Why Event Collateral Is Different

Event positions violate the assumptions that overcollateralized lending relies on. Four properties matter most.

Four properties that break lending

• Jump repricing. Information does not arrive smoothly. A position can move from one level to a far lower one in a single block, faster than any liquidation can react after the fact.
• Liquidity vanishes with news. Depth thins exactly when a lender would need to sell, because the same event that moves the price also clears the book.
• Bounded terminal payoff. The claim resolves to a fixed value. At resolution, collateral that looked sound can settle to nothing, with no recovery path.
• Adverse selection. A borrower who wants leverage against a specific position is, by construction, more likely to know something about how it resolves than the lender does.

Three design consequences

These properties force the design, rather than decorate it.

First, collateral is marked at executable value: the conservative mark, the most pessimistic of several prices, never the last trade. Value the system cannot defend on sale does not back credit. The mechanics are on the risk engine page.

Second, borrowing capacity is removed before resolution, not at it. Capacity decays as resolution approaches and the origination gate stops new credit entirely near the event, so positions are not financed through the window where they can gap to zero.

Third, liquidation runs through specialist auctions rather than generic keepers. Bonded specialists who understand the collateral clear lots that ordinary bots would misprice or abandon. See liquidations.