Risks and Guarantees¶

Futarchy markets provide useful signals, but they do not remove risk.

Mechanical guarantees¶

The protocol can mechanically guarantee only what the contracts enforce.

Examples:

Conditional assets are backed by collateral in the Conditional Token Framework.
Splitting creates matching conditional claims for mutually exclusive outcomes.
After resolution, the winning outcome is redeemable according to the reported payout vector.
Merging matched YES and NO positions can recover underlying collateral before resolution.

These guarantees depend on the deployed contracts, token approvals, and the correct oracle path for the market.

What markets do not guarantee¶

Markets do not guarantee:

that a proposal is morally good,
that a proposal will succeed after approval,
that liquidity will be deep,
that prices are manipulation-proof,
that every relevant participant has traded,
that the chosen metric captures everything the community values.

Futarchy produces a price signal. Governance, execution, and interpretation still matter.

Main risk categories¶

Market risk¶

Thin liquidity, high slippage, low participation, or short trading windows can weaken the signal.

Metric risk¶

If the system optimizes the wrong objective, markets can faithfully optimize the wrong thing. This is the usual Goodhart problem.

Oracle risk¶

Current markets rely on an oracle path to decide whether the YES or NO world occurred. Bad question wording, unclear resolution sources, or disputes can delay settlement.

Smart contract risk¶

Contracts can have bugs, integration assumptions can be wrong, and wrappers or AMMs can behave unexpectedly under edge cases.

Governance risk¶

Advisory signals can be ignored. Binding signals can be overused before trust is earned. Projects should increase authority gradually.