Elevated US tech stock valuations have made some question whether there could be a tech ‘bubble’ reminiscent of the 90s dot com era. Jon Cunliffe explains the inherent differences between then and now.
Headlines warning about stretched valuations and bubble like price action (especially in US technology stocks) have naturally made investors nervous. High valuations, however, are not a reliable timing tool for exiting markets. They mostly reflect expectations: investors will pay more for each dollar of earnings when they believe those earnings are durable, resilient, and likely to grow faster than average. Today’s market leaders, the so-called Magnificent Seven, span software, semiconductors, hardware, cloud infrastructure, and AI-driven technologies. These companies tend to generate substantial free cash flow, maintain strong balance sheets, sell business critical products on recurring contracts, and sit at the centre of long, multi-year trends like Artificial Intelligence (AI), automation, and digital transformation. This business mix can support elevated valuations for longer than many expect, even if day-to-day news flow is noisy and volatility remains part of the investment journey.
Why some tech share valuations are at a premium
It also helps to remember that a single valuation number is an average laid over a very diverse set of businesses. ‘Tech’ now encompasses profitable chipmakers, cloud platforms, enterprise software, and device ecosystems – models with different cycles, margins, and capital needs. Within that mix, some companies command premium valuation multiples because their earnings compound rapidly, while others are priced more modestly. This dispersion matters. When profits rise strongly, even a noticeable compression in valuation does not have to derail total returns; multiples don’t need to expand if earnings keep doing the work. Many tech leaders also benefit from self-reinforcing ecosystems: advantages in distribution, data, developer communities, and access to capital can widen competitive moats over time. That allows innovation to be funded internally rather than through less reliable external financing, and global revenue bases reduce reliance on any single region. None of this eliminates risk, but it raises the bar for the kind of shock needed to force a deep, lasting de-rating (i.e. decrease in valuations) across the group.
Could we see a repeat of the 90s dot-com bubble?
Against this backdrop, it’s tempting to draw parallels with the late 1990s dot com era. Yet the underlying environment is different in several ways. First, corporate balance sheets are healthier. Across the US, euro area, Japan, and the UK, non-financial companies have been net savers for much of the post Global Financial Crisis era. For example, in the US there is currently a substantial net financing surplus, which means that the corporate sector generates more cash than it requires for capital expenditure, debt servicing and working capital. This contrasts to the late 1990s when many firms ran deficits into the 2000 peak. Stronger cash generation and less leverage tend to cushion earnings through slowdowns and make the system less fragile when borrowing costs rise.
Second, global investors as a group are not currently overextended in their equity exposure. While US households look heavily exposed to stocks, a broader measure that considers equities as a share of all assets – stocks, bonds, and cash – held by non-bank investors worldwide sits well below its early 2000 peak. Against this background there is relatively less vulnerability to a sharp, position-driven unwind if sentiment cools. Third, the supply backdrop is boosting holdings of bonds and cash. Reflecting the growth of the global money supply, estimates are that the bond and cash universe is expanding by about $7 trillion per year, roughly 5.7% of global equity market value, compared with around $1 trillion per year (about 4.5%) in the late 1990s. At the same time, net equity supply (i.e., the global supply of public equities) has been negative in recent years (roughly $90 billion annually) because buybacks and takeovers are removing shares from the market faster than companies issue new ones, whereas the late 1990s featured heavy net equity issuance. Mechanically, this means equity prices must rise – or investors must actively shift money – just to keep stock weights steady as the asset pool for defensive assets continues to rise strongly.
Fourth, the capital expenditure (capex) boom is much narrower than in the late 1990s. Spending on AI and data centres is undeniably strong, but it’s concentrated in technology and adjacent infrastructure rather than spread across the entire economy, as was the case in the 1990s. Measures like private non-residential investment relative to GDP haven’t surged in the sweeping way they did in the dot com boom, and even intellectual property investment, while healthy, hasn’t risen unsustainably. The upshot here is the current environment lacks the kind of economy-wide capex excess seen in the late 1990s, implying a lower risk of a widespread investment downturn undermining broader economic growth.
Today’s tech leaders benefit from strong cash generation, resilient business models, and long-term structural trends.
Fifth, and by no means last, valuations, while elevated, are far from dot com extremes. Market leaders in early 2000 traded at price/earnings ratios of around 70× earnings after rising from roughly 20× in the mid 1990s. Today, the mega cap cohort has mainly been in the 30×-40× range for years, and the equal weighted S&P 500 sits closer to 18× earnings. The headline S&P 500 looks expensive in part because the Magnificent Seven giants now account for roughly 35% of index weight – about three and a half times their 2015 share – reflecting outsized earnings growth as much as multiple expansion.
The role of interest rates in valuations
Interest rates clearly influence valuations because they discount future profits back to the present. Markets have spent the last few years adjusting to higher rates, and companies have adapted by improving pricing, efficiency, and capital allocation. Recently, policymakers have signalled additional cuts as labour conditions soften, which, if delivered, would ease the discount rate headwind. Interest rate markets in the US now discount a much steeper path of rate cuts which, if they materialise, are likely to support risk appetite. Even so, the more important driver from here is likely to be the path of earnings and free cash flow, which remains encouraging across key platforms and suppliers tied to AI, cloud, and automation. For the long-term investor, that means focusing on business quality and cash generation rather than trying to second guess the twists and turns of the interest rate cycle.
Volatility, of course, is the price of admission for equity returns — particularly in higher beta technology. Pullbacks of 5% to 10% are common even during healthy bull markets. Larger, more persistent declines typically require a deterioration in fundamentals: a sharp contraction in economic activity that forces earnings downgrades and tightens credit conditions; a policy mistake that slows money growth; or a significant rise in bond yields if investors question the sustainability of public finances. Outside of those conditions, turbulence around earnings seasons, regulatory headlines, or geopolitics is common but not necessarily trend breaking. One risk worth monitoring is the growing concentration of the US market, where a handful of mega cap companies have driven an outsized share of index returns. If optimism about their growth potential fades, the overall market can move abruptly, which makes periodic rebalancing and diversification more important.
The importance of diversification
Diversification makes sense because large cap US equities trade at roughly a 50% valuation premium to the rest of the world. As we’ve already discussed, this doesn’t necessarily suggest a reversal, but it argues for spreading exposure. Eurozone equities, with Germany at the centre, could benefit from over €1 trillion of planned spending on defence and infrastructure and from more flexible EU deficit rules. This policy mix should support more cyclically sensitive mid cap companies, and the rollout looks set to continue into 2026. In the UK, equities still trade at a steep discount to the US, and their mix – heavy in consumer staples, health, and commodities – can hedge inflation, interest rate volatility and geopolitical uncertainty. Emerging Asia should gain from favourable growth differentials versus developed markets and from a weaker dollar, an environment that can also allow additional, market friendly monetary easing. Japanese equities continue to have merit, as a ‘good’ rise in inflation coincides with better consumption and gradual monetary normalisation. Corporate reforms in Japan aimed at improving capital efficiency have already encouraged more buybacks, asset disposals, and dividend increases, while households are, for the first time in decades, shifting savings from cash and bonds toward equities.
High valuations often reflect confidence in durable, fast-growing earnings, allowing leading technology companies to sustain premium multiples despite volatility and noisy headlines.
Finally, as the world edges from globalisation toward an order dominated by the US and China, with rising economic nationalism and reshoring, companies that derive a larger share of revenue domestically may find tailwinds. Elevated valuations in US technology mostly reflect expectations of durable growth from businesses with unique assets in a world that is still digitising.
At the same time, it is prudent to hold a meaningful allocation to regionally diversified equity exposure alongside large cap US stocks to offset concentration risk and to access a wider set of return drivers. Coupling that equity mix with sensible allocations to high quality bonds and diversifiers, rebalancing when allocations drift too far from target, and resisting the urge to time markets based on headlines remains a sound way to navigate an environment that is fundamentally different from 1999, even if it doesn’t always feel like it.
