Financial Historian Warns of AI Bubble
Financial historian Edward Chancellor argues the current AI investment climate mirrors the 1990s internet bubble, with inflated valuations and vendor financing distorting the sector's health. He suggests investors consider shifting toward real assets or holding off until a market correction.
Edward Chancellor is a financial historian known for his expertise in financial bubbles and speculative excesses. His book, *Devil Take the Hindmost: A History of Financial Speculation*, chronicles financial manias from the South Sea Bubble to the dot-com era. He also authored *The Price of Time*, which examines the impact of low interest rates on asset prices and economic stability. Chancellor sees parallels between the AI investment boom and previous bubbles, particularly the late 1990s internet bubble. He points to inflated valuations and unsustainable business models as warning signs. AI startups often achieve high valuations based on perceived technical differentiation, even with limited revenue. The median enterprise value (EV) to revenue multiple for AI startups has been around 29.7x, reflecting investor confidence in future growth. However, M&A exits typically see more conservative multiples. Some analysts question the AI industry's investment structure, citing profitability and cash flow concerns. Investment in AI has surged, with AI startups capturing over half of global venture capital funding in 2025. Deal values for AI private companies jumped from $23 billion in 2024 to $150 billion in 2025. However, a National Bureau of Economic Research study in February 2026 found that 90% of firms reported no impact of AI on workplace productivity. Some analysts are concerned about circular financing, where major AI firms use investments to artificially boost valuations. Nvidia invested $1 billion across 50+ startups in 2024, and companies like OpenAI, xAI, and Anthropic received over 60% of AI sector funding in 2025. This can create a closed loop that distorts pricing and inflates costs. Despite the hype, AI companies face challenges in demonstrating profitability. OpenAI is projected to have annual losses through 2028, potentially reaching $74 billion in operating losses that year. Deutsche Bank estimates OpenAI's losses could total $140 billion between 2024 and 2029.
