Multi-AVS Restaking Slashing-Adjusted Net Yield Calculator
EigenLayer and Symbiotic let you opt the same restaked ETH into several AVS at once, and each one you add stacks its own reward APY on top of its own slashing probability and severity — against your whole restaked stake, not a carved-off share of it. This tool sums the expected drag across up to three AVS, shows each one's break-even slashing probability, and — if they share an operator — the correlated tail-risk scenario that independent-probability math hides.
✓ No signup✓ Net APY (expected)✓ Per-AVS break-even✓ Correlated tail risk✓ Same-operator toggle
AVS 1
AVS 2
AVS 3
Net APY (expected case)
—
Per-AVS breakdown
Reward, expected annual drag and the break-even probability for each AVS on its own. A highlighted row means that AVS's own slashing probability input is already above the rate at which its reward stops covering its risk.
AVS
Reward APY
Expected drag
Break-even probability
Status
Correlated tail risk: one operator, one bad day
The independent-sum math above treats every AVS as its own coin flip. If one operator runs the infrastructure for all of them, a single fault can trip all three at once — this compares the spread-out expected drag against that one-event scenario.
Why stacking AVS changes the risk, not just the yield
Every AVS you opt into pays its own reward APY, and it's tempting to just add them up and assume the risk scales the same gentle way. It doesn't. Each AVS's slashing condition is written against your full restaked balance, so three AVS means three full-severity ways to lose stake on the same capital, not one severity split three ways. This calculator sums the expected drag across your chosen AVS and nets it against the combined reward, then separately checks what happens if those AVS all sit behind one operator — the case where a single infrastructure fault can trigger more than one slashing event in the same incident. Compare the position sizing question this answers against liquid restaking risk for a single LST/LRT, and against restaking points if you're farming the same deposit for an airdrop on top of the yield.
The math
Each AVS's expected annual drag is its slashing probability times its slashing severity: dragI = (probI/100) × (severityI/100) × 100. Summing the reward APYs and subtracting the summed drags gives the expected net APY — the number you'd converge to on average across many independent years, since each AVS's fault is treated as its own separate event.
Each AVS's break-even probability is rewardI ÷ severityI × 100 — the annual slashing probability at which that AVS's own reward exactly cancels its own expected loss. An AVS is net-negative on its own once its real-world slashing probability climbs above this line, regardless of how the other AVS in your stack are doing.
When AVS share one operator, the tail scenario assumes a single fault can trigger all of them together: tailSeverity is the sum of every AVS's severity (capped at 100% of stake), and tailProbability is approximated as the highest single AVS's own annual fault rate — the "some fault happens somewhere in the stack" rate. tailExpectedLoss = tailProbability × tailSeverity nets against the same reward total to give the correlated net APY, which is usually close to the independent-sum result on average but represents a far larger single-event loss if it actually happens.
Why can slashing on one AVS affect my whole restaked position, not just that AVS's share?
EigenLayer restaking doesn't split your ETH into separate labeled buckets per AVS the way a portfolio allocation chart might imply. You opt the same restaked balance into each AVS you validate for, and each AVS's slashing contract is written against that whole balance, not a pre-carved fraction of it. If AVS 2 has a fault and its slashing condition triggers, the penalty is calculated as a percentage of your total restaked stake — the severity number in this calculator — regardless of how many other AVS you're also running or how much reward each one pays you. Running three AVS instead of one doesn't dilute any single AVS's claim on your stake; it just gives you three separate, full-severity ways to get slashed on the same underlying capital. That's the mechanic this calculator is built around: your total expected drag is the sum of each AVS's own probability-times-severity, not a shared severity split three ways.
How do I read the break-even probability number, and why can a high-severity AVS still be worth running?
Break-even probability is the annual slashing probability at which an AVS's own reward APY exactly cancels its own expected loss — above it, that AVS is a net-negative bet in expectation; below it, the reward more than compensates for the risk you're carrying. It's rewardI ÷ severityI × 100, so a high severity number pushes the break-even threshold down, but a high reward pushes it back up, and the two can offset. In the default example, AVS3 has the worst severity of the three (8%) but also the highest reward (3.0%), which puts its break-even at 37.5% — meaning that AVS would need to slash more than a third of the time every year before it stopped being worth running, a rate no live AVS has come close to. Severity alone tells you how bad a bad day is; break-even probability tells you how often a bad day would need to happen before the reward stops covering it, which is usually the more useful number for deciding whether to opt in.
If the independent-sum math doesn't change, why does running all my AVS through one operator matter?
It matters because the independent-sum math assumes each AVS's slashing event is its own separate coin flip, and that assumption quietly breaks down when one operator runs the validator infrastructure for all of them. A single operator fault — a bad client update, a mismanaged signing key, a misconfigured node — doesn't respect AVS boundaries; it can trip the slashing conditions on every AVS that operator services at once, in the same incident. The expected-value total (0.44% a year in the default numbers) describes an average drip spread across many independent bad-luck draws over time. The tail scenario this calculator shows when you check the operator box describes something different: one event where you could lose the combined severity across every AVS on that operator, capped at 100% of your stake. Both numbers are true at the same time — the average outcome is small, and the worst realistic single outcome is much larger — and operator concentration is what turns the second number from a footnote into something worth checking before you delegate three AVS to one address.
How is this different from the restaking points calculator and the liquid restaking risk calculator?
The restaking points calculator is an upside tool — it projects what a points-farming deposit is worth once it converts into an airdrop, and has nothing to do with slashing. The liquid restaking risk calculator looks at a single LST or LRT position and the slashing plus smart-contract risk baked into that one token. This calculator sits a level below both: it assumes you've already decided to restake and asks what happens once you start opting the same collateral into multiple AVS at the same time, which is the actual EigenLayer/Symbiotic mechanic that increases risk as you chase more reward APY. Stacking three AVS instead of one doesn't split your risk three ways, it stacks three full-severity exposures on the same stake, and this tool is built specifically to show that compounding effect and the operator-concentration tail risk that goes with it — neither of which the points or single-LST tools are designed to capture.