Stop wondering whether we’re in an AI bubble and sell 20%

If I could only offer employees one piece of advice during this AI hype cycle, it’s this: if and when you’re presented with the opportunity, sell 20% of your holdings in a tender offer.

We are probably in an AI bubble. If you work at a startup, that means tremendous opportunity and tremendous risk.

In an interview with the BBC yesterday, Sundar Pichai said the AI boom has some “irrationality” and that investment cycles can “overshoot.” When asked about the bubble bursting, he said, “I think no company is going to be immune, including us.”

Derek Thompson recently wrote a great essay where he makes what I think is a smart prediction:

“Some people think artificial intelligence will be the most important technology of the 21st century. Others insist that it is an obvious economic bubble. I believe both sides are right. Like the 19th century railroads and the 20th century broadband Internet build-out, AI will rise first, crash second, and eventually change the world.”

What does this mean if you work in tech, particularly privately held startups?

You want some exposure to these rapidly appreciating companies. You probably don’t want 80%, or in many cases 100%, of your net worth in one illiquid stock. This is basically always true, but it is particularly critical today.

Unfortunately, most startup employees I talk with nod along that we’re in a bubble overall, but think they work at the one exception that will continue its unabated march to a $100B market cap, if not higher.

I don’t think enough people in the startup world have internalized this: there will be companies privately valued above $10B that go to zero.

I’m worried many employees are going to see millions on paper evaporate, and wish they had sold 20%.

If you are a VC and a few decacorns go to zero while one becomes worth $100B, that’s a great outcome. The winners pay for the losers. If you are an employee with 95% of your net worth tied up in a company that fails, that is quite obviously a disaster.

You don’t need a very long memory to conjure cautionary tales. Between 2020-2022, valuations soared in a frothy market before they came crashing down.

If you had perfect information, you would wait until right before the AI bubble pops to sell. But you don’t have perfect information, and when the music stops, the opportunities to sell private shares often dry up completely. There are many employees who passed up opportunities to sell their equity between 2020 and 2022, and haven’t had an opportunity to sell since.

Why 20%? It’s often the max you’re allowed to sell in a tender offer, but I think companies have landed on that maximum because it’s a healthy tradeoff: it’s enough to be a meaningful amount of money, but little enough that you still have the majority of your stake invested in the company.

One of the main reasons Prospect shows our company projections as a range of outcomes is that it’s helpful for framing tradeoffs like this. If you sell 20% and your company keeps going up and to the right, your remaining 80% (plus any refresh grants you get) will be worth millions, and you’ll be pretty happy. If your company loses value, you’ll be relieved you at least sold 20%.

A few years ago I sold a small percentage of my Rippling shares in a tender offer. The price has more than doubled since, but I don’t think back on that decision at all.

I bet if the price had gone down and I hadn’t sold, I would think about it very often.