AI-related equities have risen sharply in recent years and now represent more than a third of the S&P 500. But are AI equities in a bubble about to burst, or is the real risk lurking elsewhere? As always, the timing of a potential correction is the most intriguing question, and it hinges entirely on collective confidence. That confidence, in turn, depends on a web of underlying factors, among them steady growth, resilient supply chains, and ample liquidity.
There Are today many parallels to the dot-com era ...
Over the past five years, ever fewer companies represent an ever-larger share of total market value. Expressions such as “the Magnificent Seven” have entered the financial vernacular, and both mainstream media and social platforms are filled with speculation over which of these giants will rise or fall next.
In the autumn of 1929, Joseph Kennedy, father of President John F. Kennedy, famously remarked: “When even the shoeshine boy is giving stock tips, it’s time to get out of the market.” The quote has become a classic sentiment signal, a reminder that when speculation turns into a national pastime, prices are driven by hope (euphoria) rather than reason (information). This is the defining feature of a bubble: risk awareness fades when market psychology takes over the narrative. In the early stages, euphoria usually emerges within a narrow segment of the market. Gradually, however, the momentum blinds investors across sectors. When the bubble finally bursts, it tends to begin precisely where the first gains appeared.
... to the worry of several of the biggest AI companies
Two of the largest shareholders in leading AI companies, Jeff Bezos and Sam Altman, recently remarked that “We’re in an industrial bubble, but the benefits will be gigantic.” Bezos drew a parallel to the biotech bubble of the early 1990s, which, despite its collapse, ultimately paved the way for major scientific advances in e.g. epigenetics. From the ashes of that collapse arose a powerful wave of consolidation, that gave rise to a generation of “phoenix” companies, many of which are now global leaders. The same pattern characterised the dot-com era, from which Bezos’ own Amazon famously emerged as its most enduring Phoenix.
Financial markets have always been cyclical, oscillating between bubbles and corrections. Some have caused only minor ripples, while others have triggered deep and widespread crises of confidence. Bubbles appear to be a structural condition of collective trust, a phenomenon that cannot realistically be prevented through regulation or foresight alone.
The more interesting question, therefore, is not whether a bubble exists, but when the signs of a systemic loss of confidence begin to appear, the ideal exit point. To narrow that answer, we must first understand what is genuinely new in today’s market, how far valuations have risen, what forces are driving them, and whether the underlying fundamentals justify the momentum.
Yet the greatest expectations now concern the market as a whole
Since the global financial crisis, equity markets have risen almost uninterruptedly. Companies have expanded both their sales and earnings, and that growth has been remarkably consistent. This has reinforced the market’s collective confidence that the upward trend will continue. As a result, investors have priced in an ever-larger share of future earnings into today’s valuations. Confidence in continued stability has never been higher, not even in 1929, during the dot-com boom, or in the run-up to the financial crisis.
Confidence is broad-based and encompasses not only the largest corporations, as noted above, but also extends across small and mid-sized enterprises. The trust in equities is now so deep and widespread, that investors are now financing their purchases through record levels of leverage.
Since the global financial crisis, central banks have played a central role in the market’s uninterrupted upward trajectory. There have been occasional periods of widespread uncertainty triggered by substantial risks, such as the European sovereign debt crisis of 2011–12, China’s debt turbulence in 2015–16, and the Covid shock in 2020. Yet each time uncertainty crept in, and major investors began sell-offs, central banks stepped in and expanded the money supply. This restored an over-supply of liquidity between buyers and sellers (bid-to-cover), and with every successful intervention, market confidence deepened further. As a result, risk premiums (e.g. interest rate spreads) began to react ever more swiftly, and, strikingly, even positively, whenever key data on growth or employment disappointed. The market’s profound confidence, thus rests above all on a belief that central banks not only have the ability to anticipate and influence markets, but also the resolve to intervene whenever necessary.
There is, however, reason to be concerned about uninterrupted gains ...
At first glance, the track record of recent decades gives ample reason to believe that central banks will once again be able and willing to restore financial stability swiftly at the first sign of distress. A market correction, therefore, need not necessarily be deep, broad, or prolonged.
However, it is important to remember that central banks now operate under fundamentally different conditions. Their long periods of large-scale monetary expansion took place during an era of falling inflation, driven by accelerating globalisation, specialisation, and efficiency. Many came to believe that conventional economic theory was outdated, that the money supply could expand indefinitely without triggering inflation.
... in particular as central banks are in a new situation ...
That perception changed dramatically from the Covid period. The unprecedented monetary expansion was accompanied by direct fiscal transfers to consumers, fuelling demand for goods at precisely the moment when supply chains were collapsing under lockdowns. Inflation surged abruptly, forcing central banks to raise interest rates to contain it. Those rate hikes, in turn, caused sharp valuation losses on central bank balance sheets. In April 2022, for example, the Federal Reserve’s balance sheet stood at roughly one-third of US GDP, with the Eurozone showing a similar ratio.
In accounting terms, such losses are merely formalities for institutions that issue money. Yet confidence is a central bank’s true capital, and as trust wavered, many governments were compelled to reassure markets, either through explicit guarantees or public statements of support. At the same time, higher interest rates in high-debt economies sharply increased governments’ debt-servicing costs. This led to widening fiscal deficits, that is, a growing demand for money, even as the overall money supply stabilised. As a result, many countries now face the risk of debt spirals if this process repeats and interest rates remain elevated. That risk is not theoretical. It has become increasingly real, not least due to the United States’ new tariff-based trade policy.
Source: Rangvids blog
... where slow steady price drops could be of benefit
Therefore, central banks are likely to raise interest rates more quickly at the first signs of rising inflation, thereby curbing economic activity. At the same time, in periods of financial turbulence, they will probably expand the money supply by smaller amounts, over shorter periods, and perhaps also with greater delay. The result is that shorter market declines may no longer be followed by long, uninterrupted periods of rising asset prices. In fact, to secure long-term financial stability, it may even be beneficial for central banks if markets experience mild and controlled corrections.
All else being equal, this means we are now closer to the point where markets have already priced in as much of future earnings as they dare, a top. When confidence is as historically strong as it is today, the market becomes increasingly sensitive to even small fractures in collective trust. Should such a fracture occur, its impact could be amplified as central banks begin to hold back from prolonged and large-scale interventions.
These are new dynamics, and it may take time for markets to find a new psychological equilibrium.
Read part 2 in next weeks blog post