Restaking Goes Institutional: Who’s Actually Deploying Capital

EIGEN, SSV and BABY all trade more than 90 percent below their all-time highs, and the points-farming frenzy that drove restaking’s first wave has mostly faded. Yet in the same eight months that those prices kept sliding, a Nasdaq-listed Ethereum treasury company staked $170 million through EigenCloud, a publicly traded ETF market maker became an operator on an EigenLayer credit market, and one of Asia’s largest institutional custodians started routing client Bitcoin directly into Babylon. This is not a story about points or airdrops. It is a story about who actually shows up once the speculative froth clears, and what that says about restaking’s second act.

A Different Restaking Story

Restaking has been one of the most written-about corners of crypto since 2023, and most of that coverage has followed the same arc: explosive total value locked growth in 2024, a brutal drawdown through 2025 and into 2026 as EigenLayer’s own total value locked fell from a peak near $19.7 billion to a range that has bounced between roughly $4.7 billion and the low teens of billions depending on the tracker and the week, and a parallel collapse in token prices as EIGEN, SSV and BABY all gave back more than 90 percent of their all-time highs. Layered on top of that is a real security incident: the April 2026 Kelp DAO exploit that left Aave holding close to $292 million in bad debt.

What gets less attention is a second, quieter trend running underneath all of that. A small number of regulated, publicly accountable institutions have been building real positions in restaking infrastructure since late 2025, not through speculative token purchases but through direct protocol integrations, custody arrangements and governance proposals. SharpLink Gaming, Flow Traders and Hex Trust are not crypto-native funds chasing points; they are a Nasdaq company, a Euronext-listed market maker and a regulated Asian custodian, respectively. Their presence is a different kind of signal than a total value locked chart, and it is worth examining on its own terms.

Restaking in Plain Terms

Ethereum’s proof-of-stake consensus already asks validators to lock up ETH and follow protocol rules in exchange for staking rewards, currently in the area of 2.7 to 2.8 percent annually plus a modest MEV component. Restaking, a term EigenLayer coined, lets a validator or a holder of a liquid staking token opt that same capital back in to secure additional services on top of Ethereum itself. Those services are usually called Actively Validated Services, or AVSs: oracle networks, cross-chain bridges, data availability layers and rollup sequencers all need some form of economic security, and rather than bootstrap a new token and validator set from scratch, they can rent Ethereum’s existing stake and slashing infrastructure instead.

The trade-off is straightforward to state and harder to price. A restaker earns extra yield on top of ordinary staking rewards, but also accepts new slashing conditions specific to each AVS they secure. Get it wrong, a bad software update, a compromised operator, a bug in an AVS’s own logic, and the penalty is not just missed rewards but a haircut on the underlying stake. Bitcoin’s version of restaking, built by Babylon, sidesteps part of this by using Bitcoin’s own timelock scripting rather than a smart contract, which is covered in more depth further down.

None of this mechanical detail is new; it has been the subject of dozens of explainers over the past year. What has changed recently is who is reading past the mechanics and actually allocating capital.

EigenLayer and EigenCloud: Still the Center of Gravity

EigenLayer, which rebranded to EigenCloud as it pushed further into verifiable compute and data availability products, remains the largest restaking platform by total value locked and the only one with a widely traded token. As of 22 July 2026, EIGEN trades around $0.234, giving it a market capitalization of roughly $174 million against a circulating supply of about 741 million tokens; that is down approximately 96 percent from its all-time high of $5.65, set in December 2024, according to CoinGecko. Total value locked figures are genuinely contested across trackers this year, ranging from roughly $4.7 billion to double-digit billions depending on the source and the week, but every reading agrees on the direction: down sharply from the 2024 peak, even as the protocol itself has kept building.

Part of that building has been a rework of how EIGEN emissions actually get distributed. The Eigen Foundation’s Programmatic Incentives v2, whose first payouts went out in October 2025, roughly doubled total annual emissions from 4 percent to 8 percent of supply, but the more telling change is where that extra emission goes: the share allocated to EIGEN stakers and operators directly quadrupled from 1 percent to 4 percent, while a new 1 percent slice now funds business development and ecosystem growth, according to the proposal posted to the EigenLayer forum. The stated intent is to reward capital that is actually securing live, fee-generating AVSs rather than capital sitting idle waiting for a future airdrop, a direct response to criticism that the 2024 points era rewarded speculation more than usage.

EigenCloud’s other bet is that restaked ETH can secure more than blockchains. EigenCompute and EigenVerify extend the same cryptoeconomic security model to general off-chain computation, which the project has pitched partly as infrastructure for verifying autonomous AI agents, a theme HOGE Wire has explored in more depth in its look at crypto’s autonomous agent economy. Whether that bet pays off is a separate question from restaking’s more immediate institutional story, but it is the backdrop against which EigenCloud is trying to reposition itself as more than a yield product, and it matters for how seriously institutional risk teams end up treating the platform over the next few years.

Liquid Restaking Tokens and the Kelp-Aave Warning

Most restaked ETH does not sit in a wallet earning a fixed yield; it gets wrapped into a liquid restaking token, an LRT, that can then be reused elsewhere in DeFi as collateral. ether.fi’s eETH is the largest LRT by a wide margin, with Renzo’s ezETH and Puffer’s pufETH among the next tier. The pitch is capital efficiency: the same ETH secures AVSs and backs a loan or a liquidity position at the same time.

That reuse is also where the sector’s worst 2026 incident happened. On 19 April, attackers exploited a cross-chain bridge used by Kelp DAO to mint roughly 116,500 rsETH out of thin air rather than draining existing deposits, then used the inflated supply as collateral on Aave to borrow against, leaving Aave with bad debt CoinDesk put at close to $292 million. Aave’s total value locked fell by roughly $6 billion over that single weekend. The recovery effort that followed is itself a data point about how tightly the restaking and liquid staking ecosystems are now wired together: Lido, ether.fi, Consensys and Aave founder Stani Kulechov personally pledged funds toward covering the shortfall, and rsETH’s backing was fully restored by around mid-2026 after Kelp migrated its bridge from LayerZero’s OFT standard to Chainlink’s CCIP with additional attestor requirements.

The episode is a useful gut check for anyone framing 2026 as restaking’s institutional coming-of-age: the underlying technology still has sharp edges, and the same bridges and wrapped-token mechanics that make LRTs useful are exactly what got exploited. It is also part of why, as detailed below, institutional entrants have so far leaned toward direct, custodied exposure over holding LRTs picked up on the open market; the attack surface is different, but it does not disappear. Readers tracking how attackers increasingly target the softer parts of crypto infrastructure, wallets and bridges, rather than protocols themselves, may recognize the pattern from HOGE Wire’s coverage of 2026’s drainer economy.

Babylon: Restaking Comes to Bitcoin

Babylon applies the restaking idea to an asset EigenLayer cannot touch: Bitcoin. Rather than wrapping BTC into a synthetic token or bridging it to another chain, Babylon locks native Bitcoin on the Bitcoin network itself using Taproot-based timelock scripts, then lets that locked BTC secure proof-of-stake networks that opt in as what Babylon calls Bitcoin Supercharged Networks. No custodian holds the coins in the sense a wrapped-token bridge would; the security model rests on Bitcoin’s own scripting rather than a third party’s multisig.

Babylon’s mainnet and its BABY token both launched in April 2025, and the protocol has consistently shown more than 56,000 BTC staked (its own dashboard has read 56,853.16 BTC, worth roughly $5.6 billion, across several checks between mid-June and July 2026), which most trackers describe as the largest Bitcoin staking system by that measure. BABY itself has been a rougher ride than the staking figures suggest: the token trades around $0.0128 as of 22 July 2026, down about 92 percent from its April 2025 all-time high of $0.1661, per CoinGecko, a valuation gap against the multi-billion-dollar Bitcoin position it secures that is worth noting on its own.

Institutional access to Babylon has moved faster than its token price. Hex Trust, a regulated Asian digital asset custodian, began supporting Babylon staking directly from custody in 2025, letting clients earn Babylon rewards on Bitcoin held in Hex Trust’s insured, air-gapped Hex Safe environment rather than moving coins to a separate staking wallet. “Supporting Babylon and the growing Bitcoin ecosystem represents an enormous opportunity for Hex Trust and our clients,” said Calvin Shen, Hex Trust’s Chief Commercial Officer, adding that “unlocking the value of Bitcoin through staking, and ultimately through DeFi, is a huge milestone for the industry,” according to Hex Trust’s own announcement. Babylon’s Genesis Phase 2 upgrade later extended that model, letting institutions stake directly from custody without a separate bridging step.

The more forward-looking institutional move is a proposed integration with Aave. In December 2025, Babylon Labs and Aave Labs announced plans to let native BTC locked in Babylon’s Trustless Bitcoin Vaults serve as collateral on Aave V4, using zero-knowledge proofs to mirror the locked Bitcoin as a transfer-restricted token on Ethereum without ever wrapping or bridging the underlying coin. “Trustless Bitcoin vaults allow native BTC to participate directly in DeFi while preserving its fundamental security guarantees,” said Babylon co-founder David Tse, while Aave Labs founder and CEO Stani Kulechov called it a demonstration of “how easily new markets can be launched using Aave V4’s Hub and Spoke model,” per Babylon Labs’ announcement. The plan, formalized in a Temp Check posted to Aave’s governance forum, targeted testing in the first quarter of 2026 and a launch around April. As of this writing it has not yet appeared live on Aave V4, so the more accurate read is that it remains in progress rather than shipped, months past its original target. Readers curious how Bitcoin’s own price cycle factors into decisions like this can find more in HOGE Wire’s ongoing look at halving-cycle math.

Symbiotic’s Pivot: From Restaking to Collateral Markets

Symbiotic launched in January 2025 as EigenLayer’s most credible direct competitor: it accepts any ERC-20 token as collateral rather than just ETH derivatives, and it built slashing into the protocol from day one instead of adding it later. More than two years after its points program began, and with roughly $34.8 million raised across a Paradigm and cyber.Fund-led seed round and a Pantera-led Series A, Symbiotic still has no public token, an unusual position for a protocol carrying somewhere in the range of $1.6 to $1.7 billion in total value locked.

Rather than close that gap, Symbiotic used a 1 July 2026 upgrade called Core V2 to redefine what it is building toward. The framing shifted from restaking specifically to shared collateral markets more broadly: the same pooled capital that once secured AVSs can now back onchain insurance, credit and tokenized real-world asset products, with Symbiotic claiming roughly 70 percent better capital efficiency than isolated liquidity pools. The first Core V2 product, Liquid Lane, targets instant settlement for real-world assets that normally sit behind 60 to 180 day redemption windows, with asset managers Midas and Fasanara Capital among its early participants, according to The Block’s coverage of the launch. Nexus Mutual founder Hugh Karp welcomed the shift for his own use case: “Shared collateral is especially important for onchain cover, where demand for protection is outgrowing what any single balance sheet should carry alone,” he said, adding that Symbiotic lets “delegated capital sit behind Nexus Mutual as reinsurance capacity, with first-loss and second-loss exposure defined separately.”

It is a pragmatic pivot, and arguably a warning sign for restaking as a standalone category: the protocol best positioned to be EigenLayer’s peer competitor has effectively concluded that a broader collateral market, not AVS security specifically, is where the durable business is.

Karak’s Exit: From AVS Security to OpenGDP

Karak took that same logic further. The protocol, which raised a $48 million Series A led by Lightspeed Venture Partners in December 2023 at a valuation above $1 billion, spent 2024 and 2025 building itself as a third major restaking platform alongside EigenLayer and Symbiotic. By mid-2026 it had rebranded entirely, dropping the Karak name for OpenGDP and repositioning around real-world asset tokenization, stablecoin settlement and institutional coordination rather than AVS security at all. OpenGDP’s own site describes its goal as unifying the controls, liquidity, privacy, audit and autonomous systems around tokenized economic activity, language that has little to do with restaking, slashing or shared security.

No specific rebrand date or founder quote accompanied the change in any source found for this piece, and no live, liquidly traded OpenGDP or KAR token has surfaced on a major exchange. What is clear from OpenGDP’s own site is that Karak’s legacy V1 and V2 staking products and its K2 bridge are now referenced only as legacy infrastructure, not as the core product. Combined with Symbiotic’s Core V2 pivot, Karak’s exit means that of the three protocols that spent 2024 pitching themselves as EigenLayer alternatives, only Babylon, which plays in a different asset class entirely, is still built primarily around restaking as the product rather than as a stepping stone to something broader.

The Institutional Deployments Actually Happening

Put together, five concrete institutional moves since late 2025 give a clearer picture of where real capital has actually gone, as opposed to where retail attention has gone.

None of these are large relative to total crypto market capitalization, and several, the Aave and Babylon vault chief among them, are still proposals rather than live products. But the pattern is consistent: publicly accountable entities choosing restaking-adjacent infrastructure specifically for the yield stacking it enables, and doing so through custodians and delegate structures rather than by running AVS infrastructure themselves.

Why Institutions Are Showing Up Now

Three things changed between restaking’s first wave in 2023 and 2024 and this quieter second one. First, plain ETH staking yield on its own, at roughly 2.7 to 2.8 percent, is a modest number for a treasury company under pressure to show a return on a large ETH position; layering EigenCloud restaking rewards, Linea ecosystem incentives and ether.fi yield on top, as SharpLink has done, is a more defensible pitch even before counting any token-price upside, with early estimates putting the combined stack in the 7 to 9 percent range.

Second, the tokenized real-world asset market that Symbiotic and others are now chasing has grown large enough to matter on its own, crossing roughly $320 billion by mid-2026 even before most of that value can be redeemed on demand, which is precisely the liquidity gap that products like Liquid Lane are built to close. Restaking infrastructure, originally built to secure oracles and bridges, turns out to be reusable as the collateral engine for that entirely separate growth story, and that reusability is arguably a bigger deal for the sector’s long-term relevance than any single AVS launch.

Third, and least glamorous, Aave V4’s hub-and-spoke architecture, live on Ethereum since March 2026 and expanded to Avalanche in July, has made it mechanically simpler to spin up a new, isolated lending market around a specific collateral type such as vaultBTC, without redesigning the whole protocol each time. Infrastructure maturity is a less exciting story than a price chart, but it is arguably the more important one for whether integrations like Babylon’s proposed BTC vault actually ship on schedule going forward.

Custody Is the Real Unlock

Ask why institutional restaking adoption took until late 2025 to show up in any visible way, and the answer has less to do with yield or even regulation than with custody. A publicly traded company cannot self-custody a nine-figure ETH position in a browser wallet, and until qualified custodians built direct integrations with EigenCloud and similar protocols, there was no compliant path from a treasury balance sheet into restaking at all. SharpLink’s Linea deployment routes through Anchorage Digital Bank as qualified custodian specifically because that path exists now; Hex Trust’s Babylon integration is built the same way, letting Bitcoin earn Babylon rewards without ever leaving Hex Trust’s insured Hex Safe environment.

This is the same underlying tension HOGE Wire examined in its look at multisig wallet security and blind signing: institutional-grade custody is not just about who holds the keys, it is about whether the signing and delegation flow is auditable and insurable end to end. A custodian that can stake, restake and report on a client’s Bitcoin or ETH without that client ever exposing a raw private key to a staking application is solving a genuinely different problem than a retail user connecting a hardware wallet to a restaking app directly, and it is that problem, more than restaking’s underlying yield, that has been the real bottleneck for institutional capital.

The Regulatory Gray Zone: What the SEC Has and Has Not Said

The SEC has, in a narrow sense, gotten more comfortable with staking generally. Its Division of Corporation Finance issued a staff statement on 29 May 2025 concluding that certain protocol staking activities, covering solo staking, self-custodial staking through a third party, and custodial staking, do not by themselves involve the offer and sale of securities. A follow-up statement on 5 August 2025 extended similar reasoning to certain liquid staking arrangements and the staking receipt tokens they mint. Both statements were widely read as a green light for the plain-vanilla end of the staking market, per the SEC’s own release.

Both statements also explicitly carve restaking out. Liquid staking, restaking and other variations are excluded from the August 2025 guidance by name, and staff statements in any case reflect staff views only, not a Commission rule, regulation or exemptive order, meaning they carry no legal force and remain fact-specific. That leaves restaking-as-a-service arrangements, AVS operator agreements and the yield-stacking products institutions are now using in a position no different, legally, from where they stood before either statement: unresolved. For a broader map of how fragmented the current US approach is across agencies, see HOGE Wire’s overview of US crypto enforcement in 2026.

The CLARITY Act, the market-structure bill that would give the CFTC and SEC clearer, statutorily defined jurisdiction over digital assets including staking-adjacent products, has not resolved this either. The Senate Banking Committee advanced it 15 to 9 in May, but as of late July it remains stuck behind three disputes, over enforceable ethics language for officials’ crypto holdings, a law-enforcement carve-out and a stablecoin-yield loophole, with prediction markets pricing 2026 passage odds somewhere in the mid-40s to mid-50s percent range and roughly three working weeks left before the Senate’s August recess. Institutions moving into restaking today are doing so, by definition, ahead of any statutory clarity on the activity itself.

What Institutional Capital Is Actually Underwriting

None of the institutions covered here are running AVS infrastructure themselves or taking on uninsured smart contract risk directly; they are, in effect, underwriting the assumption that EigenCloud, Babylon and their peers have engineered slashing risk narrowly enough to be worth the yield. EigenCloud’s own architecture is built around that exact assumption: its Operator Sets and Unique Stake Allocation system, introduced in 2024, is designed so that a given unit of restaked ETH can only be slashed by a single operator set at any time, isolating a failure in one AVS from cascading into every other service the same capital secures, according to EigenCloud’s own description of the model.

Whether that isolation holds up under real stress, rather than in documentation, is precisely the open question institutional capital is now testing. The Kelp DAO incident showed that the bridge and LRT layer around restaking can still fail in ways that have nothing to do with slashing at all; a bad AVS or a compromised operator set is a different, still largely untested failure mode at institutional scale. Every institutional deployment covered in this piece has so far leaned toward the more conservative end of that spectrum, holding restaked positions through regulated custodians rather than operating AVS infrastructure directly, which limits their exposure but does not eliminate it. It is a deliberately cautious first move, not a wholesale endorsement of restaking’s risk model.

Restaking by the Numbers

The token side of restaking still tells a very different story from the institutional-adoption side. Figures below are via CoinGecko as of 22 July 2026.

Underlying value secured tells a different story again. Babylon’s roughly $5.6 billion in staked Bitcoin and EigenCloud’s total value locked, contested as it is across trackers but still sitting somewhere in the multi-billion-dollar range, dwarf the combined market capitalization of EIGEN, SSV and BABY several times over. That gap, a large amount of capital secured against a comparatively small amount of token value, is exactly the setup that makes the sector’s emissions and incentive design, EigenLayer’s Programmatic Incentives v2 chief among them, matter as much as any single institutional deal: token holders are, in effect, being asked to capture value from an asset base many times larger than what the token itself currently represents.

What Comes Next

A handful of concrete events will say more about restaking’s institutional trajectory over the next two quarters than another total value locked chart will. Whether the Babylon-Aave BTC vault actually goes live, months past its original April target, is the clearest near-term signal of whether proposed institutional DeFi integrations in this sector ship on the timelines their announcements promise. Whether Symbiotic ever launches a public token, or decides its Core V2 collateral-markets business works better without one, will settle a two-year-old open question about whether restaking-adjacent infrastructure needs a native token at all.

On the regulatory side, the Senate’s August recess is a hard deadline of sorts for the CLARITY Act’s 2026 prospects, and a lapse would leave restaking exactly where it sits today: judged safe enough by a handful of institutions to build around, but still outside the perimeter of the SEC’s own staking guidance. Watch, too, for whether EigenCloud’s push into verifiable compute for AI agents produces anything with real usage attached, since that, more than another AVS launch, is the bet the protocol is actually making with its rebrand.

Frequently Asked Questions

What is restaking in crypto?

Restaking lets someone who has already staked ETH, or who holds a liquid staking token, redeploy that same stake to help secure additional blockchain services, called Actively Validated Services or AVSs, such as oracles, bridges, data availability layers and rollup sequencers. In exchange, the staker earns extra yield on top of ordinary Ethereum staking rewards, but also accepts new slashing conditions tied to each service. EigenLayer, now EigenCloud, pioneered the model on Ethereum; Babylon later built an equivalent for Bitcoin using native timelock scripts instead of a smart contract.

Is restaking safe for institutional investors?

It depends heavily on which layer of the stack an institution touches. Direct AVS operation carries smart contract and slashing risk that protocols like EigenLayer try to isolate through mechanisms such as Operator Sets and Unique Stake Allocation. Indirect exposure through a liquid restaking token carries bridge and depeg risk, as the April 2026 Kelp DAO exploit that produced roughly $292 million in bad debt on Aave demonstrated. Institutions entering so far, such as SharpLink and Flow Traders, have generally done so through regulated custodians and audited delegate structures rather than running infrastructure themselves.

What is the difference between EigenLayer and Symbiotic?

EigenLayer, now EigenCloud, is Ethereum-focused, has a live EIGEN token, and is the largest restaking platform by total value locked. Symbiotic accepts any ERC-20 as collateral, launched with slashing built in from day one, and still has no public token more than two years after its points program began. As of July 2026 it has pivoted its Core V2 product away from pure restaking toward broader shared collateral markets for credit, insurance and tokenized real-world assets.

Can Bitcoin be restaked?

Yes, through Babylon, which lets Bitcoin holders lock BTC directly on the Bitcoin network using its own timelock scripting, without wrapping the asset or bridging it to another chain. The locked BTC then secures proof-of-stake networks that opt in as Bitcoin Supercharged Networks. Babylon has consistently shown more than 56,000 BTC staked, worth several billion dollars, and Babylon Labs has proposed a further integration that would let that same locked BTC also serve as collateral on Aave.

Does the SEC regulate restaking?

Not directly, and that is part of what keeps institutional adoption cautious. The SEC’s Division of Corporation Finance issued staff statements in May and August 2025 concluding that certain protocol staking and liquid staking activities do not, by themselves, involve the offer and sale of securities, but both statements explicitly excluded restaking from that treatment. That leaves restaking-as-a-service arrangements and AVS operator relationships without clear federal guidance, and staff statements carry no legal force in any case.

Written by the HOGE Wire markets desk.