The Unraveling of Private Equity
A $13 trillion industry organized around the fact that nobody prices the assets in real time.
Robinhood started selling SpaceX and OpenAI exposure to retail investors in Europe last June: derivative tokens on Arbitrum, backed by shares in an SPV, with no voting rights and no cap table recognition. OpenAI disowned the whole thing and Elon called it fake equity. Thousands of people bought it anyway.
SpaceX has been private since 2002. OpenAI since 2015. Both are worth hundreds of billions. Both sit behind accredited investor requirements, fund minimums, lockup periods, GP relationships. For decades those walls justified themselves: you paid for access because access was scarce.
Those walls have gotten porous. Secondary platforms trade pre-IPO shares daily. Data vendors publish real-time pricing on companies that have never filed a public document. And Robinhood showed that people will take a synthetic with no rights attached over no exposure at all.
Congress restricted certain investments to people who could afford the losses after 1929, and ninety years of infrastructure got built on that restriction: accredited investor rules, limited partnerships, capital call structures, fee models that only make sense when your investors can't leave. "Private markets" became a $13 trillion industry, and the word "private" stopped referring to anything about the assets. It refers to who's allowed to buy them.
You pay for illiquidity (locked in for a decade), opacity (the GP controls what you see), and the quarterly mark (assets priced once a quarter instead of continuously, so the portfolio looks steadier than it is). GPs like this because it flatters their track record and insulates the fee conversation. LPs like it because their risk models stay intact, their boards don't see drawdowns, and the allocation looks like it's doing what it was supposed to do.
The big firms have started dropping minimums and loosening access on their own, letting more investors into the same product. More checks on the same fund means more management fees. The rest is getting solved from the outside: secondary markets creating liquidity, data vendors publishing pricing, derivative products offering exposure without ownership.
State Street launched $GLD in November 2004. Before that, gold meant dealer spreads, vault storage, insurance on bars sitting in a warehouse you'd never visit. Ten years later, gold investment had tripled and dealer margins were functionally zero. The custodians argued that ETF holders didn't "really" own gold, that paper claims weren't the same as allocated bars. They were probably right, which didn't matter. Exposure had gotten cheaper, and cheaper won.
Some of this is real. When Apollo buys a retailer's defaulted debt at 40 cents and spends two years on the steering committee negotiating the restructuring, the recovery rate changes because Apollo is in the room. A passive holder of the same bonds gets whatever Apollo negotiates for them. You can't synthetic your way into a steering committee seat. Sequoia's name on a Series A cap table helps the company raise its B. A GP who rezones a dead mall into housing and sells at 3x built that return with operating work. Those end up being real skill premiums.
That's maybe a quarter of private markets by AUM, a third if you're generous. Distressed credit, early-stage venture, value-add real estate, operational turnarounds. The other $9 or $10 trillion sits in large-cap buyout, core real estate, direct lending, late-stage growth, strategies where the GP is mostly buying an index with extra steps. Those look more like gold before $GLD.
The other frictions are dissolving on schedule. Secondary platforms handle liquidity. Data vendors handle transparency. Feeder funds and interval structures handle minimums. But commercial disruption needs a frustrated buyer or a frustrated seller, and the quarterly mark has neither.
Every tool eating away at private market frictions also makes staying private more viable. Companies that can offer employee liquidity through secondaries and point to third-party valuations don't need to IPO. So the pool keeps growing while fees compress (management fees are already down to 1.6%, a fifth below where they sat for decades). What's left is a $13 trillion industry whose most important feature is that nobody prices the assets in real time.