Summary
Reasons to believe that the US is exceptional.
The consensus view is: why buy any stock market outside the US? In this paper, I challenge this view.
Conclusion:
Is the US so exceptional, and if so, can this exceptionalism last?
Certain things are quite clear and cannot be disputed.
First, in terms of US economic growth, we have seen outperformance versus the rest of the world with superior productivity growth in particular (see chart). We cannot deny that the US economy has grown strongly over the last few years.
Second, at least in dollar terms, it is still the largest economy in the world with an unchallenged reserve currency status for the US dollar. There’s no doubt that the greenback is the de facto global currency accepted pretty much everywhere.
Third, when we look at the stock market, it is quite apparent that the US stock market has outperformed for 10+ years on the back of strength in the technology sector, which has posted robust earnings growth over time. Over the past decade, US equities have delivered an impressive 14.8% annualised return per year, outpacing global ex-US equities (which have delivered only 7%) and euro area equities (+7.8% per year). Furthermore, this performance has been compounded by a strong US dollar, which has been the case since 2011.
Why buy any stock market outside the US?
I have some challenges to this consensus view. First, I believe that US economic growth is set to slow, as many consumers are showing poor sentiment, reflected in the University of Michigan Consumer Sentiment survey. I think this is largely the outcome of the very high inflation rates, in the wake of COVID and the war in Ukraine, up until recently.
Second, I expect that it will be difficult to maintain the debt-funded government spending splurge that the Biden administration undertook over four years (see chart), particularly due to COVID-era pandemic spending. But this deficit spending also includes the Inflation Reduction Act and the Chips Act, both aimed at boosting investment in the US.
Third, since October 2022 (the start of the current bull market), the excellent performance of US large caps is at least half due to the revaluation of the S&P 500, particularly the Magnificent 7 stocks which have driven them to lofty valuations. The valuation of the S&P 500 was 15.5x forward earnings in October 2022. Today’s the P/E is 22x, over 40% more expensive, and the Magnificent 7 are now trading at around 30x forward earnings, a very high valuation versus any point in previous history. Due to these high US valuations, Vanguard, a fund manager, forecast a 10-year future return for US stocks of only an average 1.7% per year from here. They expect these P/E valuations to return slowly closer to long-term average, representing a drag on US stock performance. In contrast, they see 6.1% average future returns over the next 10 years for global stocks outside the US.
We continue to see excitement around AI, which has largely fuelled the outperformance of the Magnificent 7. Four hyperscalers out of the Magnificent 7, namely Amazon, Meta, Google, and Microsoft, are looking to invest even more on AI in 2025, totalling investment capex of USD 325 billion. This is substantially more than the already massive USD 220bn that they have spent in 2024, and they intend to earmark further colossal investments in AI in 2026 and 2027. If we look at this 4-year period, we are talking a total investment potential of a trillion dollars or even more.
The future of AI
There are questions, in my view, over the eventual return on investment from this massive multi-year investment wave. Minds have been focused of late by the launch of the Chinese large language model DeepSeek R1, and also the arrival of the Chinese more broadly, including competing models from technology giants such as Alibaba.
I think this will lead to the commoditisation of AI, which will be eventually good for economic productivity, because the cheaper entry point. Thanks to the massive offer of generative AI and other AI models on the market, prices will fall for all. This, in turn, will accelerate the adoption of AI in the wider economy, outside of just technology.
AI should boost economic productivity, taking US economic growth to something like a 2% growth in productivity over the medium term. This is about as high as it ever reached in previous technology booms in the early 1990s with the introduction of the PC, and around the year 2000 with the acceleration of the adoption of the internet.
Many sectors, including software, banking, professional services and healthcare could benefit from accelerated AI adoption in the wider economy. However, I am not so sure this will be good for companies supplying AI models, the hyperscalers, even given these higher volumes, with the lower potential prices and lower profit margins.
Conclusion
While the US may still outperform for a little longer, I think this outperformance trend is getting long in the tooth both at the economic and at the stock market levels.
I would advise those who have heavily invested in US stocks and, of course, have done extremely well from this exposure over the last few years, to diversify away from this heavy US stock exposure and the heavy exposure to the technology sector, as nothing lasts forever.
Bear in mind that the US stocks today comprise nearly 74% by weight of the MSCI world index, and that is a record concentration to have in one country today. In other words, even if you invest in the MSCI World index, you may think that you are investing in the global stock market, but actually nearly three-quarters of that investment ends up in the US, which is not really that well diversified.
I would suggest considering other asset classes, but at least to think about other stock markets, such as the Euro area, the UK, Japan, and perhaps even China particularly starting with these technology companies, given that they are very cheap, still growing well, and are also starting to exploit the same AI trends that are unfolding quickly in the US.