Yves here. Richard Murphy argues that stock prices could decrease by as much as 50%. While that may sound dramatic, it is important to recognize the significant overvaluation of AI stocks. A recent article from the Bank of International Settlements warns that an AI crash could lead to an investment drought, economic contraction, and potentially a crisis.
The relevant section of the BIS Annual Economic Report is included at the end of this post. The Financial Times highlighted it as their main story:
The significant chart from their report:
This graphic illustrates only tech-related bubbles, excluding real estate, which constitutes a significant portion of collective wealth and is often leveraged, resulting in deflationary consequences if value declines sharply. Interestingly, the dot-com bubble appears mild compared to today’s situation, despite its alarming magnitude during its peak. The 1920 boom is also worth noting for its severity, primarily due to high leverage that caused a more significant backlash against the economy. High levels of margin debt led to substantial losses for banks, exacerbated by leveraged investment vehicles resembling today’s collateralized debt obligations.
While stringent securities laws now restrict margin debt, current levels are sending warning signals:
There hasn’t been a better time in a long while to review the quality of your portfolio.
History offers a clear verdict on rising margin debt.
Don’t let FOMO sway you. https://t.co/plx60LYayI pic.twitter.com/q0JXa7VvVE
— Thierry from arvy 🇨🇭 (@ThierryBorgeat) June 27, 2026
The current AI frenzy makes the dot-com bubble look relatively mild, which should raise alarms.
By Richard Murphy, Emeritus Professor of Accounting Practice at Sheffield University Management School and a director of Tax Research LLP. Originally published at Funding the Future
Stock markets in both the United States and the United Kingdom are signaling increasingly severe warning signs. Share prices are experiencing extreme valuations, confidence is declining, and the artificial intelligence boom is beginning to falter.
In this video, I discuss how these risks intertwine and why their impact may extend beyond those directly invested in shares.
The cyclically adjusted price-to-earnings ratio, known as the CAPE ratio, is currently reaching levels reminiscent of the 1929 Wall Street crash and the 2000 dot-com bubble. While history doesn’t provide exact timelines for market drops, it demonstrates that valuations of this magnitude can’t persist indefinitely.
The risks extend beyond stock markets. Banks and shadow financial institutions have heavily financed inflated assets. A sharp decline in share prices could ripple through the financial system, jeopardizing pensions, undermining lending, and triggering broader economic turmoil.
AI could serve as the catalyst. The technology is costly, often unreliable, and its implementation is taking longer than anticipated. Should anticipated profits fail to materialize, the companies driving the current inflated market valuations may experience sharp declines.
Are the authorities prepared for this eventuality? There is little evidence to suggest they are.
This is the audio version:
The Debate Ammunition for this video is available here.
This is the transcript:
I continue to express my belief that stock markets in the UK and the USA may be on the verge of a crash, and I will reiterate this. I foresee a significant drop in their value soon. How much? Following historical trends, it could be up to 50%. The data supports this, but this video encompasses more than just data.
Numerous factors are converging to suggest a looming crisis, and I am not alone in this assessment. Martin Wolf from the Financial Times shares similar sentiments. The financial warnings are growing, and it’s essential to take them seriously. This represents a crisis that will unfold during Andy Burnham’s time in office. Is he equipped to handle it? That is my inquiry.
The warning signals in our economy are increasingly difficult to overlook. There are three primary concerns.
Firstly, we see a nationwide decline in confidence.
Secondly, the AI boom is beginning to exhibit serious vulnerabilities, and I believe this trend will worsen.
Lastly, share prices have reached unsustainable valuations.
These risks are interrelated. They compound upon each other and have the potential to impact other areas of the economy, which is what I am apprehensive about.
Let’s delve into the first issue: growing alienation within our society. Many people feel detached from politics and public life. It’s a sentiment I hear everywhere. Individuals are tuning out from news channels, distancing themselves from the media altogether. UK politics appears disconnected from reality, and behaviors exhibited by figures like Donald Trump contribute to a destructive political atmosphere. This madness is pervading the UK, leading people to embrace ignorance as a coping mechanism. I often hear that they are reducing their engagement with social media, avoiding newspapers, and turning off news broadcasts. This loss of engagement has economic consequences.
People are withdrawing because they feel overwhelmed; however, this sense of overwhelm leads to tangible effects. Markets thrive when there are confident and informed individuals, both of which seem to be diminishing concurrently. The same individuals are often experiencing both conditions, resulting in decreased consumer spending, which will negatively impact the economy.
Simultaneously, AI is beginning to unravel. This sentiment appears to be widespread. Conversations reveal frustration, with many expressing, “I’m trying this, but it’s not working as anticipated.” AI is not delivering on investor expectations. Seven leading AI companies dominate U.S. stock market valuations, representing 40% of the market’s overall worth. Yet, the reality is that AI isn’t living up to its hype.
AI is encountering substantial operational challenges. Companies like Facebook have even advised employees to reduce their reliance on AI due to high costs. Furthermore, AI is proving costly to deploy and is not resolving expected issues effectively. Bugs in AI-generated software highlight its unreliability, as it often fails to execute routine tasks consistently, leading to user frustration and wasted time. Rather than saving labor, AI is consuming it.
This sluggish adoption rate may be surprising. I recall the dot-com bubble when I was involved in a dot-com company. Many claimed the rise of the internet would eradicate jobs, but that did not unfold due to the delayed implementation and the need for adaptation. The same is occurring with AI; adaptation will progress far slower than anticipated. Consequently, AI enterprises are unlikely to generate the anticipated profits.
Another challenge facing AI is a growing public backlash. Discontent with AI-generated content is evident on social media, where many are actively avoiding it. Platforms like YouTube are delisting this content, as are others. Automated customer services often frustrate users, who eventually realize they are interacting with a machine rather than a person. I personally experienced annoyance when repetitive automated responses made me feel undervalued by the company, which was evidently prioritizing automation over genuine customer service. This trend may result in companies losing valuable goodwill. If businesses overly depend on AI, their profits could suffer, compounded by rising social tensions as people grapple with job losses attributed to AI. This creates further stress within society.
Now, let’s discuss the extraordinary share valuations primarily driven by AI’s presence in the market. This phenomenon significantly influences U.S. stock values. While the UK may not mirror this substantial AI presence on the London Stock Exchange, the ripple effect is undeniable; when U.S. markets soar, many others—including the FTSE 100—follow suit.
This leads me to the third point: the exceptionally high valuations in the U.S. stock market. The following chart illustrates this:
This chart depicts Robert Shiller’s Cyclically Adjusted Price-to-Earnings Ratio (CAPE ratio), highlighting the long-term valuations of shares in U.S. markets. It contrasts stock prices with the previous decade’s earnings, with the long-term average being approximately 16. Conversely, today, the ratio is around 42, which is over two and a half times beyond normal standards. The displayed chart showcases how unusual current valuations are; it spans from 1900 to the present and features three peaks. The initial peak occurred in 1929, followed by the second during the dot-com era of 2000. The third peak, nearing the dot-com high, is the present day.
The dot-com era reached a peak CAPE ratio of 44, while today’s sits at 42. When the dot-com bubble crashed, the ratio fell to 15—a dramatic decrease occurring in two stages, with the second phase triggered by the 2008 global financial crisis.
Certainly, I cannot predict outcomes, but trends point to a concerning pattern. Markets are presently exorbitantly priced. Individuals purchasing shares in the U.S. stock exchange pay rates that correspond to 42 years’ worth of their companies’ earnings. This signifies an alarmingly low rate of real return on investment—one that typically leads to a substantial market correction, as witnessed in 1929 and 2000, and which could feasibly occur again now. While the market might claim, “this time it’s different,” history indicates that it rarely is.
Several factors could incite a market downturn; economic, social, and political confidence plays a crucial role. Other triggers include global conflicts, with figures like Donald Trump increasing uncertainty regarding warfare in the Gulf. Climate-related costs, largely ignored until this summer’s disastrous events, are also pressing realities. Political instability is on the rise, and amid this chaos, it appears that AI firms are overvalued, with the hype surrounding them starting to fade away.
Altogether, this creates an environment where a bubble may burst. Ignoring the potential is unrealistic. Once confidence erodes, prices can drop swiftly—historically demonstrated by the rapid plunge in 1929 and the significant fall in 2000, though slightly slower than 1929. Today’s decline could mirror those sharp drops, suggesting that share prices might indeed halved.
The likelihood of such a correction is growing. The Financial Times often discusses it, framing it as nearly inevitable.
The concern, however, is that a crash would not be confined solely to stock markets. If it remained isolated, one might ponder: what’s the harm? Although it risks some retirement prospects, the implications extend far beyond that.
Banks have significantly lent against the inflated share prices, and within the shadow banking sphere, trillions of dollars are expected to be involved. Should the shadow banking sector face turmoil, financial stability across the board may be at risk, affecting the mainstream banking system as well.
The Bank of England recognizes these risks as credible. This isn’t a conspiracy; it highlights serious mainstream concerns. They foresee this contagion as a potential trigger for an economy-wide crisis, which may become the most significant challenge Andy Burnham confronts during his premiership—perhaps well before 2029. Ignoring these political risks is deeply irresponsible, yet our leaders seem to be doing just that.
They persist in discussing growth as their sole economic strategy, while preparations for financial disruption should take precedence. Failure to plan for such disruptions could jeopardize Andy Burnham’s administration long before the 2029 general election, yet his silence on this subject is troubling.
