Societe Generale strategist Albert Edwards has issued a stark warning about AI stocks, presenting compelling evidence that analyst optimism is rapidly deteriorating even as markets hover near all-time highs. The veteran bear, famous for calling the dot-com bubble, argues that the AI-driven tech rally may be running out of steam.
Key findings from Edwards’ analysis reveal troubling trends: The 12-month moving average of analysts upgrading earnings-per-share forecasts has plummeted from approximately 58% to 50% since early 2024, despite the Nasdaq 100’s continued surge. Historically, such downtrends in analyst sentiment have coincided with the Nasdaq falling below its 200-day moving average. Edwards describes it as “a minor macro miracle” that the index remains at record highs given this rapid decline in optimism.
The disconnect between analyst expectations and actual earnings performance has widened significantly, with reality consistently lagging behind projections. Most concerning is that estimates for the S&P Composite 1500 have begun turning downward for the first time since their ChatGPT-driven rebound. Edwards notes that analyst optimism for the S&P 500 now shows “a series of lower highs and lower lows,” with both 6-month and 12-month moving averages trending downward.
Valuation concerns compound the risk: Tech stocks now comprise a larger percentage of the overall market than during the infamous dot-com bubble, while US equities represent an extraordinary 75% of global market capitalization. Edwards emphasizes that while declining analyst expectations wouldn’t normally threaten equity investors, “it is potentially a big risk when we are at nose-bleed-high valuations and optimism.”
Recent market disruptions have added fuel to concerns. Chinese firm DeepSeek’s late January release of a cost-effective AI chatbot challenged OpenAI’s ChatGPT dominance and raised questions about massive US AI infrastructure spending. President Trump’s inconsistent tariff policies have injected uncertainty, while Walmart’s lowered sales expectations and persistent consumer inflation concerns revealed in the University of Michigan’s Consumer Sentiment survey suggest broader economic headwinds.
Despite most Wall Street strategists maintaining bullish year-end S&P 500 targets, Edwards’ comprehensive chart analysis suggests investors should remain vigilant about the sustainability of the AI-driven rally.
Key Quotes
If the rapid decline in analyst optimism for the Nasdaq 100 is anything to go by, the tide is going out fast. Indeed, it is a minor macro miracle that the index is still trading above its 200 mav let alone record highs.
Albert Edwards, Societe Generale strategist, highlights the stark disconnect between deteriorating analyst sentiment and continued market strength, suggesting this situation is unsustainable and historically unprecedented.
It is the chart below that investors should be really nervous about. Notwithstanding the ‘blips’ from games played around reporting rounds, analyst optimism for the S&P 500 has been a series of lower highs and lower lows. Both the 6 and 12 month moving averages are now turning down.
Edwards emphasizes the most concerning trend in his analysis—the systematic deterioration of analyst confidence across multiple timeframes, which he views as a critical warning signal for investors in AI and tech stocks.
In ordinary times this would not be a serious threat to equity investors, but it is potentially a big risk when we are at nose-bleed-high valuations and optimism.
The strategist explains why declining analyst expectations pose an existential threat to the current AI-driven rally, noting that extreme valuations leave no room for disappointment or missed expectations.
If US analyst optimism is turning downward might this be enough to pull the rug from under what many see as an extremely expensive equity market, flirting with all-time highs?
Edwards poses the critical question facing AI investors: whether the shift in professional sentiment could trigger a broader market correction, particularly given the concentration of market value in AI and tech stocks.
Our Take
Edwards’ analysis deserves serious attention from AI investors, not because market timing predictions are reliable, but because it highlights fundamental valuation risks that have been obscured by AI enthusiasm. The DeepSeek disruption serves as a wake-up call that AI competitive advantages may be less durable than assumed, and that capital-intensive approaches may not be the only path to AI capabilities.
The comparison to the dot-com bubble is particularly apt—not because AI isn’t transformative (the internet certainly was), but because expectations and valuations have detached from near-term business fundamentals. The internet eventually revolutionized commerce and society, but many 2000-era internet stocks never recovered.
What’s most concerning is the timing disconnect: analyst optimism is declining while prices remain elevated, suggesting institutional investors may be trapped in momentum trades. When sentiment and fundamentals finally align with price action, the correction could be swift. For long-term AI believers, this may ultimately present buying opportunities, but near-term pain appears increasingly likely.
Why This Matters
This analysis carries significant implications for the AI industry and broader technology sector. Edwards’ warning comes at a critical juncture when AI stocks have driven market valuations to historic extremes, with tech representing a larger market share than during the 2000 dot-com crash. The deteriorating analyst sentiment suggests that even Wall Street professionals are questioning whether AI companies can deliver on their lofty promises.
The DeepSeek disruption exemplifies a key risk: if AI capabilities can be achieved at dramatically lower costs, the massive infrastructure investments by US tech giants may not generate expected returns. This threatens the fundamental investment thesis underpinning current valuations. For businesses heavily invested in AI infrastructure, this raises questions about capital allocation and return on investment.
The broader implications extend beyond stock prices. A significant AI stock correction could dampen venture capital funding for AI startups, slow enterprise AI adoption as budgets tighten, and potentially trigger a reassessment of AI’s near-term economic impact. For workers in the tech sector, a market downturn could mean hiring freezes and layoffs. The disconnect between expectations and reality that Edwards identifies may force a more realistic assessment of AI’s timeline for delivering transformative value.
Recommended Reading
For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources: