Over the past decade, US equities have decisively outpaced Australian equities on a like-for-like currency basis. The ASX 200 compounded at roughly 9.2% p.a. in AUD with dividends reinvested (10.6% with franking credits), versus 13-14% p.a. in USD for the S&P 500, an outcome so strong it has only been matched once since 1871. The gap was driven almost entirely by US mega-cap tech, while the ASX’s bank- and resources-heavy index delivered steadier, dividend-anchored growth.

The 5-year outlook is less clear-cut. The bull case rests on broadening AI earnings: Goldman forecasts 12% S&P 500 EPS growth in 2026 and 10% in 2027, with the Magnificent Seven contributing 46% of that, a shrinking share, suggesting wider participation. Hyperscaler AI capex is expected to reach ~$725 billion in 2026, on the
assumption the cycle runs through 2027-28.

But this concentration is also the risk. The S&P 500’s CAPE ratio, at 40.9, has only been this high once before 1999-2000, and the Buffett Indicator sits 56.6% above trend. Under a moderate slowdown, US returns could run 6-9% p.a. USD, still ahead of Australia, but by a much narrower margin. Under a genuine capex disappointment,
returns could fall below Australia’s steadier base case entirely. Since the capex cycle is expected to be tested within this 5-year window, any sharp re-rating is more likely here than deferred to later years.

This argues for reducing US concentration in favour of Australian equities:

• Different return driver: Australia’s return is dividend- and franking-led, not AI-dependent.
• Lower concentration risk: the ASX doesn’t hinge on a handful of stocks sustaining extraordinary growth.
• Narrower outcome range: Australia’s ~9-10.6% p.a. base case is far less dispersed than the US’s 6-14%+ range.
• A direct hedge: since the dominant US risk is AI-capex-specific, Australian equities offset it rather than merely diversifying by geography.

The historical case for US allocation is real, but the next 5 years carry wider risk tied to
one identifiable factor. Reducing US weighting isn’t a bet against it, it trades some upside for a smoother outcome.