Public v Private Debate, Space, Tariffs

Private v Public Debate

Bloomberg columnist Matt Levine published a great commentary this week on the ongoing public vs private market tension. The punchline is, “And if an increasing share of the economy — and particularly of the fastest-growing and most valuable firms — is in private assets, investors can’t just index. You can’t get “the market return” by buying every public stock if half of the market is in private companies. You have to buy a mixed public-private fund, and pay for it.”

Matt Levine is referring mostly to late stage VC and certain PE assets. There are many late stage VC assets that are teetering on the edge of going public like Stripe, SpaceX, DataBricks, for example. If those three companies were publicly traded their most recent market cap would put them in the top S&P 100. OpenAI just raised at a $300B valuation, which would put them in the top 25 next to Bank of America and Coca Cola.

Additionally, PE firms often rush into public markets when there is a crash to take advantage of low price opportunities. Thoma Bravo famously took dozens of small software companies private in 2022 near the market bottom, removing them from the index pool. Walgreens, which used to hold one of the coveted 30 seats in the Dow Jones Index, fell 90% after taking on too much debt and is now being taken private by Sycamore Partners. Public market holders will lose the opportunity to participate in the Walgreens recovery.

Meta famously went public in 2012 because in 2011 they had exceeded the maximum number of private shareholders. The JOBS act in 2012 raised the threshold requirement to go public from 500 investors to 2,000, excluding employees. Feeder funds, SPVs, are used to circumvent this threshold. A feeder fund might have 1,000 participants but only takes up one shareholder seat at the parent level. ChatGPT estimates that SpaceX has only 1,000 shareholders which is probably accurate enough. I would bet there are well over 100,000 shareholders in SpaceX and OpenAI when you look through all of the holders in each layer of the feeder funds. If you added in LPs and beneficiaries you would easily exceed 1m shareholders. So in other words, the maximum shareholder rule has been effectively rendered useless and the current incentive system keeps it that way.

Most investors today agree to pay a 20% carry to access quality private companies because that is the only way to gain that exposure. This incentive creates a pipeline for private companies to raise capital without ever needing to go to public markets. The marketplace of investors has effectively voted with their wallets saying that If the growth is there then the fees are OK. There are gradients of quality in every asset class, public and private, and there is absolutely no doubt a large swath of overvalued private market assets that would not survive public market pricing.

There are many examples of private market assets that will never be public mostly because they are too small, niche or too short duration in nature. Private credit is a perfect example. Small real estate holdings fall into this bucket as well. I’d go so far as to say that the emergence of private credit following 2008 regulatory changes is one of the greatest contemporary finance evolutions. Again, there will always be gradients of quality in every asset class and private credit isn’t perfect. Overall private credit is well positioned to continue to provide massive benefits for investors for a very long time. The biggest issue with private credit is whether or not the opportunity gets competed away.

There is room for a mix of public and private assets in every portfolio.

Data Centers in Space

Space innovation continues to be an amazing place to find inspiration. The creativity and boundary pushing while people “on earth” squabble about politics and tariffs is a breath of fresh air. The latest development I’ve come across is increasing conversation around a tangible pathway to put datacenters in space. I’ve briefly mentioned this in the past but the concept is getting more attention lately.

There are power, cooling and real estate constraints on earth that may be better handled in space. Direct sunlight can be used for solar power and -250F ‘air’ temperatures can be used for cooling. Iceland has long been popular for datacenters for this same reason, cold air outside and volcanic activity below. Starlink started launching their satellites in 2019 and today they have 7,000 in orbit with plans for between 12,00-34,000. Once we collectively decide that it’s time for datacenters in space then it’s only a matter of a few years before they start to appear. It doesn’t seem that far away at this point.

Whenever we decide to push through hard engineering challenges we end up unlocking new technologies that can be used to improve other industries. Space exploration is critical to American innovation for this reason. Unsurprisingly there is not a lot of representation of this theme in public markets. It’s a niche market.

Tariff Comment

I don’t have all of the hard data handy to back this comment but after researching I’m confident that it’s directionally correct. One of the major missing narratives of the current tariff fights is that the US is the global leader in technology and digital services exports. Software created by a US company that is used abroad is effectively a US “export”. If the official statistics included all digital exports, the US would run a trade surplus and not a trade deficit.

I’ve repeated this for a few weeks but I feel it helps reinforce the core point amidst the noise. The core point of the tariff game is domestic manufacturing independence. In a world where conflict is unavoidable, independent production of critical materials is necessary for robust national security.