E120: Banking crisis and the great VC reset

The speakers rejected the idea that SVB was merely a VC-driven panic, arguing that the week exposed broader banking stress from rate hikes and duration mismatches. They distinguished SVB, Signature and Silvergate’s liquidity problems from First Republic’s loan-book pressure and Credit Suisse’s separate confidence shock.

Chamath Palihapitiya argued that SVB’s deepest Silicon Valley issue was conflict of interest, not depositor panic. He described VCs receiving fund investments or credit from SVB and then directing portfolio-company deposits there—relationships he said would face much stricter disclosure in public markets.

The Fed’s emergency lending facility was portrayed as a one-year bridge, not a cure. Palihapitiya said banks could borrow against depressed securities at par, estimated roughly $2 trillion of underwater assets outside the four largest banks, and warned the structure could postpone rather than remove the problem.

The most concrete reform proposed was continuous, software-based bank supervision. Palihapitiya argued regulators should see real-time dashboards of assets, liabilities and duration risk, while the group also floated fee-based “bank vault” accounts that would safeguard deposits without lending them out.

For startups, the practical treasury advice was redundancy: spread cash across institutions and favor highly liquid government-backed instruments. They discussed insured cash sweeps distributing balances in $250,000 FDIC-sized chunks, while Sacks preferred large, Treasury-bill-backed money-market funds over startups managing their own bond ladders.

The venture market was described as undergoing a hard reset across funds, valuations and staffing. Founders Fund was cutting its eighth fund in half, Stripe had taken roughly a 50% valuation haircut, Tiger marked its private book down 33%, and Y Combinator eliminated its late-stage team and 17 jobs.

Jason Calacanis said the downturn is already changing founder quality and funding behavior. His Founder University had made roughly 20–30 $25,000 checks to small teams reaching an MVP, and he said laid-off builders now looked more product- and customer-driven than the hype-heavy founders of the prior cycle.

Ranga Dias’s new paper claimed room-temperature superconductivity at about one gigapascal, but his credibility remained a central problem. His 2020 Nature claim was retracted after failed replication; Palihapitiya said he had nevertheless pursued a spinout two years earlier, until the University of Rochester blocked it.