Restaking Promised Free Money. The Bill Is Coming Due.

The pitch was elegant: you already staked your ETH to secure Ethereum, so why not stake it again to secure other networks and earn extra yield? EigenLayer turned that idea into a $15 billion TVL machine by mid-2024 and spawned an entire ecosystem of liquid restaking tokens — EtherFi's eETH, Renzo's ezETH, Puffer's pufETH, Kelp's rsETH — each promising to let you restake while staying liquid. Two years in, with BTC grinding at $75K and ETH struggling to hold relevance in the Layer-1 conversation, it's worth asking the uncomfortable question: is restaking genuine infrastructure, or is it the DeFi equivalent of taking out a second mortgage to buy lottery tickets?

What Restaking Actually Does (and Doesn't Do)

EigenLayer's core concept — shared security — is real and useful. Instead of every new oracle network, data availability layer, or bridge bootstrapping its own validator set from scratch (expensive, slow, and usually centralized at launch), they can rent Ethereum's existing validator set. Validators opt in, put their staked ETH at additional slashing risk, and get paid for the service.

That part makes sense. Ethereum has ~$55B in staked ETH sitting there. Letting other protocols tap into that security pool is a net positive for the ecosystem. The problem isn't the idea. The problem is what got built on top of it.

The Leverage Layer Cake

Here's where restaking went from infrastructure to financial engineering. The typical user journey in 2024-2025 looked like this:

1. Deposit ETH into Lido → receive stETH (yield layer 1) 2. Deposit stETH into EigenLayer → earn restaking points/rewards (yield layer 2) 3. Receive a liquid restaking token (LRT) from a wrapper like EtherFi → use it as collateral elsewhere (yield layer 3) 4. Deposit that LRT into a lending protocol like Aave → borrow against it (yield layer 4) 5. Loop the borrowed ETH back into step 1

Each layer adds yield. Each layer also adds a claim on the same underlying ETH. One unit of ETH is now supporting four or five layers of obligations. This isn't innovation — it's leverage with extra steps. And leverage works beautifully right up until it doesn't.

The Slashing Problem Nobody Talks About

Here's the part the restaking bulls gloss over: slashing on EigenLayer isn't theoretical. The entire value proposition depends on validators being slashed if they misbehave while securing AVSs (Actively Validated Services). But slashing hasn't been meaningfully tested under stress. As of early 2026, only a handful of AVSs are live, and none have triggered significant slashing events.

Now imagine a scenario where an AVS experiences a bug or an attack, triggering slashing for restaked validators. That slashing doesn't just hit EigenLayer deposits — it cascades. The LRT backed by that restaked ETH suddenly has less collateral. The lending protocol using that LRT as collateral faces bad debt. Liquidations trigger more selling. The same ETH that was "working" across four protocols is now failing across four protocols simultaneously.

This isn't fear-mongering. It's how leverage unwinds in every financial system. The 2022 stETH depeg to 0.93 ETH caused real pain, and that was just one layer of wrapping. Restaking stacks three or four layers deep.

The Yield Reality Check

Strip away the points farming and token incentives, and restaking yields are modest. Genuine AVS fees — the sustainable revenue source — generate low single-digit APY at best. Most of the "yield" restakers earned in 2024-2025 came from EIGEN token emissions and points programs, not from actual demand for shared security.

With EIGEN trading roughly 85% below its airdrop highs, the subsidy model is clearly fading. The protocols that survive will be the ones where AVSs actually pay meaningful fees for security. That's a much smaller market than the $15B in TVL suggested.

If you're evaluating restaking projects, check the actual AVS revenue on Invesaro's screener rather than relying on projected APY figures that bundle token emissions with sustainable fees. The gap between the two tells you how much of the yield is real.

Who Wins, Who Loses

EigenLayer as infrastructure has staying power. Shared security is a genuine unlock for new protocols that need validator sets. EtherFi carved out a real niche by making the UX painless and now functions more like a broad DeFi access layer than a pure restaking play.

But the long tail of LRT protocols — the fifth or sixth liquid restaking wrapper offering marginally different point structures — those are dead products walking. Consolidation is inevitable. And the leverage loopers who stacked four layers of yield on a single ETH? They're sitting on a position that looks great in a spreadsheet until one link in the chain breaks.

The Bottom Line

Restaking is 20% infrastructure breakthrough and 80% financial engineering. The infrastructure part — letting new networks rent Ethereum's security — will persist and matter. The rest is a leverage trade dressed up in governance tokens and point systems. With markets in "very bearish" territory and ETH already under pressure, the collateral chain is only as strong as its weakest link. If you're restaking, know exactly which risks you're stacking. If you're just chasing the highest quoted APY across four nested protocols, you're not farming yield — you're the crop.