The VC problem in SV
After recounting all the incredibly whiny VC libertarianism-for-you-socialism-for-me dudes demanding an immediate bailout of SVB, Edward Ongweso ties this to some fundamental problems with venture capital and tech:
The reality, however, is that VCs are herd animals. The industry is overconcetrated—enmeshed, as Geri Kirilova at venture capital firm Laconia Group puts it—and structurally drives capital into a few well-connected hands who pile it into larger funds, cut it into larger checks, and hand it off to a tightly knit network of entrepreneurs and startups. This overreliance on established actors or social networks may seem like a shortcut when you’re risk-averse or unable (and unwilling) to vet every single prospective investment, but it has at times left venture capitalists unable to weed out well-connected or charismatic charlatans.
In a comprehensive case study of the VC industry, UC Davis law professor Peter Lee argues that these are structural deficits that fundamentally undercut venture capital’s ability to actually provide social utility. But venture capital isn’t just wearing blinders. It uses capital as a weapon to crush the competition and corner a market. It works to rewrite laws and regulations, as VC-backed firms tried to do for the gig economy and the crypto industry. Sometimes that means lobbying, as the industry did in the 1970s and 1980s to achieve reforms that cut capital gains taxes, made stock-based compensation attractive, and loosened pension regulations that give VCs access to new funds (and secure a massive subsidy from the government). Being loud and emphasizing their role in creating value has worked for VCs in the past. This past weekend was another example.
What does all of this have to do with SVB? By all accounts, SVB was the beating heart of the valley. In 2015, the New York Times reported that it serviced 65 percent of “all existing start-ups and many of the most prominent venture capital firms.” The bank’s collapse came out of a panic and a bad bet on interest rates, but it got into this situation because everyone involved seems to have helped build a risky system. VCs required portfolio companies to bank with SVB, SVB offered mortgages and wealth management services to VCs, and if SVB offered services similar to other banks serving Silicon Valley, then it likely made the terms of those deals incredibly attractive. First Republic Bank, for example, gave Mark Zuckerberg a mortgage at an interest rate that was below inflation—essentially offering a loan for free.
The risk introduced to SVB by overreliance on low interest rates in both its depositor base and portfolio investments is the same risk embedded in the core of the venture capital model. Profligate fundraising and investment have operated on the assumption that money would be cheap, allowing it to make increasingly exotic bets. Venture capitalists have had a decade of negative to zero real interest rates to build the future through their intrepid noses for value, so what did they give us? We got benefits largely limited to the realm of consumer goods and services, like cheap on-demand delivery and ride-hail (so long as you ignored the exploitation that powered them) and cheap streaming services (until they began hiking prices), namely. But what were the costs? Startups that revolutionized the militarization of our border and our migrant deportation operations, helped weaponize robots, offered A.I. services that exploit invisible underpaid workers in the Global South, and roiled urban transit, rental, and restaurant markets. These projects and others generated billions for investors who got in on an early fundraising round, but they also degraded the quality of life for people across the world.
As he goes on to point out, the pumping up and collapse of crypto is also a highly relevant tale, and I expect more demands for bailouts are coming.
I’m ambivalent about whether big depositors at SBV should have taken a haircut, but I am pleased that the shareholders are going to get wiped out and the executives were fired.