Every finance major agrees, at some point, that it would be cool to be able to buy and sell a share of someone's future earnings. Is your friend considering quitting their job in order to pursue their dream of playing in a band or writing a novel, but unable to afford to? Why not invest enough money for them to get started and get a piece of the upside if it works? There could be a whole market here, where people sell off a slice of their future income when they start a medical residency in order to have a decent standard of living at the start in exchange for a slightly-less-fantastic one once they're actually practicing. Maybe schools would be more incentivized to educate their students well if those students paid tuition partly in basis points.

And, for that matter, why can't we hedge against a rise in the price of our favorite meal, a drop in real estate prices in a neighborhood where we bought a home, or an increase in rents in a neighborhood where we didn't.

All of these markets would be useful, but most of them don't exist as markets, and for a few good reasons.

First, markets work best when the products in question are fungible, because that allows you to do real-world hedging (e.g. you sell wheat futures and deliver actual bushels of wheat to close out your position), and, increasingly, because you can create derivatives against them and hedge with the underlying asset.¹ That's one reason financial centers are often colocated, not just with ports, but with entrepôts, i.e. places where one of the local businesses is accepting bulk imports and then parceling them out to specific destinations. That business creates a local market in grading and comparing different variants on the same product, since they aren't perfectly substitutable but can be similar enough to be compared to one another.

Markets also need intermediaries who are willing to provide liquidity. You need a willing buyer and a willing seller to make a trade, but if the economic buyer and seller don't want to transact at exactly the same moment, the trade won't happen. If there's an intermediary who can temporarily warehouse the product in question, it's an easier product to trade, but that means that markets are more likely to come into existence when there's a critical mass of buyers and sellers who are more time- than price-sensitive, and who can provide a good living to price-sensitive liquidity providers that are willing to hold risk on their books for a while. This, too, is an entrepôt-friendly criterion, because it means there's already a population of market participants who, by default, have goods in their inventory and risk they may want to take off their books.

The last issue, which is related to the one above, is that adverse selection is a killer in new markets. If an active market doesn't exist, and someone offers to trade with you, the one thing you know with certainty is that they have reason to believe that being on their side of the transaction is a good idea. There are plenty of forces that can reduce this—someone with natural demand to hedge has an obvious explanation for why they'd want to do so (then again, the farmer who might be selling wheat futures just to hedge is also someone who's pretty well-informed about how the wheat season is going). If someone offers to sell you 1% of their lifetime earnings at a favorable price, it might be because they're diversifying their income or trying to better align incentives with a consortium of powerful backers, but it also might be because they've decided to choose a lower-income career path or gotten some bad medical news. Adverse selection must exist in markets for them to be worth participating in, because the only alternative is for everyone to have identical information at which point none of them have an incentive to trade. But too much of it means that every trade is a market for lemons situation, where the optimal answer to every trade offer is "Of course not! You're just ripping me off."

From that perspective, the thing to explain is: why do markets exist? The answer varies by asset classes:

• In commodities, if the product is truly fungible, or even if the level of fungibility can be an input into a model, you can have a lively market because everyone's optimal position over time is flat, but producers have an accidental long position (when the oil or copper or corn comes out of the ground) and consumers have a short position (hunger, or the need for energy, constitute a short position in nutritious food and conveniently accessible energy).

• Stocks and bonds are a way to create artificial fungibility: every share of Procter & Gamble is identical to every other share because that's how they're defined, and since no company's cash needs are precisely identical to their cash flow generation, there's room for mutually-beneficial trades.³ Bonds have an even better solution to adverse selection: if you hold to maturity, you're going to make what you were promised as long as you were right about the company's creditworthiness, so you always have the option to do whatever amount of fundamental research it takes to have high conviction even if your counterparty may know something you don't.

• In this sense, derivatives are like bonds, in that they generally derive their value from some underlying asset that can be traded to hedge them. This doesn't apply to every derivative—weather bets are their own thing, for example. On the other hand, the more weather derivatives there are, the easier it is to isolate other bets (if you're long soy futures because of geopolitics, you don't want your day to be ruined because of rain, so you can take a bigger position if you can hedge the rain-risk out). And that means that over time, since those derivatives end up implicitly referencing or correlating to agricultural commodities, natural gas, and sometimes reinsurance equities or catastrophe bonds, you still get opportunities for indirect hedging.

This piece started with a few examples of betting on people's earnings in specific high-variance fields, and those examples were chosen because there have been economic models built around this, though they aren't really markets. Music labels used to basically function as VCs, paying lots of bands more than they were worth to do early recordings in order to occasionally find the next Madonna or Metallica.² Publishing works the same way: advances mostly lose money, but occasionally they mean that the publisher gets a big win and a potential lock on a valuable franchise. Pre-seed investing in pre-product, pre-idea businesses is also an implicit bet on the human capital of the founder and the future earnings stream it will create—and even if the idea doesn’t work out, there’s a sort of Gentleman’s Pro-Rata where well-behaved investors who back a team’s first failed startup get a chance to back the next ones. Colleges try to do this indirectly, too: when they spend a lot on student amenities, what they're partly doing is trying to buy goodwill that can be converted into donations from their most successful alumni, though this is a weaker effect.

What all of these have in common is that they create an artificial market where one side has a big information advantage, and where the other side has not-purely-financial motivations to do a deal: "We're recording our first album" and "I got a book deal" are both worth bragging about even if they don't go anywhere.⁴ In that market, everything is non-fungible, so there are high returns to careful evaluation—it's as if every company got to invent their own financial metrics and most of the job of an investor was to figure out how to back into some number approximating revenue, gross profit, opex, etc. And these markets all end up being somewhat unpopular, because any time there's a series of messy deals with a wide bid-ask spread and a skewed distribution of outcomes, every deal will be regretted by at least one side of the transaction. Markets are a unique and special thing, they mostly can't come into existence unless circumstances are perfect, and they're valuable things to have.

The Diff has covered markets, and when they do and don't work, in many contexts.

• We've argued that air conditioning created more concentrated trading venues ($).

• Sometimes markets break down because they're segmented in a way where buyers and sellers can't trade at the same prices ($), as when homeowners are underwater on their purchase but massively in-the-money on their mortgage.

• Prediction markets also suffer from a big adverse selection problem.

• Sometimes, odd markets come out of nowhere. Remember NFTs ($)? (The general NFT concept is now back to being wildly underrated, but there are challenging behavioral problems to solve before they get big.)

• Commodities trading firms occupy an important niche ($).

Join the discussion!

1  Some commodities markets price specific products against some reference commodity; WTI is a specific kind of oil, but you can promise to deliver some other form of oil at WTI-minus-10 or whatever benchmark you like.

2  That got harder in part because the people who get the worst of this bargain are mega-celebrity musicians, who have a much more rabid fanbase than the typical record executive. So the record company gets an option on a big success, but that option has an implied knockout where the more nonlinear it is, the more likely the star is to want to recut the deal.

3  But this also requires a huge accretion of legal standards and legal precedents, and the fact that the Delaware Court of Chancery has yet to announce that they've finished figuring out corporate law and will be sending everybody home indicates that this is a continuous process rather than a one-time thing.

4  This, in fact, is part of the adverse selection risk labels and publishers ran—which, ironically, meant that the greedy, shareholder-maximizing approach was to back genuine artists who cared about making something beautiful over people who wanted social approval and would be less motivated once they got it.