Portfolios Spread Across Regions Still Share the Same Shocks

What's New

Geographic spread no longer protects a portfolio when its holdings depend on the same chips, energy, and policy. EQT makes this case in The New Rules of Capital. The report draws on more than 30 interviews with CIOs, CEOs, and partners at major allocators. It argues that diversification built for monetary cycles breaks down when shocks come from inflation, policy, and physical constraints. Allocators should test their next commitment against each holding's dependencies: energy source, technology position, value creation lever, and policy exposure.

Why It Matters

Allocators who measure diversification by region and asset class carry the most exposure here. Their risk reports can show broad spread while every sleeve relies on the same semiconductor supply chain or power grid. The finding challenges the long-held view that equities and bonds offset each other. One CIO told EQT that investors now have to "diversify your diversifiers." Acting on it requires company-level dependency data that asset-class labels do not capture.

By The Numbers

• About half: share of global trade now conducted between aligned blocs. Dependencies in services, technology, and critical commodities are deepening at the same time.

• 1%: public market average annual total shareholder return in South and Central America, 2010 to 2024. Private equity median IRR in the region was 11%. EQT notes the two measures are calculated differently.

• 50%: estimated share of global EBIT earned by private businesses. Institutional portfolios allocate an average of 10% to private markets.

• Under 10%: share of managers who deliver consistent above-median returns. That group earns up to 30% higher end-of-period MOIC over a decade than occasional top-quartile managers.

What To Watch

• Exposure tagging in risk reports. The first sign of adoption is reporting that tags holdings by energy dependency, technology position, and policy exposure alongside region and asset class. Expect early moves over the next 12 to 18 months.

• AI concentration across sleeves. EQT sees AI value forming at two ends: physical infrastructure and the operators that turn adoption into measurable productivity. Data center capital needs to 2030 total an estimated $5.2 trillion. That end depends on energy access, grid capacity, and permitting, the same inputs EQT flags as shared exposures.

• Manager selection criteria. Watch whether LPs weight repeated above-median performance over single top-quartile vintages in 2027 diligence.

The Wrap

Dependency-based construction gives allocators a clearer read on correlated risk than region and asset-class labels. The case is strongest while shocks keep coming from energy, chips, and trade policy. It weakens if monetary cycles regain control and the old equity and bond offsets return. It also depends on allocators building exposure data themselves, since EQT supplies the framework and leaves the measurement open. The first real test comes in 2027 commitment cycles, as allocators reset pacing and manager lineups.