According to Business Insider, Evercore ISI’s chief equity strategist Julian Emanuel is warning that massive debt accumulation among AI companies could trigger a market correction similar to the dot-com bubble. Tech giants like Amazon, Meta, and Oracle are leading a global surge in bond issuance in 2025 to finance rapid AI expansion. Emanuel notes the parallel to the late 1990s when companies built out internet infrastructure using debt without corresponding revenues. While he remains mostly bullish on AI, he describes current investor fears as “rational” given the uncertain payback timelines. The strategist, who previously predicted market corrections and called government bonds the “greatest bubble ever” in 2019, sees particular risk as companies no longer benefit from low interest rates.
Deja Vu All Over Again
Here’s the thing that should make everyone pause: we’ve seen this movie before. Emanuel’s comparison to the dot-com era isn’t just casual chatter – it’s coming from someone who’s built a career spotting bubbles. The pattern is eerily familiar: massive infrastructure build-out financed by debt, uncertain revenue timelines, and companies racing to outspend competitors. Sound familiar? It should. The difference this time is that interest rates aren’t at rock-bottom levels anymore, which means servicing this debt could become painful if AI revenues take longer to materialize than expected.
The Debt Reality
So what’s actually happening? Companies are basically borrowing like there’s no tomorrow to fund AI infrastructure. We’re talking about hyperscalers – the big cloud and tech players – leading a bond issuance frenzy. They’re building data centers, buying GPUs, and scaling operations at a pace that requires serious capital. The problem? Nobody really knows when this spending will translate into actual profits. And with interest rates where they are today, the cost of carrying all this debt is substantially higher than it was during the zero-rate era. It’s a high-stakes gamble that assumes AI will deliver returns quickly enough to cover the borrowing costs.
Bullish Despite Concerns
Now here’s where it gets interesting. Despite all these red flags, Emanuel remains mostly optimistic about AI. Why? Valuation perspective. He points out that compared to the late 1990s bubble, current AI names aren’t ridiculously expensive relative to their own history. They’re not cheap, but they’re not in nosebleed territory either. This creates what he sees as compelling opportunities for investors who can handle some volatility. Basically, the risk might be worth the potential reward if you’ve got the stomach for it. But let’s be real – that’s exactly what people said during the dot-com boom too.
What It Means For Investors
Look, the warning signs are there. When a strategist who correctly called previous bubbles starts drawing comparisons to 1999, you should probably pay attention. The combination of heavy borrowing, uncertain revenue timelines, and higher interest rates creates a perfect storm if AI growth slows even slightly. Companies that rushed to scale might find themselves stuck with expensive debt and delayed returns. But here’s the counterpoint: unlike many dot-com companies that had literally zero revenue, today’s AI players are often established tech giants with other profitable businesses. The question is whether those existing profits can carry the weight of AI ambitions long enough for them to pay off. Only time will tell if this debt-fueled AI expansion ends in triumph or tears.
