Mini Series: Why the "Bubble" Narrative is Dead Wrong

20th February, 2026

Comparing today’s AI-driven market to 2000 is like comparing apples to oranges—it just doesn’t work.

1. Real Demand vs Speculation

During the DotCom era, the bubble was built on fantasy. Telecom companies laid massive amounts of fiber assuming demand would explode. It didn’t. Most of that infrastructure sat unused, and the industry collapsed.

Today, it’s the opposite. AI data centers aren’t speculative—demand is real.

• They already consume 4–4.4% of U.S. electricity, and by 2030, that could double or triple.

• Utilization rates show vacancy around just 2%, compared to up to 90% unused fiber in 2000.

The key difference: today’s infrastructure is actively used, not a bet on the future.

2. Profitability Matters

In 2000, companies traded at extreme multiples with little or no earnings. Today:

• Forward valuations are high, but still well below bubble peaks.

• Market leaders are highly profitable, with strong balance sheets, real cash flow, and growing earnings.

High valuations backed by real profits are completely different from paying up for hope.

3. Leverage and Interest Rates

Margin debt is high in nominal terms, but as a share of market cap, it’s below historical peaks.

• Today: <2% of the S&P 500

• Long-term average: ~2.5%

• DotCom peak: ~3%

Macro conditions are also different. Pre-DotCom, the Fed aggressively raised rates from 5% to 6.5% in under a year. Today, we are in an easing cycle, supporting risk assets.

4. The New World Order

One of the most important shifts: AI adoption is now essential, not optional.

With aging populations and slow workforce growth, productivity gains are critical—and AI is the most powerful lever available.

This has created a new kind of Cold War, not defined by ICBMs, but by AI and compute power.

• The U.S. and China are competing aggressively for AI leadership.

• Initiatives like the Genesis Mission in the U.S. show urgency reminiscent of the Manhattan Project.

• In this environment, advanced semiconductors have become the new oil.

Demand is broad-based—spanning consumers, enterprises, and governments—and it’s accelerating.