One investor expects billions in returns while another warns the same company's IPO could trigger a market correction—welcome to the split-screen reality of AI valuations in late 2024.

The Summary

The Signal

Hussain Sajwani's confidence in his Anthropic stake represents the bull case playing out in real money. The Dubai real estate magnate went deep on AI investments over the past few years, and now those bets are converting to paper billions. His timing reflects the pattern of sophisticated capital flowing into foundation model companies during the 2022-2023 window, when valuations were high but not yet stratospheric.

Jack Selby at Copper Sky sees the same IPO through a different lens: lockup risk. When Anthropic goes public, early investors get liquidity. When early investors get liquidity after a vertical run-up in AI valuations, some of them sell. When enough of them sell, markets reprice the entire sector.

"The blockbuster IPO may become the catalyst for major reversal in the artificial intelligence trade."

This is the classic late-cycle tension:

  • Founders and early backers want to realize gains
  • Later investors need continued momentum to justify entry prices
  • Public market buyers inherit the valuation question with none of the early upside

The wedge here is what Anthropic's IPO actually tells us. If Sajwani is right, we're still early in the infrastructure build-out and foundation models remain underpriced relative to the economic value they'll capture. If Selby is right, we're at the top of the hype cycle and about to learn that training models at scale doesn't automatically translate to durable moats or profit margins.

The Implication

Watch what happens in the 90 days after Anthropic's lockup expires. If insiders hold, the bull thesis strengthens. If they distribute and valuations hold, the market is pricing in real earnings power. If they sell and the sector reprices downward, you'll know which investors understood the difference between a good company and a good price.

For anyone building in the agent economy: your cost structure depends on foundation model pricing. If Anthropic and peers face margin pressure post-IPO, that flows downstream to you as cheaper inference. A correction might hurt portfolios but it accelerates adoption.

Sources

Bloomberg Tech | Bloomberg Tech