Shaun Maguire on Sequoia's new $950M funds, defense tech timing, and why VC isn't an asset class

Summary

Sequoia Capital is launching two new funds totaling $950 million: Venture Fund 19 at $750 million and Seed Fund 6 at $200 million. The seed vehicle is deliberately separated from the main venture fund to signal a sharper emphasis on early-stage conviction. A growth fund operates on a separate cycle and was announced independently. Shaun Maguire, a Sequoia partner who joined six years ago, frames this vintage as the firm's first where a post-generational-transition team has been working together for five-plus years, naming partners including Stephanie, Constantine, and growth-side names Pat and Andrew as examples of that cohort hitting stride.

Fund sizing discipline is explicit. Sequoia targets a minimum 10x net return per fund and acknowledges it got pulled into faster deployment during the COVID era. The reset is deliberate: fewer companies, deeper involvement, and a stated preference for founder-and-market rather than founder-or-market.

Venture Capital's Structural Problem — and Sequoia's Answer

Maguire endorses Roelof Botha's framing that venture capital is not an asset class, agreeing that passive, late-stage VC generates reward-free risk. His counter is that Sequoia operates differently: the firm is often involved before incorporation, helping assemble founding teams and shape initial strategy. He cites David Kahn structuring Kela, an Israeli defense tech company, by recruiting co-founder Homotom Meridor to pair with founder Alon Drawer, as a recent example. Jim Goetz incubating Palo Alto Networks and Sequoia working out of the same office as early Airbnb and YouTube teams are cited as historical precedents.

The honest admission is that Sequoia passed on Coinbase partly because buying Bitcoin directly looked like a cleaner trade — and Bitcoin did outperform Coinbase at most points in time. The broader point is that in mega-capital technology cycles, value often accrues to infrastructure incumbents (NVIDIA being the obvious AI-era example) rather than to VC-backed startups.

Defense Tech: Window Closing

Sequoia made three to four defense tech investments in the roughly six-to-twelve month window after Russia's invasion of Ukraine when the opportunity was still non-consensus. Named positions include Neros, Stark, and Kela. Maguire is direct that defense tech has now become consensus, valuations are high, and the alpha window has largely closed. The government shutdown is acting as a real-time filter: companies with must-have capabilities are still receiving informal award signals and commitments pending congressional approval, while nice-to-have vendors are getting hurt materially. Maguire describes it as a clear tale of haves and have-nots.

Space: Macro Bull, Micro Skeptic

Maguire started a failed space launch company in 2007 and has thought about orbital data centers for over a decade. On Varda (Founders Fund-backed), he says he passed early and frames it as the right macro idea with the wrong micro idea — ZBLAN fiber optic cable manufacturing in orbit is redundant given Starlink's long-distance data capabilities. He is nonetheless bullish on Varda's trajectory because talent concentration and operational expertise in space will compound as SpaceX Starship drives launch costs down.

On data centers in space more broadly, two unsolved technical problems dominate: heat dissipation in vacuum (heat cannot efficiently radiate away from a satellite and builds up in a local pocket) and high-energy particle damage to electronics without Earth's magnetic shielding. Maguire considers both genuinely open engineering questions, not marketing problems. Timing risk around Starship launch cadence adds a third variable. He is a ten-to-fifteen-year macro bull on the space economy but acknowledges some current entrants may be too early.

Telecom and SpaceX Direct-to-Cell

SpaceX capturing 30% of global mobile data within a decade is the bull case Maguire models, but he does not think this destroys incumbent telecoms. Total mobile data consumption will roughly double over the same period — driven by always-connected autonomous vehicles streaming high-resolution content, Optimus robots, and the elimination of dead zones — leaving existing carriers with a net increase in traffic even in the SpaceX success scenario.

Founder Skills in the AI Era

On whether founders should learn to code, Maguire expects coding to become analogous to arithmetic — useful mental training but not a core human activity. His recommended cognitive investments are pure mathematics or physics for maximum logical rigor, or alternatively skipping structured education entirely in favor of building social capital and deal-making skills. He sees the ability to marshal capital, customers, governments, and talent simultaneously as an increasingly differentiated founder capability, though he frames it as a resurging archetype rather than a new one.