This article was issued and approved by Titan Investment Solutions Limited which is registered in England and Wales (10636265) and is authorised and regulated by the Financial Conduct Authority.

An update for September 2026:

A year ago, we asked whether artificial intelligence (AI) had pushed the US stock market into bubble territory. Since then, 2026 has given us a real-world test of that question, including a significant shift in market leadership. 

Here's what happened, where things stand now and what it means for how we manage your money.

What actually happened in 2026: “the Great Rotation”

Through the first half of 2026, investors grew increasingly nervous about how much they had paid for a handful of AI-related "mega cap" stocks (the very largest technology companies). Money moved out of those names and into cheaper, more traditional parts of the market such as energy, industrials, materials and more defensive sectors.

This shift is well documented. Morningstar's sector analysis  showed energy as the best-performing US sector, with technology being the worst over the prior three months. Morningstar's chief multi-asset strategist described this reversal as "a stark shift" from the trend of recent years. 

That stark shift continued through the summer. An S&P Dow Jones Indices sector dashboard  recorded energy as having the highest relative strength among large-cap sectors that month, while a widely reported single-day event on 30 July 2026 saw a broad technology-sector selloff (including major chipmakers) coincide with an energy-sector rally, tied to a spike in oil prices amid Middle East tensions.

Meanwhile, value stocks (shares priced modestly relative to current profits) meaningfully outperformed growth stocks (companies priced on the promise of fast future profit growth) at multiple points in 2026 too. J.P. Morgan Asset Management's own research  noted that value stocks had outperformed growth names year-to-date in 2026, building on a 21-percentage-point outperformance of international value over growth in 2025. Separately, sector strategists reported large value assets beating large growth counterparts by "more than 11 per cent" in the opening weeks of 2026.

However, while these figures may sound like a volatile backdrop for US and its AI names, there was good news for diversified investors. The wider US market still ended up comfortably positive. In fact, so far this year, the S&P 500 has returned around 13% (including dividends) according to data from Slickcharts and ChartRow , to 8 September 2026. That return figure shows how spreading risk across sectors, rather than concentrating it in a handful of AI winners, continued to work even during a turbulent year.periods of market confidence, supported by strong investor demand and generally healthy corporate balance sheets.

So, how expensive is the US market today?

The S&P 500 is the main index of the 500 largest US-listed companies. "Forward P/E" compares share prices to companies' expected profits over the next year. In short, the higher the number, the more investors pay per dollar of expected profit.

Currently, the S&P 500's forward P/E ratio is roughly 20 times earnings today  or closer to 21 times according to BlackRock's estimate . That's above the long-run average where the 10-year average forward P/E is approximately 19.0x . The magnitude of that difference may not look immediately significant to many. However, over time, it can result in a meaningful effect on portfolios — particularly when other, cheaper markets can produce similar returns. 

Concentration also remains a live concern. The ten largest companies in the S&P 500 made up approximately 40% of the whole index at various points in 2026  . That’s noticeably more concentrated than at the peak of the dot-com bubble in 2000, when the top 10 made up roughly 26–27% . 

When taken together, higher P/E and concentration figures mean investors relying on a passive, US-heavy portfolio are today paying a historically high price for a historically narrow slice of the market. For us, that’s the backdrop against which we look at where else capital can be put to work.

Looking domestically 

The UK is one clear example of where it’s possible to find similar returns but at a cheaper starting point, especially when you consider that the UK stock market has had a strong run this year. The FTSE 100 broke through 10,000 points for the first time on 2 January 2026, the first trading day of the year, helped by strength in mining, defence and banking stocks  and has continued setting fresh highs since.

Yet, even after these gains, UK shares still trade at a notable discount to the US. The FTSE 100's forward P/E ratio was approximately 13.35x as of 1 January 2026 . That’s compared with approximately 20x or more for the S&P 500 (see above). The gap between these figures is a key reason we continue to see value in maintaining meaningful exposure outside the most expensive parts of the US market.

Is this AI boom different from previous bubbles?

With America’s S&P 500 looking expensive and highly concentrated, the subsequent question of whether we are in a bubble is understandable. However, while headline numbers can be cause for concern, some comfort can be found in earnings results. 

For example, Nvidia's most recent figures, reported on 26 August 2026, illustrate the scale of real, delivered AI demand: quarterly revenue of $96.2 billion, up 106% on a year earlier, with Data Centre revenue of $89.0 billion, up 117% . This is real, delivered demand for AI computing power today, not just a promise about the future, though sales growth on this scale doesn't by itself tell us how much of it will ultimately convert into profit.

In fact, investors are already asking harder questions about how quickly AI spending translates into profit for the companies buying all this computing power, whether prices for AI services can hold up as competition increases and how much market share cheaper Chinese AI models are taking from Western providers. The market is therefore arguably becoming more discerning about which AI winners deserve their valuations, rather than rewarding anything with "AI" in the name — which would be behaviour we would likely see in a bubble. 

Furthermore, it's worth stepping back and looking at expected returns over the next 10–15 years, rather than the next ten months. J.P. Morgan's latest Long-Term Capital Market Assumptions  cover the decade-and-a-half ahead.

It shows that while US large caps (with their large concentration of tech and AI-story names) will no longer be the lead performer, the asset class is still expected to return a healthy 6%, if not a little more. 

But, could there still be a bigger correction from here?

While the US and its large caps projected future performance seems healthy (albeit down from its current point), it does seem prudent to approach the asset class with care. Not only are investors already worried about the lofty valuations, there are other several risks that could weigh on the outlook. 

Chief among them is the possibility that AI companies take longer than expected to convert their heavy spending into reliable profit, particularly as competition erodes pricing for AI products and services and squeezes margins for today's winners. Supply constraints could also persist, with advanced chips and the energy needed to power data centres both remaining potential cost pressures.

Monetary policy adds a further layer of uncertainty. The new Fed Chair, Kevin Warsh, who was confirmed by the Senate on 13 May 2026 and installed as chair two days later, has taken a markedly hawkish stance since his first FOMC meeting on 17 June 2026. At that meeting, the Fed held rates at 3.50–3.75% while dropping its prior easing bias. Warsh also sharpened that message further at the Jackson Hole symposium on 28 August 2026, saying the Fed was "not ruling out" a rate hike given still-elevated inflation. Taken together, this points to a "higher for longer" rate environment that could weigh on risk assets.

Finally, geopolitical or energy-market shocks remain a near and present risk, having already contributed to volatility earlier in 2026, as seen in the July rotation discussed above.

The bottom line

Ultimately, the US market is elevated by most historical valuation measures, with a forward P/E around 20–21x, against a 10-year average closer to 19–20x, and the AI trade showing some features that warrant caution, including record levels of index concentration. 

But "dangerously overvalued" depends on your time horizon and your level of diversification.
2026 has shown that even a significant rotation within the market doesn't have to mean a bad year for a well-diversified investor: the S&P 500 is still up around 13% for the year even after a sharp mid-year swing away from the previous AI leadership. 

Our stance is unchanged therefore: take a long-term view, diversify across regions and asset classes, favour quality and cash generation over narrative and use alternatives to help smooth the ride when volatility picks up.

If you would like to talk through what any of this means for your own portfolio, please get in touch with your adviser.

The Titan Square Mile Investment Team

Sources:  
Morningstar, 19th February 2026
Dow Jones Indices sector dashboard, 31 July 2026
J.P. Morgan Asset Management, 31 March 2026
StoneX Wealth Management, 19 February 2026
Slickcharts and ChartRow, to 8 September 2026
Trendonify, 26 August 2026
BlackRock, 7 July 2026
FactSet Earnings Insight; Trendonify 10-year median, both as of August/September 2026
RBC Wealth Management, data to 31 December 2025; BigGo Finance, late August 2026
ClaritX, citing Forbes reporting, mid-2026
Morningstar, Yahoo Finance UK, Hargreaves Lansdown, all reporting 2 January 2026
Siblis Research, FTSE 100 P/E & Earnings Growth data
NVIDIA Corporation, Q2 FY2027 results, SEC Form 8-K, 26 August 2026
J.P. Morgan Asset Management, 2026 Long-Term Capital Market Assumptions, 10 September 2026

Important Information
This document is marketing material and is issued and approved by Titan Investment Solutions Ltd (“TIS”) which is authorised and regulated by the Financial Conduct Authority. The company is registered in England and Wales with Company Number 10636265. The independent research is provided by Square Mile Investment Consulting and Research Limited ("SMICR") which is not authorised or regulated by the Financial Conduct Authority and does not undertake regulated activities. TISL and SMICR are wholly owned subsidiaries of Titan Wealth Holdings Limited. Titan Square Mile is a trading style of Titan Investment Solutions.

This document is aimed at professional advisers and regulated firms only and should not be passed on to or relied upon by any other persons. It is not intended for retail investors, who should obtain professional or specialist advice before taking, or refraining from, any action on the basis of this document. It is published by, and remains the copyright of, TIS. TIS makes no warranties or representations regarding the accuracy or completeness of the information contained herein. This information represents the views and forecasts of TIS at the date of issue but may be subject to change without reference or notification to you. This document does not constitute investment advice, a recommendation regarding investments or financial advice in any way and shall not constitute a regulated activity for the purposes of the Financial Services and Markets Act 2000. Should you undertake any investment activity based on information contained herein, you do so entirely at your own risk and TIS shall have no liability whatsoever for any loss, damage, costs or expenses incurred or suffered by you as a result. TIS does not accept any responsibility for errors, inaccuracies, omissions, or any inconsistencies herein. Past performance is not an indication of future performance.