EigenLayer Restaking Decline: Is It Dead?

ether.fi built weETH, the largest liquid restaking token in the sector, by routing user ETH into EigenLayer to earn extra yield on top of ordinary staking rewards. As of this month, that's largely over. The company has split weETH back into a plain liquid-staking product and spun the restaking piece into a separate, much smaller token (weETHs) built on Symbiotic, EigenLayer's main rival, currently sitting at only about $18 million, roughly 0.5% of ether.fi's total staking base. Full technical severance, removing the EigenPod withdrawal credentials that link ether.fi to EigenLayer, is scheduled for the fourth quarter.

This matters because ether.fi wasn't a marginal player deciding restaking wasn't for it. It was the single largest channel of capital into EigenLayer. When the biggest distributor unwinds this completely, that's the market's most informed participant concluding the yield no longer compensates for the risk, not one fund manager's opinion, but a verdict from the entity that had the best view of the economics.

The mechanism is straightforward. EigenLayer's slashing went live in April 2025, meaning restakers could now actually lose principal if the software they were securing, the actively validated services, or AVSs, misbehaved or got exploited. Before slashing activated, restaking was close to a free option: extra yield, limited downside. After it went live, restakers were taking on real tail risk for AVS fees that, in practice, stayed thin. That's the core repricing: risk went up, reward didn't follow, and the trade that looked free stopped looking worth it.

Trust took a second hit in April 2026, when Kelp DAO, another major LRT, suffered roughly a $300 million exploit that triggered roughly $5.4 billion in ETH withdrawals from Aave alone, as panicked users pulled funds over fears of bad debt from Kelp's stolen collateral, part of a broader $13 billion pulled from DeFi in 48 hours. That's not evidence EigenLayer itself was hacked, but it's exactly the kind of event that makes a rational allocator ask why they're holding extra smart-contract and slashing risk for AVS yields that were never that large to begin with. ether.fi's exit lands on top of that reset, not in isolation.

Zoom out and the TVL trajectory tells the same story: EigenLayer peaked around $22 billion in August 2024 and has fallen toward roughly $5-7 billion since. Crucially, that capital hasn't rotated into Symbiotic, the most obvious rival restaker, Symbiotic's own TVL sits around $1.6-1.7 billion, far too small to have absorbed what left EigenLayer. This is money leaving the restaking category outright, not shopping for a better version of the same trade.

For weETH holders, the practical effect is that their token is quietly reverting to something closer to plain liquid-staked ETH, lower yield, lower risk, and no more restaking upside. That's arguably a rational trade for most holders given the last 18 months, and the market appears to agree: ether.fi's own token, ETHFI, is up roughly 22% over the past week, a rally the company itself attributes to a neobank push into tokenized stocks and lending rather than restaking. The market is rewarding the exit, not punishing it.

EIGEN holders are in a tougher spot. The token has stayed range-bound around $0.19-$0.22 through most of August, a small-cap shadow of its 2024 launch, and its price has tracked the TVL collapse closely. If the largest distributor is walking away and no rival restaker is absorbing the outflow, EIGEN's near-term value depends less on restaked-ETH volume, the metric it was built to capture, and more on whether EigenLayer's own pivot pays off.

That pivot is real and already underway. EigenLayer rebranded as EigenCloud in June 2025 and has been building toward verifiable-compute and AI-adjacent infrastructure, EigenDA for data availability, EigenCompute for off-chain verified execution, EigenVerify for proof systems, with partnerships reportedly including Google and LayerZero. EigenCompute's mainnet is targeted for the third quarter of 2026. The logic is that AVS and compute fee revenue can eventually stand on its own, independent of how much ETH is restaked. That the protocol itself is chasing this path is a tacit admission that restaking-for-yield underperformed as a standalone business model.

The base case here is continued shrinkage: restaked-ETH TVL keeps falling through the rest of 2026 as ether.fi completes its exit, and other large LRTs, Renzo, Puffer, Kelp, face the same incentive to unbundle restaking from their core tokens rather than defend a shrinking, low-conviction category. EIGEN likely stays weak as long as its price is implicitly tied to restaking volume that keeps declining.

That case breaks bullish if EigenCompute's mainnet launch produces real, paying AVS or compute usage this quarter, evidence the protocol can generate fee revenue that has nothing to do with restaked ETH, letting EIGEN re-rate on a different business entirely. It breaks further bearish if another major LRT follows ether.fi's playbook before that revenue materializes, compounding the outflow with no offsetting narrative to hold EIGEN's price up. Watching which of those two events lands first is the real signal here, not whether restaking is dead, but whether EigenLayer can replace it with something that pays.