Atman Monthly Update — March 2023
Dear Atman Capital subscribers,
We hope this update finds you well. We want to discuss the recent Silicon Valley and Signature Bank failures and provide an update on our investments and how we manage risk at Atman.
Given banking redundancies built in early Q1, Atman had no direct exposure to the crisis and only one portfolio company with direct SVB vulnerability. They managed to rescue all their capital after the swift FED and FDIC intervention.
The source of our SVB paranoia came from Byrne Hobart’s February 23rd newsletter. We were fortunate to have anticipated the bank run two weeks before it happened. You can find the full excerpt that raised our eyebrows at the end of our letter.
As you know, Atman is a firm built on a community of inevitable founders. We call it the Atman Egregore, a safe place for our CEOs to discuss anything, align interests, and connect via online and offline events, occasionally inviting guests to speak. Once we posted this inside our group and many founders responded that they migrated from SVB or were in the process, we knew there was a true risk of a bank run and acted accordingly.
We anticipate a reform in the regulatory measures that followed the 2008 GFC since the flow of information and digital transformation now allow for digital bank runs, which can put the entire system at risk. As issues arose about the safety of customer funds, $42B was transferred from SVB in 24 hours, probably making it the largest “digital bank run” in history.
After the FED intervention, two primary sources are available for US Banking, as Naval puts it:
There are now swap lines on the US Banking deposit base of $17.6T. The FED balance sheet is $8.4T, effectively transforming the FDIC into the FED, which some call quantitative easing into infinity. Time will tell the systematic issues of this policy and the FED’s fight against inflation. One thing is certain: more regulation will come as trust has been temporarily eroded, and no business can afford not to have the certainty to access their deposits.
What happens next to the early-stage venture market?
Since April of 2022, we have advised all our startups to stop focusing solely on growth but rather build a cash position for 40+ months with a plan for profitability.
Many startups received preemptive rounds that shouldn’t have taken place and now hang on without product market fit, a path to sustainable growth, and have overhired at a premium price.
The fourteen years of low-interest rates and the enormous amount of money printed during COVID-19, combined with overfunding of sub-par companies, built the base for this scenario for what will come over the next three years.
We are still in the early stages of the havoc.
A large cohort of companies raised enough in 2019–2022 to have 2–4 years of runway, but unlikely will grow into the priced valuation of the previous round. Pedro Sorrentino was live on Brazil Journal (Portuguese), anticipating that 70% of startups that raised in the past three years will likely fail. We passed on several investment opportunities that raised rounds at ludicrous prices and now have no traction to find additional capital but still refuse to adjust to lower prices.
Flat or down rounds are the new black.
While this shift may concern some, it presents opportunities for early-stage investors like Atman. Increasing demand for early-stage investments will offer attractive returns in the coming months.
The Inevitable CEO of today is now also responsible for understanding macroeconomic risk and properly building a basic treasury structure for their startup. The bar continues to get higher, and we are prepared to take advantage of this new scenario as we have a fresh early-stage fund that has no exposure from the noise of 2018–2022.
We remain excited to continue to partner with Inevitable people and remain out of the esoteric noise. We are approaching the beginning of the bottom of the cycle as companies that overhang without product-market fit and bloated teams go through this necessary correction.
Excerpt from Byrne Hobart’s February 23rd newsletter:
There are two basic attitudes toward risk:
If you’re going to take it in one area, you want to compensate by being conservative everywhere else; Silvergate, the connection between crypto companies and the regulated banking world, put its balance sheet into fairly conservative assets, for example.
Risk tolerance is expressed everywhere, and the organization learns to live with high variance.
Silicon Valley Bank leaned a bit towards the latter: their deposits were always heavily weighted towards VC funds and the companies they backed, so their deposit growth was basically a lagged function of venture fundraising (now at a nine-year low ($, WSJ)). Their assets also leaned in this direction, with exposure to venture debt and with warrants to capture the upside. But they also took another less relevant risk: putting almost half of their assets into long-date mortgage-backed securities ($, FT). This doesn’t have meaningful credit risk, but it does have significant rates risk, and the result is that they have an unrealized loss of $15.1bn as of last quarter. That compares to equity of $16.3bn. And in the quarter before that, unrealized losses were $15.9bn against equity of $15.8bn.
This is a pretty striking occurrence: on a mark-to-market basis, they were broke last quarter, albeit still liquid. And that liquidity matters; one reason banks aren’t required to mark assets to market is that they can hold them indefinitely as long as they have deposits. On the other hand, Silicon Valley Bank’s deposits are less sticky than other companies’, since they’re provided by mostly money-burning companies. The lack of mark-to-market accounting is a reflection of general banking reality, that banks don’t get explicitly margin-called by customers. They do sometimes experience runs. It would take an absolutely titanic bank run to actually impair the company’s liquidity, so a run is unlikely. And even if the company did run into trouble, there are good political reasons to think that depositors wouldn’t be harmed: the people who donate the legal maximum to political campaigns are disproportionately likely to bank with Silicon Valley or work for companies that do.
On the other hand, the tech world is more risk-averse than it used to be, and depositing money into a bank that’s still levered 185:1 on an asset base that includes, among other things, “Premium wine”-backed loans in an amount almost equal to last quarter’s marked-to-market equity. No one wants to look paranoid by being the first to move their money out, but no one wants to deal with the consequences of being last. And if money does flow out, they eventually have to start selling assets, which turns an “unrealized losses” footnote into a headline loss number.