The AI stock bubble, which gained significant attention during the latter half of 2025, has reportedly burst, according to insights from John Higgins, the chief markets economist at Capital Economics. Higgins delves into the implications of this phenomenon and raises concerns about what might lie ahead for the tech sector.
Traditionally, a market bubble occurs when asset valuations are significantly inflated compared to their intrinsic worth. Investors often see prices surge even in the face of concrete financial indicators suggesting weaker performance. Higgins explains that if one assesses current valuations against historical norms, it appears that the AI bubble has indeed deflated.
In a recent note to clients, Higgins pointed out that while the price-to-earnings (P/E) ratio for sectors like information technology had spiked in recent years, it has now declined to levels not seen since before the pandemic. The P/E ratio for the IT sector peaked at nearly 75% in late 2024, significantly below the over 150% seen during the dotcom bubble of the early 2000s.
The artificial intelligence (AI) landscape witnessed explosive growth through late 2025, showcasing 498 AI unicorns valuing a total of $2.7 trillion—revealing a significant surge in new startups, with 100 having been founded in 2023 alone. Notably, OpenAI experienced a remarkable jump in valuation, climbing to $730 billion.
However, this boom is now facing considerable resistance. A phenomenon dubbed the “SaaSpocalypse” has emerged, marked by a swift sell-off in software-as-a-service (SaaS) stocks. Investors are wary that advancements in generative AI could undermine traditional software business models, leading to notable declines in companies like Salesforce and ServiceNow, both shedding about 30% of their market values since the year’s start.
Higgins indicates that the challenges extend beyond just the SaaS sphere. The semiconductor industry, crucial for AI technology, has also encountered hurdles. A past surge in demand led to a chip shortage, further exacerbated by geopolitical strains like tariffs and conflicts, notably the ongoing war in Iran, which have caused supply chain disruptions.
Looking ahead, Higgins warns of a potential second bubble forming amidst these industries’ struggles. While tech companies have reported substantial earnings growth recently, questions arise regarding the sustainability of this trend. According to Bloomberg Intelligence, the “Magnificent Seven”—a group of leading tech companies—are projected to see earnings growth around 18%, a stark contrast to the 11% expected from the rest of the S&P 500.
Higgins suggests that, unlike traditional bubbles where market prices diverge from fundamentals, this situation may present a unique scenario where the bubble lies within the earnings projections themselves. The fundamental concern is what happens if these high earnings experienced by tech giants begin to recede.
Several factors could trigger a downturn in AI-related earnings. A notable risk is that the anticipated demand for AI might not materialize as strongly as forecasted. Despite a reported utilization rate of 88% among companies in AI, adoption appears to be stalling, in part due to employee fears of job displacement linked to AI technologies.
The broader economic landscape also poses risks, particularly as ongoing geopolitical conflicts disrupt resources. The war in Iran, for instance, has halted helium production in Qatar, which is integral to chip manufacturing. Furthermore, the heightened prospects of conflict not only threaten infrastructure, such as data centers, but could also lead to increased energy prices that inflate operational costs for tech firms.
Higgins concludes that if economic conditions falter, this could reverberate through the stock market, adversely affecting the earnings of AI-related companies, independent of the actual demand for AI technologies. This caution reflects a broader apprehension regarding the sustainability of recent growth trends within the tech sector amidst evolving challenges.
