How to Survive the A.I. Bubble: Lessons from Jeremy Grantham
Strategic Business Thinking
Jeremy Grantham is one of the most respected long-term investors and market historians of the past half-century. As co-founder of GMO, he built his reputation by studying asset bubbles and warning about them before they burst—from the Japanese equity bubble in the late 1980s to the dot-com boom and the U.S. housing crisis.
What makes Grantham different is his focus on valuation and historical patterns. He doesn’t try to predict short-term market moves. Instead, he looks at how far prices have drifted from long-term fundamentals and asks a simple question: are investors paying too much for future growth?
His track record shows a consistent theme. The most dangerous moments in markets are not when ideas are weak, but when they are so compelling that investors stop questioning price.
That perspective is what he brings to A.I.
Jeremy Grantham’s warning on A.I. is often misunderstood.
He’s not saying A.I. is irrelevant. He’s saying something more uncomfortable: a great technology can still be a bad investment if you pay the wrong price.
That’s what makes this moment dangerous.
The biggest bubbles are rarely built on bad ideas. They’re built on great ideas taken too far. Railways changed the world. The internet changed the world. Both also destroyed capital when investors paid any price for the future.
So the real question isn’t whether A.I. is real. It is.
The question is how to participate without becoming the last buyer.
1. Don’t confuse the future with the price
A.I. will transform industries. That doesn’t mean every A.I. company is worth any valuation.
This is the classic mistake: right trend, wrong price.
A company can be part of the future and still be a bad investment if expectations are already priced to perfection.
Before asking how big A.I. can become, ask how much of that future is already in the price.
2. Follow cash flows, not narratives
In a bubble, stories expand faster than profits.
Everyone talks about automation, intelligence, platforms, proprietary data. Some of it is real. A lot of it is packaging.
I reduce everything to economics:
Does it cut costs?
Improve margins?
Increase pricing power?
Drive retention?
Generate cash?
Improve ROIC?
If I can’t answer clearly, I don’t have an investment case.
A.I. should show up in the numbers, not just in the pitch.
3. Avoid anything that requires perfection
At extreme valuations, there’s no room for error.
The company has to execute flawlessly: growth, margins, competition, hiring, capex, regulation. That’s not investing. That’s betting on perfection.
Even if the story is right, the price can still be wrong.
The higher the valuation, the more fragile the outcome.
4. Watch the capex cycle
A.I. isn’t just software. It’s infrastructure: chips, data centers, energy, cooling, talent.
That creates opportunity, but also risk.
When companies start spending defensively—because they’re afraid of falling behind—returns usually suffer.
The key question isn’t how much is being built.
It’s what return that capital will generate.
5. Keep cash as optionality
In a bubble, cash looks lazy. It isn’t.
Liquidity gives you flexibility. It lets you avoid forced decisions and act when prices reset.
Most people don’t just lose money because they buy high. They lose because they have no room to move when the cycle turns.
Cash is boring in the mania. It’s invaluable after.
6. Prefer quality over excitement
Bubbles reward excitement first, quality later.
When things correct, the market goes back to basics: profitability, balance sheet strength, pricing power, discipline.
Quality businesses can fall, but they’re more likely to survive and come out stronger.
The goal isn’t to avoid volatility.
It’s to avoid permanent loss of capital.
7. Don’t short the future
The opposite mistake is becoming blindly bearish.
A.I. is real. Some companies will build exceptional businesses. Some assets will become strategically critical.
The point isn’t to reject A.I.
It’s to reject bad underwriting.
Stay exposed, but stay disciplined.
The Vitruvian view
At Vitruvian, the approach is simple: selective, analytical, execution-driven.
A.I. will reshape industries. But capital discipline still matters.
The winners won’t be the ones talking about A.I. the most. They’ll be the ones using it to improve economics, simplify operations, and build durable advantage.
That’s the real lesson.
Bubbles don’t happen because people believe in nothing. They happen because people believe so much they stop asking what something is worth.
A.I. may be revolutionary.
That doesn’t remove the need for valuation, cash flow, and discipline.
Surviving a bubble isn’t about being pessimistic.
It’s about being selective, liquid, and patient enough to act when the opportunity is real.
BLOG & INSIGHTS
Exploring innovations, strategies, and the future.
How to Survive the A.I. Bubble: Lessons from Jeremy Grantham
Strategic Business Thinking
Jeremy Grantham is one of the most respected long-term investors and market historians of the past half-century. As co-founder of GMO, he built his reputation by studying asset bubbles and warning about them before they burst—from the Japanese equity bubble in the late 1980s to the dot-com boom and the U.S. housing crisis.
What makes Grantham different is his focus on valuation and historical patterns. He doesn’t try to predict short-term market moves. Instead, he looks at how far prices have drifted from long-term fundamentals and asks a simple question: are investors paying too much for future growth?
His track record shows a consistent theme. The most dangerous moments in markets are not when ideas are weak, but when they are so compelling that investors stop questioning price.
That perspective is what he brings to A.I.
Jeremy Grantham’s warning on A.I. is often misunderstood.
He’s not saying A.I. is irrelevant. He’s saying something more uncomfortable: a great technology can still be a bad investment if you pay the wrong price.
That’s what makes this moment dangerous.
The biggest bubbles are rarely built on bad ideas. They’re built on great ideas taken too far. Railways changed the world. The internet changed the world. Both also destroyed capital when investors paid any price for the future.
So the real question isn’t whether A.I. is real. It is.
The question is how to participate without becoming the last buyer.
1. Don’t confuse the future with the price
A.I. will transform industries. That doesn’t mean every A.I. company is worth any valuation.
This is the classic mistake: right trend, wrong price.
A company can be part of the future and still be a bad investment if expectations are already priced to perfection.
Before asking how big A.I. can become, ask how much of that future is already in the price.
2. Follow cash flows, not narratives
In a bubble, stories expand faster than profits.
Everyone talks about automation, intelligence, platforms, proprietary data. Some of it is real. A lot of it is packaging.
I reduce everything to economics:
Does it cut costs?
Improve margins?
Increase pricing power?
Drive retention?
Generate cash?
Improve ROIC?
If I can’t answer clearly, I don’t have an investment case.
A.I. should show up in the numbers, not just in the pitch.
3. Avoid anything that requires perfection
At extreme valuations, there’s no room for error.
The company has to execute flawlessly: growth, margins, competition, hiring, capex, regulation. That’s not investing. That’s betting on perfection.
Even if the story is right, the price can still be wrong.
The higher the valuation, the more fragile the outcome.
4. Watch the capex cycle
A.I. isn’t just software. It’s infrastructure: chips, data centers, energy, cooling, talent.
That creates opportunity, but also risk.
When companies start spending defensively—because they’re afraid of falling behind—returns usually suffer.
The key question isn’t how much is being built.
It’s what return that capital will generate.
5. Keep cash as optionality
In a bubble, cash looks lazy. It isn’t.
Liquidity gives you flexibility. It lets you avoid forced decisions and act when prices reset.
Most people don’t just lose money because they buy high. They lose because they have no room to move when the cycle turns.
Cash is boring in the mania. It’s invaluable after.
6. Prefer quality over excitement
Bubbles reward excitement first, quality later.
When things correct, the market goes back to basics: profitability, balance sheet strength, pricing power, discipline.
Quality businesses can fall, but they’re more likely to survive and come out stronger.
The goal isn’t to avoid volatility.
It’s to avoid permanent loss of capital.
7. Don’t short the future
The opposite mistake is becoming blindly bearish.
A.I. is real. Some companies will build exceptional businesses. Some assets will become strategically critical.
The point isn’t to reject A.I.
It’s to reject bad underwriting.
Stay exposed, but stay disciplined.
The Vitruvian view
At Vitruvian, the approach is simple: selective, analytical, execution-driven.
A.I. will reshape industries. But capital discipline still matters.
The winners won’t be the ones talking about A.I. the most. They’ll be the ones using it to improve economics, simplify operations, and build durable advantage.
That’s the real lesson.
Bubbles don’t happen because people believe in nothing. They happen because people believe so much they stop asking what something is worth.
A.I. may be revolutionary.
That doesn’t remove the need for valuation, cash flow, and discipline.
Surviving a bubble isn’t about being pessimistic.
It’s about being selective, liquid, and patient enough to act when the opportunity is real.
BLOG & INSIGHTS