AI: Boom, not bubble, yet – Staying grounded through diversification
Global stock indices continue to break record highs as we move towards the end of 2025, powered by strong growth in the emerging markets, as well as innovation in technology and the ongoing artificial intelligence (AI) investment boom. At the heart of the rally sit the “hyperscalers” – the dominant US technology giants driving advances in AI, cloud computing, and semiconductor capacity. Firms such as Microsoft, Alphabet, and Amazon have reported exceptional earnings growth and record levels of capital investment. Combined with strong pricing power, this has underpinned much of the S&P 500’s rise, even as other sectors have struggled to keep pace. Much of this now appears to be priced into valuations, with some arguing the US ‘has priced in perfection,’ while the real world is far from perfect. There’s growing concern that we may be in an AI bubble. Our view in the Blackfinch Multi-Asset team is that we’re not - at least not yet. However, AI stocks, particularly the hyperscalers, share some characteristics with historic bubbles.
In recent years, market leadership has become increasingly narrow. The S&P 500 equal weighted index – which removes the market capitalisation bias by giving every company an equal 0.2% share with quarterly rebalances – has significantly lagged the main benchmark, reflecting how dependent US equity performance has become on just a handful of mega-cap names. This idea is explored below, where we review both the S&P 500 and its equal-weight counterpart over the last five years to the end of October 2025.
Figure 1: S&P 500 equal-weight vs market-capitalisation – how concentration has driven US investment returns
When investors talk about a potential market ‘bubble,’ they often point to the late 1990s and the dot-com boom and subsequent bust. The period was fuelled by the arrival of the internet – a revolutionary technology that promised to reshape commerce, media, and communication. From 1995 to early 2000, the NASDAQ Composite index climbed more than 400%, driven by soaring enthusiasm for anything online. Companies with little more than a website and an understandable business plan were valued in the billions. All without the sound fundamentals to support them. This led to investors believing that the internet would create limitless opportunity meaning valuations detached from reality.
When the excitement faded, and profits eluded investors, the bubble burst – the NASDAQ fell approximately 75% from its peak, erasing trillions of dollars from the market. The technology has ultimately been transformative, but the timing and scale of those early expectations were far ahead of what businesses could deliver.
While comparisons to the late 1990s are understandable, the evidence points to a fundamentally different market backdrop today. This current rally is being led by companies with solid earnings, healthy balance sheets and clear, tangible growth prospects – not by speculation.
Figure 2, which compares the 12-month forward price-to-earnings (P/E) ratios of the Magnificent Seven (the seven mega-cap tech names at the forefront of the AI theme) and the S&P 500 equal-weighted index, helps illustrate this. Before reviewing the chart, it is worth noting that sectors like technology do typically trade on higher multiples than the wider market, but questions remain whether the current spread above the market P/E multiples is justified. Currently, many of the largest technology companies now trade at substantially higher valuations, reflecting the strength of their earnings and dominant market position. Yet, outside this narrow group, the average US company remains far more modestly priced. The equal-weight index continues to trade upward, following the red trendline, indicating that while parts of the market have become expensive, valuations across the wider index remain very much grounded.
Figure 2: Forward valuations – elevated, but not euphoric
This dispersion stands in sharp contrast to the late 1990s, when almost every technology-linked company was priced for perfection. At that time, valuation excesses were broad-based and detached from profitability. As we are today, the premium is concentrated among a small cohort of firms with demonstrable earning power and real cashflow generation.
Figure 3 reinforces this point by comparing valuations with earnings growth. During the dot-com period, forward earnings for US growth stocks were contracting even as prices soared – a clear signal of speculative euphoria. This time, the opposite is true. Forward earnings expectations have risen meaningfully, particularly across sectors linked to AI, semiconductors, and cloud infrastructure. To put this into context, the chart shows that 12-month forward earnings per share (EPS) for US growth stocks are currently rising at roughly +15%-20% year-on-year (yoy), while valuations – measured by the forward P/E ratio – have steadily climbed and somewhat stabilised around 30 times earnings. By comparison, 1999 earnings were falling at around -10% yoy, even as valuations exceeded 60 times. In other words, prices today are being supported by improving profits, not speculation alone.
Figure 3: 12-month forward P/E & EPS– Fundamentals, not euphoria, should drive valuations
This alignment between earnings and valuations suggests that investor enthusiasm for AI and technology leadership, while strong, is being driven by tangible performance. Meaning that whilst the market may appear optimistic, it is not irrational at this point.
Looking forward, AI momentum doesn’t yet appear to be slowing. As of November 11th, the time of writing this article, US earnings season – a period of several weeks where US corporations report their quarterly sales and earnings – is again showing better than expected results for Q3. So far, nearly 91% of the S&P 500 companies have reported, with earnings up 11.75% on the back of revenue growth of 8.2%. Prior to the start of the season, FactSet put expected earnings growth at 8% from a year earlier. The technology sector has continued to deliver headline-grabbing results which has been underpinned by AI investment. Of note were the hyperscalers, who continue to focus spending, known as capital expenditure (capex), into the AI arms race. This is all in the hope of being the winner of the global mega trend – as shown in Figure 4.
Figure 4: US mega-cap technology hyperscalers continue to raise guidance[1]
This theme looks likely to carry on into the new year, and arguments can be made that profits from AI will continue to consolidate upstream with bottleneck companies in the AI supply chain standing to benefit most in the near term.
To conclude, the evidence suggests we may not be in an AI-driven bubble, but rather a phase of genuine – if uneven – growth. Unlike the late 1990s, today’s mega-cap businesses are underpinned by robust profitability, disciplined management teams, and measurable productivity gains. The enthusiasm for AI appears to be transitioning beyond just narrative, though a more tangible return on investment will certainly appease most businesses that have committed significant capex. While valuations for US mega-cap technology stocks have undoubtedly expanded – trading at forward P/Es near 30x versus the broader S&P 500 at around 24x – these multiples reflect strong earnings visibility and cash flow generation. By contrast, outside this narrow group, opportunities look far more attractively priced. For example, the S&P 500 equal-weight index trades closer to 16x forward earnings.
These divergences highlight the importance of staying balanced. With concentration risk tied to a handful of US names, markets remain vulnerable to shifts in sentiment or earnings disappointments. Yet the global opportunity set is wide. Strong balance sheets, lower valuations, and earlier-cycle growth stories in emerging markets offer both diversification benefits and asymmetry. This is on top of limited downside if US multiples compress, and greater upside if capital rotates abroad.
In short, the market is not detached from fundamentals but is one with a narrower leadership. As we move into the next phase of the AI cycle, we may see sensitivity in the pricing of these richly valued companies. For investors, it is important to participate selectively, recognising that innovation remains a driver of long-term returns, but acknowledging diversification can help weather further volatility from this point on. With the current outlook, this is achieved across markets, sectors and styles.
Staying grounded through diversification
For investors, the key lesson isn’t to avoid innovation – but to balance it. The narrow leadership driving global equity markets highlights the importance of diversification. Concentration risk has left the broader US market reliant on a small group of companies – the same firms driving impressive returns but also exposing the market to greater vulnerability if earnings momentum fades or sentiment shifts.
Within our multi-asset portfolios, diversification continues to be a key source of resilience. We maintain balanced exposure between market-cap and equal-weight indices, reducing the dependence on the largest companies while retaining exposure to their strong earnings growth. Beyond the US, we deliberately allocate to regions where valuations are more attractive, and earnings expectations are less demanding. For example, many US mega-cap technology stocks trade on forward P/Es near 30x, compared with around 16x for the equal weight index and closer to 17x for Chinese or Asian technology firms. This valuation gap provides both a margin of safety and an opportunity for re-rating.
Throughout the year we have favoured a more globally diversified asset allocation. Despite the US remaining in the headlines, it comes last when buying the market in sterling. We continue to believe that investors will increasingly benefit from a more balanced mix across geographies, sectors and styles. This combination of high relative valuations in the US, and improving fundamentals elsewhere, argues for a broader approach. As we move into 2026, we continue to advocate diversification, not only as a source of protection in the event of a correction, but as a driver of future returns.
As we look ahead, regional diversification remains key. With the US market still commanding a valuation premium, opportunities elsewhere are increasingly compelling. Europe and the UK offer exposure to a cyclical recovery at discounted prices, while Asia and emerging markets combine structural growth potential with more modest expectations. These regions stand to benefit as global manufacturing, digital infrastructure, and supply chains continue to evolve. Maintaining a balanced approach can capture these trends while mitigating the risks of concentration in a single market.
If you would like to discuss how market movements could impact your portfolios, or the wider investment market, do please contact our asset management team.
Website: www.blackfinch.am
Tel: 01452 717 070
Important Information:
Your capital is at risk. The value of your investments could go up as well as down and you may not receive back the full amount you invest.
Blackfinch Asset Management Limited is an Appointed Representative of Blackfinch Investments Limited which is authorised and regulated by the Financial Conduct Authority. Registered address: 1350-1360 Montpellier Court, Gloucester Business Park, Gloucester, GL3 4AH. Registered Company in England & Wales under No. 11639647. All information correct at November 2025.
[1] Source: Oxford Economics, LSEG Datastream, from Equities: Q3 earnings – the beat goes on. © Oxford Economics 2025