Andreessen Horowitz raised $15 billion in January. Kevin Hartz raised $450 million this week.
He says that's the whole point.
A-Star is the early-stage firm Hartz co-founded with Bennett Siegel, who used to invest at Coatue. The firm just closed its third and largest fund.
It is still small on purpose. Hartz thinks buying startups in bulk is the wrong way to do venture.
Hartz is best known as a co-founder of Eventbrite, the online ticketing firm. He has been an active early-stage investor for years.
A-Star will deploy the $450 million over close to three years. It will write $3 million to $5 million checks into 30 to 40 seed firms.
The goal is 10% or more ownership in each one. If A-Star hits that mark, it won't back a rival.
Compare that to the megafund model. Bigger funds spread money across more bets, more stages, and more rival firms.
They hope a handful of winners pull the whole portfolio up. Hartz said his peers are "kind of like buying in bulk."
Siegel put it another way. Just because you back a firm doesn't mean you keep writing checks all the way to the IPO.
The wager is that being picky beats being everywhere when the goal is outsized returns.
If you want a read on which venture bets are paying off, Market Briefs covers it every weekday morning - and you get a free investing class when you join.
A-Star's first fund posted a 36% net IRR as of the end of last year. IRR - or internal rate of return - is the profit per year on the money invested.
A 36% net IRR puts that fund in the top 5% of all such funds. The second fund is at 58%.
It is fueled by stakes in firms like AI training shop Mercor and corporate spend tool Ramp, which hit a $32 billion value late last year.
Miles Dieffenbach runs the investment office at Carnegie Mellon's endowment. He said his team prefers "smaller funds that are more focused and concentrated."
His take on A-Star at a bigger size: "We would be very worried if this fund was a $1 billion fund size."
A-Star hasn't backed OpenAI or Anthropic. Hartz said those firms were already too well-funded by the time A-Star had a shot.
The firm has bet instead on the layer of firms built on top of the big AI models. That includes Decagon, a customer service startup recently valued at $4.5 billion.
There's a quiet split in venture right now. The biggest names are raising bigger funds to back the same handful of AI giants.
Smaller firms like A-Star are betting that the real returns sit one layer down. The bet is on the firms using the models, not the firms building them.
A-Star's two earlier funds say that bet has been right so far.
The next $450 million is what proves whether saying no still pays.
If you want to keep tabs on where the smart money is going, join 350,000+ investors reading Market Briefs - the daily email also includes a 45-minute investing course as a bonus.
