Governance Proposal Risk

WHY GOVERNANCE IS AN ECONOMIC RISK. A protocol's rules are code, and governance can change the code. Emission schedules, fee switches, treasury deployment, buybacks and collateral parameters are all votable in many protocols, so a proposal can alter the economics an investment thesis rests on. There is no equity analogue this direct: a shareholder vote rarely rewrites the product's revenue model between one week and the next.

WHAT IT DOES. Reads proposal events from independently verified governor contracts and scans titles and bodies for economic keywords - emission, treasury, supply, fees, buybacks - weighting by recorded vote activity.

WHERE THE METHOD IS WEAK, and this is worth teaching as a general lesson about text classification. A keyword list decides membership by the presence of a STRING, while the question being asked is about MEANING. That fails in both directions at once, which is why adding entries does not fix it:

• FALSE POSITIVES. 'Treasury' appears in routine operational proposals - reimbursing a grant, renewing a service - that change no economics. 'Fees' appears in a proposal about conference fees.
• FALSE NEGATIVES, which are worse because they are invisible. A proposal titled 'Adjust parameter X for pool Y' can be the most economically consequential vote of the year and contains no listed keyword. The economically important proposals are frequently the ones written in the protocol's own internal vocabulary.

The honest test is whether a human would need to READ the proposal to decide. Here they plainly would. This analysis sets its risk tier by string matching; a language model reads the body only to explain each proposal after the tier is set, so the tier stays a keyword screen.

VOTE-ACTIVITY WEIGHTING IS A REASONABLE SECOND DIMENSION. A proposal drawing unusual participation is more likely to matter, because participants incur a cost to vote and do so when they perceive stakes. Note the timing problem though: activity accumulates during the vote, so weighting by it means the signal is strongest as the vote closes, which is late.

WHAT IT CANNOT SEE. Off-chain governance - forum discussion, temperature checks, multisig actions taken without a vote - where a great deal of real decision-making happens before anything reaches a governor contract.