The price of
the airbag.
Protection against a liquidation cascade, priced by a market instead of an actuary. Every week settles against a number anyone can recompute.
See how a week settlesIllustrative shape. No market has opened yet. The first epoch publishes its methodology before it accepts a trade.
A worried depositor has exactly one move available: withdraw everything. There is no way to stay in and hedge.
On Kamino, in a single month.
Real positions, not a tail scenario.
October 2025, driven by a 14% SOL drawdown.
These are the problem's numbers, publicly verifiable on chain. PREMIUM has published none of its own yet. When it does, every figure will be checkable through Panta's attribution endpoint.
One market. Two people who need opposite things.
The borrower hedges
Leveraged on Kamino and afraid of next week. Pays a small premium. Nothing happens, that was the cost of sleeping. Something happens, the position pays while the deposit burns.
The underwriter earns
Thinks the week is calm, and sells into that. This is the side that brings volume: selling insurance is hard, selling yield is not. Every underwriter makes the hedge cheaper.
- Price per share
- $0.710
- Shares
- 704.23
- If the week stays calm
- $704.23
- Return
- +40.8%2130% APR
Parametric cover, not insurance. The payout follows a protocol-wide metric, not your individual loss, so it can pay when you were unharmed and pay nothing when you were liquidated.
One market, four states, nobody driving.
- 01
The factory opens the market
One market per protocol per week, created on a cron with a slug derived from the epoch. The settlement window is frozen at this moment and published before anyone can trade it.
- 02
Two sides take opposite views
A borrower buys cover against a cascade. A yield-seeker underwrites a calm week and collects the premium. The curve prices both without needing a counterparty to appear first.
- 03
The window closes and the metric is counted
Liquidated collateral across the whole protocol, summed over the published slot range at finalized commitment. No adjudicator, no dispute over what counts.
- 04
The outcome pays out on chain
The evidence bundle publishes every contributing signature and the total, so anyone can recompute it. Winners claim against the resolved market.
- slug
- kamino-liquidation-e0
- window
- 301,450,000 → 302,890,000
- threshold
- $2,000,000
- state
- open
Illustrative figures for a completed week. No epoch has settled yet.
No adjudicator. Just arithmetic.
The question is a number and two slots, published before the market opens and never adjusted afterwards. A stranger with an RPC endpoint can settle it independently and check our answer.
A market only we can settle is a trust product. A market anyone can settle is infrastructure.
- metric
- Kamino liquidated collateral, USD
- program
- KLend2g3cP87fffoy8q1mQqGKjrxjC8boSyAYavgmjD
- window
- slot 301,450,000 → 302,890,000
- price
- one Pyth snapshot, pinned at epoch start
- threshold
- $2,000,000
- commitment
- finalized
YES if metric ≥ threshold, NO otherwise.
Settled at finalized, never processed. A reorg near the boundary must not be able to change who gets paid.
Protocols priced at launch
Two, deliberately. A long tail of thin markets prices nothing well. These are protocols whose risk we price, not customers and not partners.
Read the other way, it is a public risk price.
The same instrument, seen from outside, is a continuously updating market-implied probability that a given protocol fails this week. Lending protocols can set LTVs against it. Treasuries can set allocation caps. Nobody has been able to see this number before.
GET /api/risk/kaminoSettled epochs
No epoch has settled yet.
Each settled week publishes its evidence bundle here: the slot range, every contributing signature, and the total. Recomputable without asking us for anything.
The first epoch opens with its methodology already public.
Not after. Before. You can check the arithmetic before you decide whether to trust it.