When one of Wall Street's largest asset managers publicly declares that genuine diversification is "hard to find," investors — including those in digital assets — would do well to pay attention. That is precisely the signal J.P. Morgan strategist Gabriela Santos has sent, identifying a narrow set of havens capable of offering real insulation from the artificial intelligence trade that has come to dominate global equity portfolios: U.S. treasuries, gold, and European markets.
Santos's assessment arrives at a moment when the AI thematic has embedded itself so deeply into equity indices that escaping its gravitational pull is genuinely difficult. The concentration of capital into a handful of technology and semiconductor names — the picks-and-shovels companies building the infrastructure layer of the AI economy — has made traditional diversification logic harder to execute. When an investor sells one AI-adjacent position and rotates into a broad index fund, they frequently find themselves buying right back into the same cluster of companies through a different vehicle. The overlap is structural, not accidental.
This is the crux of what Santos is describing. Diversification, in the classical sense, is supposed to mean owning assets whose returns are driven by different underlying forces. When nearly every major equity market has absorbed significant AI exposure — either directly through technology weightings or indirectly through productivity assumptions baked into valuations — the correlation problem becomes systemic. You cannot diversify away from a theme that has colonized the benchmark itself.
Which makes Santos's three-part prescription worth dissecting carefully. U.S. treasuries are the most conventional of the three: duration assets that tend to move inversely to risk sentiment and are largely indifferent to the earnings trajectory of any technology sector. If AI-driven equity enthusiasm reverses — whether due to a demand shortfall, a regulatory shock, or simply a valuation reset — treasuries provide the traditional flight-to-safety buffer that institutional portfolio construction has relied on for decades.
Gold is the more philosophically interesting inclusion. The metal has no earnings, no cash flow, and no direct relationship to any corporate technology cycle. Its value derives from scarcity, central bank demand, and the collective psychology of investors seeking stores of value outside the financial system's normal plumbing. In a world where AI has inflated equity valuations to the point that even bonds correlate uncomfortably with risk assets during stress events, gold's orthogonality to the technology earnings story makes it a genuine diversifier rather than a theoretical one. For readers of this publication, the parallel to Bitcoin's "digital gold" thesis is obvious — though Santos's framing stops at the physical metal, which remains the institutional default when genuine decorrelation is the goal.
Europe is the most tactically specific of the three recommendations. European equity markets carry materially lower technology weightings than their U.S. counterparts, and their valuations have lagged American indices substantially through the AI bull run. A portfolio rotation toward European exposures therefore accomplishes two things simultaneously: it reduces AI-correlated technology concentration and picks up assets that may have room to re-rate if the narrative eventually shifts toward value over growth. There is also a currency dimension — euro-denominated assets provide dollar diversification as a secondary effect, which matters when AI euphoria is disproportionately a U.S. dollar story.
For the digital assets industry, Santos's comments carry an underappreciated subtext. Crypto has spent years arguing for its place in institutional portfolios precisely on the diversification thesis — that blockchain-native assets move to different rhythms than traditional markets. The AI supercycle has complicated that argument. As technology stocks surged, so did many digital asset valuations, tightening correlations and weakening the clean decorrelation story. Meanwhile, gold and treasuries — unglamorous, low-velocity assets — have quietly been doing the actual work of portfolio insulation that crypto once claimed as its own competitive advantage.
That does not mean digital assets are without a role in a diversified portfolio. But it does mean the diversification thesis needs to be argued more carefully and substantiated with data rather than asserted on ideological grounds. If a senior J.P. Morgan strategist is scanning the investable universe for genuine AI diversifiers and landing on treasuries, gold, and Europe rather than crypto, the industry faces a positioning challenge it cannot ignore. The question for digital asset allocators is whether the next market stress event will finally establish Bitcoin or ether as genuine portfolio ballast — or whether, once again, the flight-to-quality trade will flow past them and into the same centuries-old hedges Gabriela Santos is pointing at today.
Santos's warning is ultimately a structural one: in a market reshaped by a single dominant theme, the assets that offer true independence from that theme are fewer and more traditional than most investors prefer to admit. The search for diversification has gotten harder, and the honest answer to where it lives is less exciting than many portfolio managers would like.
Written by the editorial team — independent journalism powered by Bitcoin News.