Slashing is the protocol-enforced penalty that destroys part of a validator's staked collateral for misbehavior such as downtime or double-signing. It is what makes proof-of-stake security credible: the validator has real money at risk, creating a genuine economic deterrent.
Treat slashing as a central variable to quantify and reserve against, not a tail event to disclose. Launching staking or restaking without a funded reserve just transfers the risk to depositors.
How it works
When a validator commits a slashable offense, such as proposing two conflicting blocks for the same slot (equivocation) or signing contradictory attestations (surround voting), the protocol destroys a portion of its staked collateral and forces it out of the active set. This is not a fee or a delay. It is permanent capital destruction.
On Ethereum the immediate penalty is at least 1/32 of the validator's balance, then a correlation penalty is layered on top. Isolated slashing carries a small correlation penalty. Hundreds of validators slashed at once can lose nearly their entire remaining balance, by design: systemic misbehavior is punished far harder than isolated mistakes.
Why it matters
Slashing is the primary loss event depositor funds must be protected against. A liquid staking token's net asset value falls when a validator backing it is slashed, and with no reserve that loss passes straight to LST holders. This is not theoretical: several Ethereum validators were partially slashed after the Merge due to configuration errors, especially double-signing from key migration mistakes.
Design consequence
For restaking, exposure multiplies. A position restaked across three actively validated services faces three independent slashing vectors at once. A single correlated incident, like one operator running the same buggy client across all three, can trigger penalties from each. Reserve sizing has to account for the worst-case correlation, not the average single-service loss.
See Tokenomics Audit for how this applies in practice.
Know the terms but not sure how they apply to your project? That is what an engagement is for. We design, document, and stress-test the whole token economy inside the Tokenomics Data Room.
80+ projects advised. Complete tokenomics in 4 to 6 weeks.