Tech VCs: blockchain shock bounces funds towards AI
US venture capitalists hope a rising tide will save them from admitting some investments are stuck in the mud.
The valuations of public tech stocks are jumping amid excitement about generative artificial intelligence. But private markets are yet to recover. Start-ups are eking out cash to avoid down rounds. That leaves billions of dollars of potential investment sitting on the sidelines.
The past decade has been very prosperous for VCs. Low rates drove more money into risky, early-stage investments. Last year set a new record for VC funds. They raised close to $163bn, according to PitchBook data. Tiger Global, for example, closed a $12.7bn fund. Global venture firms had close to $586bn of unallocated capital, so-called “dry powder”, by October last year.
The tech downturn and collapse of Silicon Valley Bank have made it harder for fund managers to invest or secure capital this year. Distributions to external investors — known as “limited partners” — are down. Exits are scarcer.
PitchBook data shows US VC firms raised less than $12bn across 99 funds in the first quarter of the year. That is down nearly three-quarters on the previous year.
In the aftermath of the dotcom bust, a handful of US venture capital firms took the unusual step of reducing fund sizes. Groups such as Accel Partners released investors from some capital commitments.
Limited partners should not expect a repeat performance. Management fees give VC funds too big an incentive to put money to work. One option is to redirect funds away from out-of-favour crypto start-ups towards the boom in AI.
Change is already under way. Paradigm has expanded the focus of its fund. Peter Thiel’s Founders Fund is reported to have cut the size of its flagship fund. But it plans to roll money pledged into a future fund, not release its claim.
The pain of crypto start-ups may eventually become a gain for AI specialists. It does not quite amount to a bait and switch by VCs. “Bait, wait and reorientate” might be a fairer description.
Read the full article Here