As artificial-intelligence stocks skyrocket and valuations across tech sectors swell, many professional investors are dusting off tactics used during the late-1990s dot-com boom to steer clear of potential busts. While they do not wish to abandon the AI wave, they recognise the dangers of buying at peak frenzy. Thus they’re repositioning portfolios away from the high-flying names that have dominated headlines and moving into assets that could benefit from, or ride behind, the next phase of the trend. This strategic pivot draws heavily on the lessons of the internet era: ride the wave, but don’t get stuck at the crest.
The AI boom has seen key stocks—particularly chipmaker Nvidia—triple or more in value within a two-year span, and the group of what are often called the “Magnificent Seven” dominates U.S. growth indices. In this context, asset managers such as Amundi and Goshawk Asset Management are signaling that the risk-reward balance is shifting. They are looking to replicate the late-1990s approach: participate selectively in the trend, but rotate capital into less crowded, undervalued themes before the broader market catches up.
Why the Dot-Com Era Playbook Works Again
The logic behind revisiting the 1990s strategy lies in the combination of exuberance, tech leadership concentration and the risk of overcapacity. During the dot-com boom, hedge funds realised they didn’t have to short everything; instead they sold excessive valuations in “obvious” names, took profits and shifted into themes or stocks that hadn’t yet been bid up. Research shows that from 1998-2000, nimble funds beat the broader market by roughly 4.5 % per quarter by doing exactly that—riding the upswing but avoiding the crash’s full damage.
Today the parallels are striking. Many AI-heavy companies are investing trillions into infrastructure, chips and data-centre builds—all chasing roughly the same promise of automation and intelligence. The risk of overcapacity, duplication and unrealistic growth assumptions is ever present. That makes the strategy of stepping away from headline players and moving into adjacent or undervalued plays increasingly appealing. It’s not about rejecting AI altogether—it’s about managing participation with discipline.
Instead of simply loading up on the headline names dominating media coverage, investors are seeking “second-wave” or supporting plays—the companies that supply, enable or benefit indirectly from AI mega-investments. For instance, robotics firms, Asian tech manufacturers, software companies servicing AI infrastructure and even uranium miners feeding power-hungry data centres have attracted interest. One manager cited uranium as a favourable bet thanks to the energy demands of large AI compute clusters, while others pointed to Taiwanese precision-parts suppliers tied to chip manufacturing.
At the same time, many portfolio managers are trimming exposure to the top-valued names, realising profit and reallocating into sectors that may be overlooked. This tactical rotation creates a hedge against the very visible risk of a sharp correction in headline AI stocks. For example, one asset manager sold out-of-favour chip names and bought into European industrials or Asian suppliers, reasoning that a broader value-reset might trigger the next wave of gains.
The Risks Behind Riding an AI Bubble
The great unknown, of course, is timing. Calling the peak of a tech bubble is notoriously difficult and often only visible in retrospect. The fact that many companies are investing ahead of proven business models or revenue streams raises concerns. Some strategists warn the current environment already displays classic bubble features: excessive speculation, valuation dislocations, and a convergence of too many companies chasing the same future market size.
Overcapacity constitutes a real threat. Just as the telecom boom in the 2000s saw fiber-optic build-out exceed demand, the AI build-out race could result in excess compute capacity, redundant data-centre expansion and empty hardware pipelines. Should a few large bets fail to pay off, the ripple effects could impact not only tech stocks but broader markets.
This is why the dot-com era playbook emphasises rotation and avoiding crowd-exposed names. The guiding principle: participate in the trend but protect the downside. By stepping into underappreciated segments and trimming overheated names, investors aim to capture upside while limiting catastrophic risk.
What Signals Investors Are Watching
Investors tracking the AI boom are monitoring several key indicators that could signal late-cycle dynamics. These include surging options trading tied to volatile tech stocks, unusually high valuations without corresponding earnings growth, and broadening investor participation—especially among retail investors chasing FOMO (fear of missing out). These signals echo late-1990s patterns when impulsive retail speculation amplified the tech peak.
In addition, supply-chain constraints or bottlenecks—for instance in advanced chips, rare-earth minerals or high-performance computing hardware—are being watched closely. If supply expands too quickly or competition intensifies, margins could collapse, creating structural pressure on many of the firms currently enjoying premiums. Investors are also attuned to regulatory or geopolitical shocks, particularly around China’s expanding role in AI and semiconductor manufacturing.
In practice, portfolio changes are subtle but meaningful. Some fund managers are increasing weights in regional suppliers, software service providers, and alternative energy infrastructure firms—which could benefit indirectly from AI expansion rather than being directly exposed to the headline hype. Others are building hedges: increasing allocations to sectors historically resilient in downturns, such as healthcare, European industrials or value-oriented stocks, as protection against a potential AI correction.
Investors are also maintaining diversification—to avoid becoming overly dependent on the high-flying few. Rather than concentrate in a handful of mega-cap names, the strategy favours a broader constellation of smaller-cap and emerging-market exposures that remain under the radar but have upside potential. By contrast, relying only on the obvious winners carries risk if sentiment shifts.
Why This Matter for Market Dynamics
The revival of the dot-com era playbook highlights how market participants learn from history—and the way signals of excess can recur across different technology waves. It underscores that bubbles, while varying in form, share core features: elevated valuations, broad participation, speculative momentum and the risk of structural correction. The AI wave may be transformational, but its timing, pacing and winners are far from guaranteed.
For investors, the message is clear: embracing innovation is important—but so is risk discipline. The many billions already flowing into AI infrastructure mean that opportunities abound. However, the likely path of returns may be longer, less glamorous and more uneven than the rush seen in the early-2000s internet boom. Adapting the dot-com playbook means owning the trend—but resisting the urge to ride it recklessly.
Through this approach, participants aim to generate meaningful returns while sidestepping the most dangerous excesses. In doing so, they might avoid becoming collateral damage in a generational tech bubble.
(Source:www.reuters.com)
The AI boom has seen key stocks—particularly chipmaker Nvidia—triple or more in value within a two-year span, and the group of what are often called the “Magnificent Seven” dominates U.S. growth indices. In this context, asset managers such as Amundi and Goshawk Asset Management are signaling that the risk-reward balance is shifting. They are looking to replicate the late-1990s approach: participate selectively in the trend, but rotate capital into less crowded, undervalued themes before the broader market catches up.
Why the Dot-Com Era Playbook Works Again
The logic behind revisiting the 1990s strategy lies in the combination of exuberance, tech leadership concentration and the risk of overcapacity. During the dot-com boom, hedge funds realised they didn’t have to short everything; instead they sold excessive valuations in “obvious” names, took profits and shifted into themes or stocks that hadn’t yet been bid up. Research shows that from 1998-2000, nimble funds beat the broader market by roughly 4.5 % per quarter by doing exactly that—riding the upswing but avoiding the crash’s full damage.
Today the parallels are striking. Many AI-heavy companies are investing trillions into infrastructure, chips and data-centre builds—all chasing roughly the same promise of automation and intelligence. The risk of overcapacity, duplication and unrealistic growth assumptions is ever present. That makes the strategy of stepping away from headline players and moving into adjacent or undervalued plays increasingly appealing. It’s not about rejecting AI altogether—it’s about managing participation with discipline.
Instead of simply loading up on the headline names dominating media coverage, investors are seeking “second-wave” or supporting plays—the companies that supply, enable or benefit indirectly from AI mega-investments. For instance, robotics firms, Asian tech manufacturers, software companies servicing AI infrastructure and even uranium miners feeding power-hungry data centres have attracted interest. One manager cited uranium as a favourable bet thanks to the energy demands of large AI compute clusters, while others pointed to Taiwanese precision-parts suppliers tied to chip manufacturing.
At the same time, many portfolio managers are trimming exposure to the top-valued names, realising profit and reallocating into sectors that may be overlooked. This tactical rotation creates a hedge against the very visible risk of a sharp correction in headline AI stocks. For example, one asset manager sold out-of-favour chip names and bought into European industrials or Asian suppliers, reasoning that a broader value-reset might trigger the next wave of gains.
The Risks Behind Riding an AI Bubble
The great unknown, of course, is timing. Calling the peak of a tech bubble is notoriously difficult and often only visible in retrospect. The fact that many companies are investing ahead of proven business models or revenue streams raises concerns. Some strategists warn the current environment already displays classic bubble features: excessive speculation, valuation dislocations, and a convergence of too many companies chasing the same future market size.
Overcapacity constitutes a real threat. Just as the telecom boom in the 2000s saw fiber-optic build-out exceed demand, the AI build-out race could result in excess compute capacity, redundant data-centre expansion and empty hardware pipelines. Should a few large bets fail to pay off, the ripple effects could impact not only tech stocks but broader markets.
This is why the dot-com era playbook emphasises rotation and avoiding crowd-exposed names. The guiding principle: participate in the trend but protect the downside. By stepping into underappreciated segments and trimming overheated names, investors aim to capture upside while limiting catastrophic risk.
What Signals Investors Are Watching
Investors tracking the AI boom are monitoring several key indicators that could signal late-cycle dynamics. These include surging options trading tied to volatile tech stocks, unusually high valuations without corresponding earnings growth, and broadening investor participation—especially among retail investors chasing FOMO (fear of missing out). These signals echo late-1990s patterns when impulsive retail speculation amplified the tech peak.
In addition, supply-chain constraints or bottlenecks—for instance in advanced chips, rare-earth minerals or high-performance computing hardware—are being watched closely. If supply expands too quickly or competition intensifies, margins could collapse, creating structural pressure on many of the firms currently enjoying premiums. Investors are also attuned to regulatory or geopolitical shocks, particularly around China’s expanding role in AI and semiconductor manufacturing.
In practice, portfolio changes are subtle but meaningful. Some fund managers are increasing weights in regional suppliers, software service providers, and alternative energy infrastructure firms—which could benefit indirectly from AI expansion rather than being directly exposed to the headline hype. Others are building hedges: increasing allocations to sectors historically resilient in downturns, such as healthcare, European industrials or value-oriented stocks, as protection against a potential AI correction.
Investors are also maintaining diversification—to avoid becoming overly dependent on the high-flying few. Rather than concentrate in a handful of mega-cap names, the strategy favours a broader constellation of smaller-cap and emerging-market exposures that remain under the radar but have upside potential. By contrast, relying only on the obvious winners carries risk if sentiment shifts.
Why This Matter for Market Dynamics
The revival of the dot-com era playbook highlights how market participants learn from history—and the way signals of excess can recur across different technology waves. It underscores that bubbles, while varying in form, share core features: elevated valuations, broad participation, speculative momentum and the risk of structural correction. The AI wave may be transformational, but its timing, pacing and winners are far from guaranteed.
For investors, the message is clear: embracing innovation is important—but so is risk discipline. The many billions already flowing into AI infrastructure mean that opportunities abound. However, the likely path of returns may be longer, less glamorous and more uneven than the rush seen in the early-2000s internet boom. Adapting the dot-com playbook means owning the trend—but resisting the urge to ride it recklessly.
Through this approach, participants aim to generate meaningful returns while sidestepping the most dangerous excesses. In doing so, they might avoid becoming collateral damage in a generational tech bubble.
(Source:www.reuters.com)
