.png)
Are We In A Bubble?
- Alpesh Patel
- Oct 28
- 5 min read
Whenever markets soar faster than fundamentals, the word bubble begins to circulate like a contagious rumour.
In 2025, as artificial intelligence (AI) transforms economies and the “Magnificent Seven” dominate global equities, many investors have invoked the spectre of 1999.
Are we in another speculative mania, or merely a rational repricing of innovation? This article argues that we are not yet in a bubble, but we are building the scaffolding for one.
Drawing on Goldman Sachs’ 2025 Global Strategy Paper and historical evidence, it will show that while valuations and concentration are high, earnings growth, strong balance sheets, and limited leverage still differentiate the present from past bubbles.
Read report:
However, the psychological and structural preconditions for exuberance are forming - making today less a bubble, and more its prequel
1. Defining the Bubble: Euphoria Meets Leverage
A financial bubble occurs when asset prices rise far beyond their fundamental value, fuelled by speculative behaviour and leverage.
Historically, bubbles share three features: rapid price appreciation, extreme valuations, and systemic vulnerability
Goldman Sachs identifies the same criteria: exuberance around transformative technology, capital inflows chasing hype, and valuation multiples that assume implausible future cash flows.
From the Dutch Tulip Mania (1630s) to the dot-com boom (1990s), bubbles start with innovation and end with disillusionment.
In the current cycle, the innovation is AI - a technological revolution akin to electricity or the internet. The question is whether markets are pricing AI’s potential or fantasy.
2. Valuations: Stretched, But Not Irrational
The rise of mega-cap technology stocks has been extraordinary: the ten largest U.S. companies now account for nearly 25% of global market capitalisation
Yet Goldman Sachs finds today’s valuations well below the extremes of past bubbles.
The median forward P/E of the “Magnificent Seven” is ~27x, roughly half that of the 1999 leaders
The price-to-sales ratios are lower, and PEG ratios (price/earnings-to-growth) stand near 1.7 — far below the 3.7 seen during the dot-com peak.
Dividend Discount Models imply 8% nominal perpetual growth, ambitious but not absurd relative to recent earnings trends.
Moreover, unlike the speculative startups of 1999, today’s tech giants generate enormous free cash flow and returns on equity exceeding 40%, with little or no net debt.
The Goldman report concludes bluntly: “Valuations are stretched, but not yet consistent with historical bubbles”
In short, prices are high because profits are high - not because investors have lost their minds.
3. Fundamentals and the AI Revolution
In most bubbles, prices lead profits. This time, profits have led prices. Exhibit 3 of the Goldman Sachs paper shows that tech sector earnings per share have grown sixfold since 2009, compared to just 80% for the rest of the market
This growth reflects genuine structural transformation. AI, cloud computing, and digital infrastructure have increased margins and productivity.
The largest firms - Nvidia, Microsoft, Apple, Alphabet - enjoy economies of scale, data monopolies, and first-mover advantages that justify premium multiples.
Goldman’s analysts argue that the current rally is driven by fundamental growth, not “irrational speculation about future growth”
That said, this justification is fragile: the higher prices rise, the more flawless earnings must remain. When valuations depend on perfect execution, the line between rational optimism and euphoria thins rapidly.
4. Market Concentration and the Illusion of Safety
While the fundamentals are strong, market structure is flashing amber. The top ten U.S. firms now equal the market capitalisation of the entire Eurozone, Japan, India, and Canada combined. This degree of concentration is historically rare and often precedes instability.
Goldman notes that market leadership has always rotated - from railroads (1850s) to energy (1920s–70s) to banks (2000s) - and that “dominant sectors rarely remain so forever”
No company from the S&P 500’s 1985 top ten remains there today. The danger is not immediate collapse, but narrow fragility.
When too much capital is tied to too few firms, even a small disappointment can trigger systemic tremors. As history shows, bubbles are rarely identified by valuation alone, but by concentration and complacency.
5. The Role of Leverage: The Missing Fuse
The defining feature of destructive bubbles from 1929 to 2008 is leverage. Excessive debt amplifies losses when prices fall.
Yet Goldman observes that today’s balance sheets are exceptionally strong:
Big tech firms finance capex largely through free cash flow, not debt
Corporate leverage is low, banks are well-capitalised, and household balance sheets remain healthier than pre-crisis levels.
Even the current surge in AI-related capex (projected to reach $432 billion by 2026) is internally funded rather than borrowed.
In other words, there is froth, but not fragility. The absence of systemic leverage makes a 2000-style collapse unlikely, even if valuations correct sharply.
6. Behavioural Undercurrents: The Early Euphoria Stage
If fundamentals are sound and leverage contained, why the unease? Because bubbles are psychological before they are financial.
Exhibit 2 in the Goldman Sachs report outlines the classic pattern: a technological breakthrough → new entrants → speculation → dominance → secondary innovations → disruption → mania → collapse
By that taxonomy, 2025 sits between stages 2 and 3: exuberant investment and rising valuations, but not yet irrational frenzy.
Retail speculation remains muted; corporate narratives dominate. Yet the language of inevitability - “AI will change everything” - already echoes the dot-com playbook.
The risk, as Goldman concedes, is that excitement and capital intensity “rhyme with previous bubbles”
The conditions are forming; the catalyst (disappointing earnings or interest-rate shock) has yet to arrive.
7. Macro Context: Cheap Money, Expensive Assets
Bubbles rarely form in isolation. Ultra-low real interest rates and excess global savings have inflated all asset prices, not just technology stocks.
Goldman notes that equity valuations are elevated across regions and sectors, from U.S. tech to European banks
Thus, what appears as a “tech bubble” may in fact be a liquidity bubble - a structural inflation of all risk assets caused by cheap capital.
If central banks sustain loose policy to offset slowing growth, valuations may stay high for years. If they tighten, air will escape rapidly.
Either way, the macro backdrop makes “not in a bubble yet” a conditional statement. The fuse is dry; all it needs is a spark.
Conclusion: The Pre-Bubble Economy
So, are we in a bubble? The answer is nuanced.
No, because valuations, profits, and balance sheets remain broadly defensible.
Not yet, because speculation, leverage, and mass participation; the true hallmarks of bubbles - are not (yet) dominant.
But soon, because concentration, narrative fervour, and capital intensity are converging.
Goldman Sachs concludes: “High valuations and competition in AI suggest investors should continue to focus on diversification”
That is the language of cautious optimism, but also of subtle warning.
If history rhymes, we are living in its penultimate verse - not in the bubble, but in its overture.
Disclaimer: This article is for educational and informational purposes only and does not constitute financial advice. Past performance is not a reliable indicator of future results. Investors should conduct their own research or seek independent, regulated advice before making any investment decisions.
Alpesh Patel OBE
Comments