The Dot-Com Ghost Haunting AI Dreams

The parallels between 2000 and today feel almost scripted. Then, it was Pets.com and Webvan burning through cash on vague promises of “digital transformation.” Now, it’s AI startups capturing two-thirds of North American and European venture funding while many barely demonstrate viable business models. The Expleo AI Pulse survey reveals French executives are split: confidence in AI’s potential sits at 64/100, yet over half believe a bubble is inflating. Only 66% trust their organization’s ability to deploy AI successfully, a figure that’s dropping fast.

What’s different this time? The current tech giants are genuinely profitable, unlike their dot-com predecessors. But as one French investor noted, profitability doesn’t guarantee immunity from speculation. The “Magnificent Seven” may be earning real money, but they’re also making massive bets on AI infrastructure that might take years to generate returns, if ever. Nvidia’s Jensen Huang insists there’s “no AI bubble”, calling it “the largest infrastructure project in human history.” Yet even he acknowledges we’re merely at the starting line of a multi-trillion-dollar buildout.

When Diversification Becomes a Dangerous Word

The standard French financial advice, “stay invested in your ETF world no matter what”, crumbles when you examine the historical charts. An investor who held MSCI USA from 2000-2010 watched their portfolio bleed red for a decade. The CAC 40’s 12-year slump meant an entire generation of French investors saw their PEA accounts stagnate through their prime earning years.

This reality check has sparked heated debates about diversification strategies. The 100% equities crowd argues that over 15-20 years, stocks always win. But as one commenter bluntly put it: “If everything collapses when you’re 65, you won’t care if it recovers in 20 years.” For a 50-year-old with a two-year-old PEA, the math is merciless. There’s simply not enough time to recover from a prolonged downturn.

The 75/25 stock-to-bond allocation emerges as a psychological lifeline. Investors who maintained this mix through the 2000s crash found the volatility bearable enough to stay the course. But here’s where French-specific constraints complicate matters: within a PEA, your bond options are limited. Many turn to fonds euros (euro-denominated funds) for stability, though recent performance has been underwhelming. Others explore hedged foreign bonds, introducing currency complexity that not every investor wants.

The “Enshittification” Factor Nobody Talks About

A fascinating undercurrent in French investment circles is the concept of “enshittification”, the observation that companies are boosting profits by charging more while delivering less quality. This isn’t just consumer complaints about subscription creep, it’s a systemic concern about earnings sustainability. If corporate profits are artificially inflated by squeezing customers rather than genuine innovation, the entire market’s valuation foundation becomes suspect.

This matters enormously for French investors because it challenges the assumption that today’s strong earnings justify high valuations. The Livret A rate dropping to 1.5% has already pushed savers toward riskier assets. If those assets are built on unsustainable profit extraction, the correction could be severe.

Infrastructure Reality vs. Financial Fantasy

The RAM shortage crisis illustrates the tangible impact of AI speculation. Three companies, Samsung, SK Hynix, and Micron, control 95% of DRAM production. They’re prioritizing high-margin AI memory over consumer products, causing RAM prices to triple or worse. A smartphone that cost €500 will soon hit €600. Dell has raised prices €55-765 depending on configuration. Even Apple is negotiating emergency supply deals at 100% premiums.

This isn’t abstract financial engineering, it’s real resource allocation with real economic consequences. The hyperscalers are absorbing production capacity for data centers, leaving consumer electronics manufacturers scrambling. Some analysts predict the shortage could last until 2029.

For French investors, this creates a bizarre scenario: the AI revolution might be real enough to cause component shortages, yet speculative enough that most AI projects fail. Gartner estimates 90% of AI initiatives will flop, burning through billions in capital expenditures that could have been returned to shareholders.

The DCA Delusion and PEA Realities

Dollar-cost averaging (DCA) gets touted as the universal solution, “the weapon against recession.” The theory: keep investing monthly through the downturn, and you’ll profit before prices recover to your initial entry point. But this assumes psychological fortitude that most investors lack. As one French investor candidly admitted: “Very few people are crazy enough to keep investing regularly in a declining asset for years while losses prevent them from sleeping.”

The PEA structure adds another layer of complexity. Early withdrawals before five years trigger account closure and tax penalties. This lock-in feature, designed to encourage long-term thinking, becomes a trap if you need liquidity during a prolonged downturn. The MSCI World PEA ban discussion highlights how regulatory risks can suddenly limit your investment options, making diversification within French tax wrappers even more challenging.

What History Actually Teaches Us

Those who DCA’d through the “lost decade” from 2000-2010 saw 5x returns 15 years later. But this statistic masks crucial nuance: the recovery was uneven, and many investors capitulated before the rebound. The Morningstar data on active fund underperformance shows that even professionals struggle to time recoveries correctly.

For French investors approaching retirement, the math is particularly stark. The FIRE movement’s growth in France reflects deep skepticism about traditional pension systems, but the tax trap on investment income means you need far more capital than simple calculations suggest. A €1 million portfolio might generate only €2,000 monthly after French taxes and social charges, not the €3,000 many expect.

Protecting Your PEA in Uncertain Times

So what’s a French investor to do? The consensus isn’t to sell everything and hide in a Livret A, especially with rates at 1.5%. But blind faith in “stocks for the long run” ignores demographic realities, climate change pressures, and geopolitical fragmentation that make this era genuinely different.

• 1. Stress-test your horizon: If you’re 50 with a new PEA, you’re not a long-term investor in the traditional sense. Your strategy should shift toward capital preservation sooner than standard advice suggests. The pension deficit crisis means you can’t count on state support to bail out investment mistakes.
• 2. Diversify within constraints: Use your PEA for French and European equities, but consider a CTO (Compte-Titres Ordinaire) for global diversification and alternative assets. The 75/25 stock-bond split remains psychologically sound, but explore hedged international bonds rather than limiting yourself to French obligations.
• 3. Focus on quality over hype: The dot-com crash destroyed speculative names but quality companies recovered. Within your PEA, prioritize firms with real profits, reasonable debt, and sustainable competitive advantages. Avoid the French tech stocks trading at 50x sales based on AI promises.
• 4. Plan for volatility, not just returns: The RAM shortage shows how AI demand creates real supply chain shocks. This isn’t 2000’s fiber-optic glut, it’s resource competition that could persist. Build cash reserves outside your PEA for opportunities when (not if) the correction comes.

The Verdict: Bubble or Infrastructure?

The truth likely sits between Jensen Huang’s infrastructure vision and bubble warnings. AI will transform industries, but not every AI stock deserves its valuation. The dot-com bubble gave us Amazon and Google, but it also vaporized billions in dumb money.

For French investors, the key is recognizing that your PEA is a tool, not a religion. The tax advantages are valuable, but not at the cost of appropriate diversification. The “lost decade” happened before, and it can happen again, especially when valuations ignore fundamentals.

The RAM shortage, executive skepticism, and 90% project failure rates aren’t signs of a healthy, sustainable boom. They’re warning lights that the market has gotten ahead of reality. Diversification, realistic time horizons, and a healthy skepticism toward hype aren’t just prudent, they’re essential survival tools for whatever comes next.