Matt Frankel

So there's a lot to like there on the bull side.

There are several bear cases other than the valuation and the general hype surrounding the IPO.

Customer concentration is huge, which Rachel can touch on in a bit.

It's important to mention that NVIDIA is developing competing products, and NVIDIA has pretty much an unlimited R&D budget, which would scare me investing in any competitor of NVIDIA.

There's a lot of future growth priced in here, and there's more that can go wrong than a lot of investors seem to think, especially if you buy the $50 billion valuation.

The short answer is a lower valuation.

But the longer version, I'll almost never buy an IPO stock on day one.

And Cerebrus is not going to be an exception.

I don't see myself buying Cerebrus on days two, three, or four either.

All of the recent AI adjacent IPOs, at least the ones I can think of, have been turbulent at best, even when they performed well.

And in most cases, they've been terrible performers.

Think companies like Figma.

The hype is too high for what I would consider a rather speculative stock at this point.

So I'll be staying away.

But to really answer your question, if they kept these growth numbers alive and the valuation went down to something that was a little bit more palatable, that's probably what would get me interested.

Yeah, so I've certainly trimmed positions in the past, especially those I already consider to be richly valued, if I saw a compelling opportunity.

A long time ago, I trimmed half of my Apple position because I thought it was richly valued and it was getting a little overheated.

But it's usually to sell...

a stock in one industry and buy something undervalued in another, not to do what I would essentially call diversifying your AI trade, which is what you're talking about with some of those big investments there.

A word of caution for me on that strategy especially would be the tax implications of selling appreciated stock.