The bears are waking up (again)
The chorus of equity market bears is growing louder. This time, they’re suggesting the emergence of new, more energy-efficient artificial intelligence (AI)-powered large language models (LLM) combined with massive investment in the infrastructure required for legacy models will result in oversupply and crashing prices akin to that seen during the development and life cycles of previous technologies.
Should you listen to them?
Here’s an executive summary of the arguments gaining a little traction amongst equity market ‘tech’ bears.
According to a growing cohort of bears, the AI theme has once again gripped investors, media outlets, and tech experts, who believe the advent of cheaper and more efficient versions of trained AI-driven inferencing LLMs will result in everlasting growth of demand for chips, bandwidth and storage.
According to the bears, the promised breakthroughs feel eerily like the heady days of the dot.com bubble. Pointing to a market narrative we’ve seen before, AI developments – particularly those highlighting efficiency and energy conservation – are signals that remind some market veterans of the 1999/2000 “it’ll go up forever” mentality.
I remain more circumspect, having read the 1999 Merril Lynch analyst tome covering the dot.com stocks that claimed the future for those companies was so bright that “almost any price can be paid” for them and a sound return made.
We’re not quite there yet.
The AI “cognitive efficiency” revolution
For consumers and perhaps the world at large, a positive development is that the new wave of AI appears to be mimicking the way our human brains and cellular networks work, conserving energy by activating only the processes needed at the required moment. DeepSeek, the AI platform promoted as leading this charge, is said to integrate principles that echo biological systems:
For example, DeepSeek’s neural network design only uses maximum energy when absolutely necessary, ensuring nothing goes to waste, and so taps into the same resource-optimisation strategies you might find in the cells of living organisms.
And by splitting tasks among specialised pathways – akin to neurons with unique roles – DeepSeek’s approach – it is claimed – routes data through only the precisely required channels at the right time. Attention mechanisms and gating networks decide where processing power goes.
Finally, DeepSeek’s yearning for lower energy usage is claimed to be an early-market signal that efficiency gains may be outpacing practical demand, setting the stage for potential oversupply.
According to the bears, today’s AI gold rush has uncanny similarities to the early 2000s. Back then, compression algorithms in telecom kicked off a euphoric phase. Everyone predicted unending expansion. Instead, that very efficiency unleashed an oversupply of infrastructure, causing valuations to plummet.
To equity market bears, what we’re seeing now with AI feels like a rerun: after a series of astonishing technical leaps, Wall Street enthusiasm might be outpacing sustainable reality. Once the initial hype cools, it’s common to discover that many rosy projections were more sizzle than steak.
- Infrastructure oversupply
According to the bears, data centres and chips will inevitably mimic the fibre-optic overbuild of the late ’90s. Once this new tech is proven, every major player may invest in graphics processing units (GPU) farms and specialised AI hardware, only to find the market saturated and prices dropping.
When new tech saves vast amounts of energy or processing capacity quickly, it often creates a glut. That oversupply doesn’t just hit hardware – labour markets, software platforms, and even hip digital assets like meme coins could find themselves on the wrong side of investor demand.
- Valuation distortions
Those same bears believe the Nasdaq’s current trajectory looks suspiciously like the telecom bubble’s last hurrah. Noting the experience with WorldCom and Cisco (NASDAQ:CSCO) around 2000, when the music stopped, they remind investors that the stocks of companies like Nvidia (NASDAQ:NVDA) may not gently slide lower but instead fall off a cliff. They add that even established chipmakers today (think Intel) are navigating rapidly shifting margins and competitive pressures.
- The efficiency paradox
Finally, the bears note something I have banged on about for a few decades. New technology has a habit of rewarding consumers with a better standard of living but a nasty habit of punishing investors who invested too much in the hope that consumer benefit equates to generous investor returns. Better technology doesn’t automatically lead to better investments. As AI hardware evolves, new algorithms or quantum tech could swiftly outdate today’s solutions, hammering profitability. Efficiency is a double-edged sword: it can help a company stand out in the short term, but it can also open the door to rampant supply gluts and price collapses.
According to the bear camp, if historical patterns hold, infrastructure players will take the first hit. The expansion of and current enthusiasm for the makers and suppliers of AI servers, specialised chips, and data centres would cause valuations to rise too quickly. Then, as oversupply sets in, inventory piles up, and prices drop. It’s the classic boom-bust cycle.
According to the bears, Nvidia’s share price has not gained for more than seven months, a sign that history is repeating itself under investors’ noses.
The bears suggest that once leading software names come under pressure – especially those tied to the AI hype – investors will exit en masse, crushing sector-wide valuations, with the negative wealth effect rippling outward like those from a pebble dropped in a pond, slowing the broader economy.
Of course, the big question might simply be: When? The answer is that it might be happening right now. Despite ongoing advances in AI, the Nasdaq semiconductor (SOX) index has remained unchanged for almost a year. For the bears, this failure to advance is a sign of fatigue.
If the market stumbles, as the bears predict, central banks typically respond with monetary stimulus – think money printing, rate cuts, and quantitative easing (QE)-style balance sheet expansion. This could rescue markets, but it could also fuel inflation fears, making tangible commodities like gold even more attractive. Any infrastructure spending response by China, for example, as policymakers try to jumpstart the economy, could spur commodity markets
The bottom line is that eventually, the bears will be right. No boom lasts forever. The bears believe the next chapter of the AI-driven market might look a lot like the final chapters of the 2000 dot.com story.
Roger Montgomery is the Founder and Chairman of Montgomery Investment Management. Roger has over three decades of experience in funds management and related activities, including equities analysis, equity and derivatives strategy, trading and stockbroking. Prior to establishing Montgomery, Roger held positions at Ord Minnett Jardine Fleming, BT (Australia) Limited and Merrill Lynch.
This post was contributed by a representative of Montgomery Investment Management Pty Limited (AFSL No. 354564). The principal purpose of this post is to provide factual information and not provide financial product advice. Additionally, the information provided is not intended to provide any recommendation or opinion about any financial product. Any commentary and statements of opinion however may contain general advice only that is prepared without taking into account your personal objectives, financial circumstances or needs. Because of this, before acting on any of the information provided, you should always consider its appropriateness in light of your personal objectives, financial circumstances and needs and should consider seeking independent advice from a financial advisor if necessary before making any decisions. This post specifically excludes personal advice.
