The nice thing about markets is that in some literal sense, everyone's on a level playing field: if you put in a limit order to buy 100 shares of something, and an order from Citadel to buy the same thing hits the order book a microsecond later or a penny per share lower, no matter how sophisticated the infrastructure and analysis behind their trade was, you're ahead of them in line and get to trade before them. Then again, the downside is that if you're participating in markets, there is no one to protect you from the fact that you're competing with Citadel, and many other very competent participants besides. To an extent, you can free-ride on all of their efforts by just buying index funds—market-makers are competing to provide liquidity to that kind of flow, and hedge funds and long-onlies are competing to make the prices for what you're buying more efficient.

There are less liquid and more exclusive markets that are adjacent to these liquid ones. For example, suppose your view on some company is: it's a good short eventually, there's no catalyst right now, but once there is a catalyst, it'll be so expensive to borrow that it probably won't be worth shorting. For an individual investor, that's just a frustrating fact of life; sometimes you pay 5% a year to short it when there's no reason for it to go down, and you'd have to pay 50% or more to maintain that position once it's actually dropping. But an institutional investor, or a sufficiently wealthy individual, has another option: they can get a bank to write them a total return swap that's equivalent to shorting the stock, then buy an offsetting position. There are two ways to look at what they've done in this case:

1. They've paid a fee to lock in the option to short the stock on prearranged terms; when they sell the stock they bought to offset that position, their economic exposure flips to being a short position, but they don't actually have to arrange any borrowing. (That's their counterparty's problem, at least if the counterparty is fully hedged.)

2. They've created a synthetic asset: a bet on how much it costs to borrow a given stock over a year. They can monetize this by shorting the stock (some anomalies that look like promising ways to spot good companies to short completely go away once you account for the cost to borrow). If the position is big enough, they could even monetize it by negotiating a borrow fee split with their prime broker.

There are plenty of other equities-adjacent businesses where not everyone has access to the same suite of products. And one of the more interesting is venture, where getting into deals is a big part of the entire business. It's also a very tough business, because there's a correlation between a fund being prominent enough to hear about an interesting round first and that fund quickly deciding whether or not it wants to lead the round or even offer to do the entire round itself. So deal flow is not just a measure of the quantity of deals a given investor gets access to, but a proxy for their quality, too. This dynamic exists in places other than venture—Berkshire Hathaway under Buffett had unique access to M&A opportunities with big family-owned companies and non-controlling stakes in distressed financial firms.¹

But all is not lost! What this actually means is that there isn't a single consistent market for access to startups. Instead, there's a graph, and as long as a given investor is a prominent node in some important part of that graph, they have the opportunity to produce alpha—there are investors who are good at staying connected to particular alumni networks, there are investors who are better than their peers at early technical diligence in their particular field, and there are the ones whose franchise is that they respond to emails incredibly fast in a way that's generally helpful.

In none of these cases will you get anything resembling a meaningful sample size before the market has moved on. By the time there were clear leaders in the field of backing pre-IPO PayPal employees, all of them had moved on to other things, like Youtube, SpaceX and Palantir, and presumably by the time it's obvious who is best at investing in ex-SpaceX or ex-Palantir companies, most of the alumni who were going to start big companies will already have done so.

One of the most interesting situations is the kind that occurs across asset classes, where a previously untradable asset suddenly turns into one with a lively, active market. This can happen for all sorts of reasons: in options, it was partly because of the publication of the Black-Scholes model and partly from the introduction of hand-held calculators; in pre-IPO stock, IPO timelines got longer right about when it became trivial to check LinkedIn for early employees of now-valuable companies; in high-yield bonds, Michael Milken pretty much willed the market into existence (though 60s-era conglomerates' habit of issuing low-rated debt, and frequent defaults elsewhere during the disruptive 1970s, certainly gave him some good raw material to work with). There's a lot of money in bringing markets into existence, for exchanges and for market-makers. But it's quite hard to do, because most of the time, a good market doesn't exist for a reason. Startups don't really need there to be a Series-A Stock Exchange, because their financial backers already have a mandate to hold for a long time, and because they want their employees to have equity in order to stay incentivized.

There's a point at which markets reach escape velocity, and an asset shifts from being something that's primarily bought for the long term to something you can put a stop-loss on. That shift usually requires a slow increase in transparency or an improvement in valuation norms, followed by a gradual shift to ad hoc over-the-counter transactions, eventually culminating in a formal market with an order book or at least a cohort of specialized liquidity providers who are willing to make a market. But for some markets, especially ones where making a market means revealing too much alpha, it's basically impossible for everything to align in a way that creates a durable market. For the truly bespoke derivatives bets, and for the earliest stages of venture, illiquidity will remain the default.

In The Diff, we spend a lot of time looking at where markets do and don't emerge, including:

• How Milken made high-yield debt a real market.

• Data companies follow a few basic templates ($).

• The many paradoxes of market efficiency.

• Many of your biggest assets aren't on your balance sheet.

• Prediction markets are a new market with an old problem.

• Stakes in venture-backed companies are also somewhat hard to trade ($).

Join the discussion!

1  It's notable that both Paul Graham and Warren Buffett, despite investing in basically the opposite kinds of companies, got a lot of exposure from writing good essays, and that both of them did a great job executing on a business model of: "We will give you a termsheet, very fast, with financial terms that are not necessarily as good as you'd get elsewhere, but having our brand name associated with you will make it much more likely that you raise more money on better terms in the future." It takes a long time to build up a franchise like this, and neither of them seemed to target that as part of their business. One of the first cases in which Buffett did this was when GEICO was recapitalizing through a big preferred stock issue in 1976, and in that case Buffett already had a quarter-century of familiarity with the company. When Paul Graham was writing about the first cohort of what was then called the Summer Founders Program, he referred to it as "an experiment."