There used to be a time when investors could buy the S&P 500 as a sensible investment, forget about it and let it grow in value in the background. That era is now over, according to Julian Bishop, manager of the Brunner Investment trust.
“If you'd asked me 10 years ago: ‘I want to buy one tracker, put it away for the rest of my life’, I probably would have said the S&P 500. It's a remarkable collection of businesses, great institutions in America, a great financial system," he said. "I think none of those things is true anymore.”
There are multiple reasons for this change of heart, the manager explained, from cash flow yield to concentration and valuation.
Cash flow over earnings
On the first measure, “if you actually look at the free cash flow yield on the S&P, it's very low compared to history – way below the cash you would get in a bank account”, he said. “That's worrying.”
Bishop uses free cash flow yield rather than the more commonly cited earnings yield because the latter, in his view, gives investors a misleading read.
When a technology company spends heavily on hardware – say, buying chips from Nvidia – accounting rules let it spread that cost across several years, even though the chipmaker books the full sale at once. The result is that both companies can appear profitable on an earnings basis at the same time, even when one is effectively handing all its cash to the other, he explained.
A narrower index than investors think
A more talked-about problem with index investing is concentration. Trackers labelled “global equities” are in practice heavily American portfolios and, within America, a heavily concentrated one.
“When people talk about global equities, you think about having exposure to a little bit of everything,” Bishop said. “But the US market is now about 70% of the world market. So when we talk about global equities, we're really talking about mostly US equities.”
Jason Hollands, managing director at Bestinvest, agreed with Bishop and called this trend ‘the Great Index Reset’. He also agreed that concentration is now the defining risk for index investors.
Within the S&P itself, concentration has narrowed even further. The top 10 stocks in the S&P 500 now represent 37.5% of the entire index, with the Magnificent Seven alone – Nvidia, Microsoft, Apple, Alphabet, Amazon, Meta and Tesla – accounting for a third.
By adding together the information technology sector, communication services (where Google and Meta sit), consumer discretionary (where Tesla and Amazon sit) and all the industrials related to AI, Bishop said it's “probably fair to say that 50 or 60% of the S&P now is tech and very dependent upon the AI narrative”.
This is not just an issue within developed markets or the US.
The MSCI Emerging Markets index, which spans more than 1,200 stocks across 24 countries, has close to 30% of its weight concentrated in just three semiconductor companies: Taiwan Semiconductor Manufacturing Company (TSMC), Samsung Electronics and SK Hynix, Hollands noted. TSMC alone represents 21% of the MSCI Taiwan index; Samsung and SK Hynix together account for 43% of the MSCI Korea index.
The consequence is that Taiwan and South Korea have become the two largest country positions within the emerging markets index, pushing China and India into third and fourth place. China's weighting has fallen from around 43% of the index in 2020 to just below 20% today.
The capex gap
At the centre of that narrative is a capital spending programme whose scale is “hard to overstate” and no revenues yet capable of justifying it.
“If you take the entire capex of the oil and gas sector, which historically was the biggest capex pool in the world, that was about $500 or $600bn,” he said. “But the entire energy sector's revenues are about $4trn.”
The AI industry, by contrast, is spending at a comparable or greater rate with a fraction of the income, with about $100bn in revenues this year from AI labs in the context of $1trn of capex. “That just cannot be sustained”, Bishop said.
“If you were to add up all the revenues of the biggest tech companies that ever existed – in 2024, Apple, Microsoft, Amazon Web Services, Google, Meta – their revenues were about $1.2trn, ” Bishop noted. “Really, AI needs to be generating a multiple of that to justify the current level of capex.”
SpaceX and what comes next
The arrival of SpaceX on public markets has heightened the concerns of both experts. Hollands noted that SpaceX will enter major benchmarks on a very limited free float, reshaping index composition and forcing passive funds to buy its shares regardless of valuation. When a company of that size enters a benchmark, existing constituents are diluted to make room.
Bishop used the SpaceX example to illustrate how valuations have detached from fundamentals.
“The market has decided that SpaceX today is worth about the same as Microsoft,” Bishop said. “SpaceX's entire revenues last year were a tenth of Microsoft's profit. We're clearly getting into fairly ridiculous territory.”
Should Anthropic and OpenAI follow SpaceX to public markets – as is widely expected – a small number of AI-related companies could quickly absorb a significant share of benchmark weightings, according to Hollands, crowding out hundreds of existing constituents and increasing concentration risk for investors who believe they hold broadly diversified portfolios.
The passive trap
Investors should not think they are insulated from these dynamics just by holding passive funds. Index inclusion, said Bishop, has become a mechanism that works in favour of the companies entering, not the investors buying them.
“Index funds are, to a certain extent, becoming like an exit route for these companies – they know they have a forced buyer of certain amounts of shares,” he said.
“Any buyer of an index fund who just blindly buys these stocks irrespective of the fundamentals must recognise they're in really speculative territory.
“If you look at put-call ratios, margin lending, momentum – they're all at extremes. Typically, those factors have not boded well.”
Alternatives for passive investors
For investors who remain committed to passive strategies, Hollands pointed to two alternatives.
The first is an equal-weight index fund, which holds the same constituents as a conventional benchmark but allocates an identical weighting to each company rather than sizing positions by market value. For US exposure, he suggested combining a standard S&P 500 tracker with the Legal & General S&P 500 Equal Weight Index fund.
The second is factor-based investing – funds that weight holdings according to fundamental characteristics such as sales, cash flow, book value and dividends, rather than market capitalisation.
Here, he pointed to the Invesco RAFI US Fundamental Value ETF, which owns the 1,000 largest US companies weighted on those four measures, and the Invesco MSCI World Equal Weight UCITS ETF for global exposure.