Justin Ernest just rewrote the playbook on venture capital. Instead of spending a year glad-handing institutional investors to raise a traditional fund, the Sabertooth VC founder deployed nearly $400 million into some of the hottest startups on the planet—Anthropic, Anduril, and SpaceX—using a network of special purpose vehicles (SPVs) and a captive pool of limited partners. It's a strategy that's turning heads in an industry where the traditional fundraising cycle can take 12-18 months and where access to unicorn deals is increasingly locked behind club doors.
The venture capital world runs on a familiar rhythm: spend months pitching a fund thesis, close on commitments, then deploy capital over 3-5 years. But Justin Ernest wasn't interested in waiting. When opportunities emerged to invest in Anthropic, Anduril, and SpaceX—companies that rarely open their cap tables—he had a different idea.
Instead of raising a traditional fund through Sabertooth VC, Ernest assembled a network of limited partners willing to move fast on individual deals. Through special purpose vehicles, he's now deployed close to $400 million into some of the most competitive startup rounds of the past few years, according to TechCrunch.
The SPV model isn't new, but Ernest's scale is unusual. Most investors use SPVs for one-off opportunities or to syndicate overflow allocation from their main fund. Ernest flipped the script entirely, building what amounts to a venture platform without the traditional fund structure. His LPs commit upfront interest, then he surfaces deals as they emerge. When a hot opportunity surfaces—say, an Anthropic secondary round—he can move within days rather than months.
This matters because timing has become everything in venture capital. The rise of AI has compressed deal cycles to a fraction of their historical norms. Companies like Anthropic went from stealth to multi-billion dollar valuations in under three years. Traditional funds, locked into annual partnership meetings and lengthy diligence processes, often miss the window. Ernest's approach treats speed as a feature, not a bug.
But there's a trade-off. Traditional venture funds build long-term relationships with founders and follow-on across multiple rounds. SPV investors typically get one bite at the apple. They're also paying management fees and carry on a per-deal basis, which can be less efficient than a pooled fund structure. For LPs, it means more administrative overhead—instead of one K-1 form per fund, they're managing documents for every individual investment.
Ernest's background offers clues to how he pulled this off. Before Sabertooth, he spent time at Playground Global, an early-stage deep tech fund that pioneered studio-style venture models. That experience exposed him to alternative structures and the kind of high-net-worth individuals who value access over traditional fund economics. His LP base reportedly includes tech executives and family offices looking for exposure to companies that rarely hit the public markets.
The portfolio speaks for itself. SpaceX remains one of the most sought-after private investments in the world, with secondary shares trading at massive premiums when they're available at all. Anduril, the defense tech unicorn founded by Palmer Luckey, has become a darling of investors betting on the militarization of AI and autonomous systems. And Anthropic, OpenAI's chief rival in the large language model race, has attracted capital from Amazon, Google, and top-tier VCs at valuations north of $15 billion.
Getting into any one of these deals requires relationships, timing, and often a willingness to accept unfavorable terms. Getting into all three suggests Ernest has built the kind of dealflow network that most traditional funds spend years cultivating. The SPV model, counterintuitively, may have helped. Founders and existing investors sometimes prefer SPVs because they consolidate multiple small checks into a single line on the cap table, reducing administrative complexity.
Still, there are questions about sustainability. Can Ernest maintain LP enthusiasm if returns take years to materialize? Will he eventually need to raise a traditional fund to compete for lead positions and board seats? The SPV strategy works brilliantly for access investing—getting into hot late-stage rounds where the heavy lifting is already done—but it's less clear how it scales if Sabertooth wants to move earlier in company lifecycles.
The broader venture industry is watching closely. Founders Fund, Andreessen Horowitz, and other top firms have experimented with flexible fund structures, but most still anchor around traditional 10-year vehicles. If Ernest's model proves durable, it could accelerate a shift toward more modular capital deployment, especially in sectors like AI where deal velocity shows no signs of slowing.
For now, Ernest has demonstrated something important: in an era where the best deals move at internet speed, the ability to deploy capital quickly matters as much as the capital itself. His $400 million bet is that the future of venture looks less like patient institution-building and more like a network of agile, deal-specific partnerships. Whether that thesis holds long-term will depend on exits, returns, and whether LPs stay hungry for the next Anthropic—even if it means filling out a few extra tax forms.
Ernest's nearly $400 million SPV experiment represents more than just an alternative fundraising path—it's a bet that venture capital's future belongs to those who can move fastest. As AI companies compress traditional startup timelines and unicorn rounds close in days instead of weeks, the ability to deploy capital at speed becomes a competitive advantage in itself. Whether this model can deliver the kind of returns that justify its administrative complexity remains to be seen, but for LPs willing to trade convenience for access to Anthropic and SpaceX, Ernest has proven there's more than one way to play the venture game. The real test comes when these investments start exiting and the industry can finally compare SPV economics against traditional fund performance.
