Sensitivity: does the verdict flip?
Same deposit, premium and loss-given-exploit assumptions, swept across a range of exploit probabilities you might believe. Your breakeven row is highlighted — above it, cover wins in expected-value terms; below it, self-insuring wins.
| Assumed exploit probability | Expected loss | Premium cost (fixed) | Net expected value |
|---|
How it works
Cover pools price premiums by assessing the aggregate risk of a protocol — audit history, TVL, code complexity, time in production, and how similar contracts have fared elsewhere — and setting a rate that (in theory) covers expected claims across everyone buying cover on that protocol plus a margin for the pool's capital providers. That rate is the pool's collective bet on exploit probability, expressed as a price. Your side of the trade is different: you're not pricing a whole pool, you're deciding whether to transfer your own risk for a fee, and that decision necessarily runs through your personal belief about how likely an exploit actually is this year — a number nobody can give you with certainty, because past exploit rates are a thin, noisy sample and the next exploit doesn't care what happened last year. The expected-value math this calculator runs (premium cost vs. probability-weighted expected loss) tells you whether cover is a good bet on average, over many repeated years at your stated probability. It does not tell you what to do. A risk-averse person can rationally buy cover even when the expected value is negative — that's the same logic behind why people buy home and car insurance despite insurers being profitable on average, because a single large loss (losing the whole deposit) can matter more than the expected-value math suggests if you can't easily absorb it. This tool models the arithmetic honestly; it is not investment advice and does not know your risk tolerance, your protocol's actual risk, or the future.
Reading the numbers
At the defaults — a $100,000 deposit, a 6% APY premium, an 80% loss-given-exploit assumption, a 5% assumed annual exploit probability, and a full 365-day cover period — the premium cost is $100,000 × 6% × (365/365) = $6,000. The expected loss for the period is $100,000 × 5% × 80% × (365/365) = $4,000. Net expected value is $4,000 − $6,000 = −$2,000: at a 5% assumed exploit probability, cover costs $2,000 more than the expected loss it protects against, so self-insuring has the better expected value here. The breakeven probability — where premium cost exactly equals expected loss — is premiumRate ÷ lossGivenExploit × 100 = 6 ÷ 80 × 100 = 7.5%. Only if you believe the real annual odds of an exploit hitting this specific protocol exceed 7.5% does buying cover flip to positive expected value at this premium and loss assumption. Shorten the cover period and both the premium cost and the expected loss scale down proportionally together, so the breakeven probability itself doesn't move — it only depends on the premium rate and loss-given-exploit ratio, not on how many days you're covering.