linked large caps, but it is an “easy bubble” relative to the hardest historical episodes because investors can still construct portfolios with normal risk that avoid the most expensive parts of the market. Inker contrasts this with the internet bubble, the global financial crisis, and 2021, arguing that each required a different kind of avoidance: in 2000 you could stay in risk assets but rotate within equities and credit; in 2007–08 you had to avoid risk assets broadly; and in 2021 you had to flee both stocks and bonds toward cash-like assets. Today, by contrast, he thinks non-U.S. …
Tactically, U.S. large-cap AI leadership looks crowded and vulnerable to supply and issuance pressure, while non-U.S. equities and other risk assets still look more defensible. The near-term risk is that earnings look strong just as financing and lockup-related supply increases.
Over the next few months, the more likely path is continued dispersion: U.S. growth may still hold up, but value, small caps, and non-U.S. assets should remain the better relative-risk trade unless AI spending keeps producing unusually durable profits. The view would weaken if capex-driven earnings stay robust and supply is absorbed without multiple compression.
Structurally, the interview argues that AI is a transformative technology whose builders may still earn poor returns once capital rushes in. That would reinforce a broader regime where investors should focus less on the story of innovation and more on who captures the economic rents.
The current AI bubble is an easy bubble for agnostic investors to handle because you can avoid overvalued US stocks while still owning risk assets outside the US.
Speaker argues you can own non-US risk assets that are priced to deliver decent returns, making this like 2000 (easy) rather than 2007 or 2021 (hard) — you don't have to go to cash.
The slope of the risk/reward line has fallen from 0.4 to 0.1 for global assets, but remains at ~0.4 when US equities are excluded.
The speaker uses this to show that non-US assets still offer decent compensation for risk while including US assets makes the picture much worse.
The AI boom represents an earnings bubble, not just a valuation bubble like in 2000.
The speaker distinguishes between valuation bubbles (2000) and earnings bubbles (current), noting earnings bubbles have historically been harder to recover from (citing European 2007/2008 example).
What makes an investment bubble easy versus hard to navigate?
He says an easy bubble is one where you can still hold a normal amount of risk and avoid most of the damage, so if you turn out to be wrong you still own a reasonable portfolio. A hard bubble forces you to abandon risk assets or even move to cash, which creates big career and client-retention pressure if you are wrong or early.
How did the internet bubble differ from the global financial crisis bubble?
He contrasts 2000 as a bubble mostly in growth stocks, where a normal 60/40 style portfolio could avoid the worst overvaluation by owning other risk assets. By 2007, every risk asset around the world looked overpriced, so diversification among risk assets no longer helped and investors had to reduce risk much more aggressively.
Why was the end of 2021 even harder to navigate than 2007?
He says both stocks and bonds were overvalued at the same time, meaning anything with duration was expensive. Avoiding losses required moving to cash or cash-like assets, but that is especially hard because clients dislike holding cash and inflation was eroding its value.
Unlock the full claims, asset map, scores, related transcripts, follow-up questions, and AI chat — shaped around your portfolio, watchlist, favorite speakers, and risks.
