Docs / The mechanisms

Decision markets, minus the jargon

Don't vote, trade — explained with a $10 example and no equations.

5 min read · Updated Aug 2026

Every organisation has to answer the same question over and over: should we do this thing? Companies answer it with a boss. DAOs usually answer it with a vote. Backable projects answer it with a market, and the slogan for it is don't vote, trade.

Why voting fails here

Token voting sounds democratic and behaves badly. Voting is free, so almost nobody researches the proposal — their single vote will not change the outcome, so the effort is wasted. The people who do turn out are the ones with something to gain from the specific proposal. And anyone who buys enough tokens can simply outvote everyone who cares. The result is governance decided by turnout campaigns rather than by whether the idea is good.

How a decision market works

When a proposal is created, the market splits into two conditional versions of the token: one that only counts if the proposal passes, and one that only counts if it fails. Traders buy and sell both. After a fixed trading period, the two prices are compared using a , so a last-second spike cannot decide anything.

If the pass-price is higher, the market is saying the project is worth more with this proposal than without it, and the proposal executes. If the fail-price is higher, it does not. Nobody counts votes; the losing branch is simply unwound.

A $10 example

Should the project spend $8,000 on a packaging machine?

PROPOSED

The founder writes the proposal. Two markets open: token-if-pass and token-if-fail.

TRADING

You think the machine is obviously worth it. You put $10 into token-if-pass. Someone who thinks it is a distraction puts $10 into token-if-fail. Everyone with an opinion and a wallet does the same.

MEASURED

Over the trading window, token-if-pass averages $0.062 and token-if-fail averages $0.058. The market's verdict: the project is worth more with the machine.

EXECUTED

The proposal passes and the treasury pays out automatically. The fail-side positions are unwound. You were right, and being right paid; if you had been wrong, being wrong would have cost.

The whole trick in one sentence. Voting asks people what they want and costs them nothing to answer; markets ask people what they believe and charge them for being wrong.

What decision markets can't do

A market is only as good as the traders in it. A tiny project with three participants produces a thin, noisy signal, and markets price what is legible — revenue, dilution, cost — better than things that are not, like whether a hire is a culture fit. They also cannot make a bad business good. What they reliably do is stop the treasury from being drained by proposals that obviously destroy value, without needing anyone trustworthy to be in charge.

Decision markets are run by MetaDAO, which is where the mechanism was built and where the deeper technical documentation lives.

Monthly budgets and proposalsWhich decisions reach a market in the first place.Reading tokenomicsTWAPs also gate the team's unlocks — here is how.