“We Price the Tail”: The Founder of a DeFi Cover Protocol on Selling Insurance Nobody Wants Until They Do
Kaisu Orrenmark runs Lanternback, an onchain cover market that has paid out on 11 exploits in two years. She talks about underwriting smart-contract risk, why cover remains a niche product, and what insurers get wrong about code.
DeFi users spend heavily on yield and almost nothing on protection. Onchain cover protocols estimate that less than 2% of the value locked in lending and liquidity protocols is insured. Kaisu Orrenmark, 38, a former catastrophe-risk modeller at a European reinsurer, founded Lanternback in 2024 to try to change that. The protocol lets stakers underwrite cover on specific smart contracts and pays claims through a vote of staked assessors. It currently holds about $212 million in underwriting capital. We spoke over video from her home in Tallinn.
The Problem With Pricing Code
Crypto Beacon: Start with the obvious question. How do you price the risk that a smart contract gets hacked?
Kaisu Orrenmark: Badly, at first. In traditional catastrophe modelling you have centuries of hurricanes. In DeFi you have a few hundred notable exploits and the code keeps changing. So we price the tail, not the average. We assume a protocol with a clean audit history, a bug bounty and three years live still has an annual loss probability in the low single digits. That gives you a premium of perhaps 2% to 3% of covered value per year. For a young protocol it can run above 10%.
CB: Who sets those numbers?
KO: The market does, within guardrails. Underwriters stake capital against a particular protocol and choose the price at which they will take the risk. If the price is too low, they lose money when a claim lands. If too high, nobody buys. We seed the starting rate with a model, but the model is a hint.
“Everyone loves insurance until they read the premium. Then they decide the risk is somebody else’s.”
Claims Without a Claims Department
CB: Your protocol has paid 11 claims. How do you decide what counts?
KO: The policy wording is narrow on purpose. We cover a technical failure of the named contract: a coding bug, an exploit, an oracle manipulation if the contract’s own logic permitted it. We do not cover a stablecoin depegging, a rug pull by insiders or a user falling for a phishing site. Each claim goes to a panel of about 40 assessors who have staked tokens, and they have seven days to vote with evidence. If they approve dishonestly they lose stake.
CB: Does that work when there is a $40 million loss and people are angry?
KO: It was tested hard in March, when a lending market we covered was drained. The claim ran to $9.3 million. There was a lot of noise on social media. Assessors reviewed the transaction traces and the governance forum, and approved about 91% of the claimed amount. The rest was excluded because those users had borrowed through a wrapper contract that we had not covered. People hate that kind of exclusion, and they are right to hate it, but it is the only way the product survives.

Why Demand Stays Low
CB: Two percent coverage. Why is it so small?
KO: Three reasons. First, the yield on offer in DeFi is a seductive number, and insurance eats into it. A vault paying 8% with 3% going to cover is a vault paying 5%, and the leaderboards do not reward that. Second, a lot of risk sits with the biggest depositors, who are institutions, and institutions want a regulated insurer with a balance sheet and a rating. Third, we still have the trust problem: cover sold by a smart contract is itself smart-contract risk.
CB: That last one seems awkward for you.
KO: It is the central irony. We do not cover ourselves. Our own contracts have been audited four times, and we run a separate pool of $6 million that users can buy cover from, but I would not call that elegant.
What Traditional Insurers Miss
CB: Several large insurers have begun offering crypto cover. Are they competitors?
KO: They are partly customers. A traditional insurer will write a policy for a custodian, but it will not underwrite a lending protocol’s code. It will sometimes lay off the risk to us. I have taken calls from three of them this year. What they struggle with is correlation. In DeFi, one oracle failure can hit a dozen protocols at the same time, so ten separate policies are really one policy. We model that explicitly, and cap how much capital can sit behind a cluster of linked protocols.
What Comes Next
CB: What would make the market grow?
KO: Two things. Make cover automatic: a protocol buys it on behalf of all its depositors and builds the cost into the fee, so users never face a checkbox. And get better data on losses. At the moment each exploit is analysed in a blog post. I want a shared, structured database so everyone can price from the same facts.
CB: Last question. Would you put your own savings in an uninsured vault?
KO: I have, and I regret one of them. That is probably why I do this job.
Sana tracks crypto regulation across the EU and UK, drawing on years of reporting on financial compliance. She is based in Dublin.
Conversation (…)
One briefing. Every morning. Free.
Join readers who start the day with Crypto Beacon: the essential Crypto & web3 stories, curated by our editors.