Decision markets aim to extend prediction markets from forecasting outcomes to informing decisions. The core idea stems from Robin Hanson’s proposal for futarchy & policy markets. The mechanism: if a company, government, or institution agrees on a metric of success, prediction markets can determine which policy or action is most likely to improve that metric.

It’s an elegant idea. If prediction markets are good at aggregating distributed information about future events, they should, in theory, be good at aggregating wisdom about which choices lead to better outcomes.

The past several years have brought a wave of experiments trying to put this into practice. Today, MetaDAO runs futarchy-style governance for startups, using token price as the performance indicator (proposals only pass when markets predict they’ll improve it). Combinator, a decision market infrastructure protocol, operates similarly, letting anyone run these markets for themselves.

After spending significant time in this space, I believe the current implementations of decision markets fail to deliver outside of a few narrow market types, and it comes down to two structural problems: a lack of informed traders and an overly-complicated architecture.

The trader problem

For a market to generate a useful price signal, you need informed traders. To attract traders, you need to stake enough capital. The Fifty Cent Dollars piece on Minimum Viable Liquidity puts a name to this threshold: the MVL is the floor below which a market doesn’t attract enough informed trading activity to price accurately.

Classical prediction markets like Polymarket and Kalshi clear this threshold for major events because the audience is enormous. Millions of people have opinions about who will win a presidential election or whether the Fed will raise rates.

Decision markets, however, tend to only be relevant to the people who are affected by, or informed about, the specific decision. If a company is choosing between two product strategies, the pool of people with meaningful information is small: employees, close advisors, and a handful of competitors who understand the space. Attracting the few traders who can actually price it accurately requires seeding enough capital to make it worth their while. The thinner the market, the higher that bar.

Furthermore, for the traders to accurately price the market, they need context: the com...

handsdiff

handsdiff