Analysis
Two exit paths are open to the biggest AI companies right now, and they're moving at very different speeds. Anthropic filed confidentially on June 1 and won't price until November at the earliest -- nearly six months of regulatory process, roadshow prep and market-timing risk before anyone sees a return. Stripe closed its purchase of OpenRouter and SpaceX closed its purchase of Anysphere in the same six-week window, putting $67.5 billion through the system while Anthropic's bankers were still building the data room.
As an early-stage check-writer, that speed differential changes how I think about backing companies with acquisition potential versus IPO ambitions. A strategic acquirer with cash or stock to deploy can close a deal in weeks once terms are agreed; an IPO candidate is at the mercy of SEC review timelines, a six-month confidential-to-public window, and whatever the Treasury market and Nasdaq are doing on pricing day. I'd rather have a portfolio company be attractive to Stripe or SpaceX than be waiting on a calendar date set by underwriters and market conditions neither the company nor I control.
“As an early-stage check-writer, that speed differential changes how I think about backing companies with acquisition potential versus IPO ambitions.”
Room for disagreement: the IPO path has a ceiling M&A usually doesn't. Anthropic's forecasted $1.14 trillion market cap dwarfs any realistic acquisition price a single strategic buyer could pay, and public liquidity lets early investors sell down gradually rather than accept whatever terms an acquirer offers in a single negotiation. If you're building something with genuine platform ambitions -- not just a feature another company wants to bolt on -- the IPO ceiling is worth the six-month wait and the market-timing risk that comes with it.
My honest answer: I want portfolio companies good enough that BOTH paths are live options at exit, and I structure term sheets accordingly. The moment a founder tells me acquisition is the only realistic outcome, I ask why -- sometimes it's a good answer, but often it just means the company never built the standalone scale an IPO requires.