If the Bull Market Returns, Which Ethereum L2s Can Turn From "Vampires" Into ETH Value Allies?

TL;DR

• L2 scaling worked, but ETH value capture remains thin.
• The data supports the "vampire L2" concern: in the latest 30D window, tracked networks generated about $10.02 million in fees paid but only about $42.0K in Ethereum rent, roughly 0.42%.
• If the next bull market returns, it is likely to reward fewer L2s, not all L2s.
• Survivors need distribution, liquidity, token value capture, and credible ETH alignment.

Ethereum L2s were supposed to be the scaling victory lap. They made transactions cheaper, moved activity away from congested L1 blockspace, and proved that Ethereum could support a larger app economy without forcing every user to pay mainnet gas. Then the narrative flipped. In the bear market, the same L2s became one of ETH's most uncomfortable criticisms: activity scaled, but ETH fee burn, settlement revenue, and token-level pricing power did not scale with it. That is the "fortune and misfortune" inside Ethereum's L2 strategy: the roadmap succeeded, but the asset-level feedback loop weakened.

CoinEx Research does not view this as evidence that L2s failed. It is evidence that Ethereum's modular roadmap worked, but with an asset-level side effect: L2 success does not automatically become ETH value capture. The real question is no longer whether L2s can scale Ethereum. It is whether that scale still matters enough to ETH.

In the latest Growthepie snapshot tracked, production scaling networks held about $37.91 billion in TVL/TVS on June 9, 2026, processed about 27.01 million daily transactions, and recorded about 2.10 million daily active addresses. Since January 1, 2023, that equals roughly 5.5x TVL/TVS, 7.3x transactions, and 4.6x active addresses. Yet the same dataset showed only about $1.4K in 7-day average daily L1 costs paid by L2s and about $205 in 7-day average daily blob costs. These figures look small because Dencun and subsequent blob-market scaling sharply reduced the cost of publishing L2 data. That is exactly the point: the cheaper Ethereum becomes for L2s, the harder ETH must work to prove value capture through other channels.

That gap is why the next bull market should not be treated as a blanket L2 rescue. The market may still want Ethereum scaling exposure, but CoinEx Research believes it will likely ask a harder question: which L2s can turn from value extractors into ETH value allies?

Why Ethereum L2s Became The Bull Market's Favorite "Vampire" Narrative

The vampire label is emotionally loaded, but the mechanism is simple. L2s benefit from Ethereum security, settlement, liquidity, and brand legitimacy. After EIP-4844 and Dencun, they also gained cheaper data availability through blobs. That was good for users. It was also good for L2 operators, because lower L1 costs improved the economics of running a rollup.

The tension is that Ethereum's own asset economics became less visible. Lower L2 costs mean less pressure on L1 fees. Lower L1 fees mean weaker ETH burn. Weaker burn makes the "ultrasound money" story more conditional. If L2s retain most of the app fees, sequencer economics, and user attention while Ethereum receives only a small rent payment, the market can reasonably question whether ETH is still the cleanest way to capture Ethereum's network growth.

The token layer makes the problem sharper. Many L2 tokens do not directly capture transaction fees. Users often pay fees in ETH or another gas asset, while the L2 token mainly coordinates governance, incentives, or ecosystem grants. That does not make the token worthless, but it changes the burden of proof. If usage rises while token value capture stays vague, the token becomes a narrative proxy rather than a cash-flow-like asset.

The latest Growthepie 30-day revenue split illustrates the issue. In the window from May 11 to June 9, 2026, the top tracked networks in the chart generated about $10.02 million in fees paid but only about $42.0K in rent paid to Ethereum, or roughly 0.42% of measured fees. This does not mean the rest is token-holder revenue, but it does show the scale of the value-capture gap.

Base is the clearest example of distribution-led success with weak direct ETH rent return. It recorded about $5.72 million in network fees paid by users and about $12.1K in rent paid to Ethereum, while Growthepie's profit metric stood near $5.71 million. Arbitrum showed about $611.6K in user fees and about $3.6K in Ethereum rent. Optimism showed about $57.0K in user fees and about $2.8K in Ethereum rent. Exact accounting depends on source methodology, but the direction is clear: for many L2s, the Ethereum rent share remains small compared with the activity and operator economics they support.

L2 Fees Stay Mostly Outside Ethereum Rent

Source: CoinEx Research, Growthepie. Window: May 11 to June 9, 2026. Data downloaded June 10, 2026.

This is not a moral argument against L2s. It is a value-capture argument. Ethereum chose modular scaling because cheap execution could attract more users and applications. The cost is that ETH now needs stronger proof that L2 growth can still translate into burn, staking demand, settlement premium, ETH-denominated liquidity, or institutional monetary demand.

The L2 Shakeout: Why The Next Bull Market Will Not Lift Every Rollup

The old bull-market assumption was simple: if Ethereum comes back, L2 beta comes back. That assumption is weaker today. The L2 market has become more crowded, more capital-intensive, and more distribution-driven. There are too many chains with similar promises: cheaper transactions, faster execution, ecosystem grants, and a token.

The latest activity data already shows head concentration. Base was the largest transaction network among tracked scaling networks, with about 9.92 million daily transactions on June 9, 2026, or about 36.7% of the tracked transaction total. Arbitrum remained the largest by TVL/TVS, with about $16.75 billion, or about 44.2% of the tracked total. The top five networks by transactions accounted for roughly 82.2% of tracked transaction activity, while Base and Arbitrum together held about 74.0% of tracked TVL/TVS.

That split matters. The market is not choosing one universal L2 winner. It is separating roles. Base leads transaction distribution, Arbitrum leads liquidity depth, and Optimism remains important as a stack/network-effect play. This is a healthier structure for the top of the market, but it is harsher for the long tail.

The long tail looks even more fragile when measured from peak TVL. Mode, Scroll, Metis, Taiko, Manta, Loopring, Arbitrum Nova, Unichain, Gravity, and Soneium were all down more than 90% from peak TVL. This list mixes different failure modes. For older or incentive-led chains, the drawdown can signal structural demand decay. For newer chains such as Taiko or Unichain, it may reflect early liquidity resets after launch peaks rather than terminal decline. They remain watchlist names only if their strategic positioning converts into durable usage.

Blast is a useful cautionary example outside the Growthepie export scope. DefiLlama's historical chain TVL data shows Blast peaking near $2.26 billion on June 6, 2024 and falling to about $29.02 million by June 10, 2026, a drawdown of roughly 98.7%. Kinto's closing FAQ provides another warning: even a highly designed modular exchange chain can fail to sustain itself if product-market fit, security, funding, or user demand breaks down.

In other words, several watched L2s have not merely corrected from peak TVL; they now retain only low-single-digit percentages of their former liquidity base.The lesson is that the L2 shakeout has two sides: concentration at the head and collapse in parts of the tail. L2 wars are not decided by technology alone. Distribution, liquidity, trusted apps, capital access, and token design matter. A chain without durable users can become a zombie chain. A token without value capture can become a value trap. A rollup without differentiation can become another low-fee commodity.

Ethereum's Counterattack: From Cheap Blobs To Settlement-Layer Premium

Ethereum is not standing still. Dencun made L2 data cheaper. Fusaka and PeerDAS push the roadmap toward more scalable data availability. Ethereum.org describes Fusaka as part of the path toward higher blob capacity and more efficient data sampling for L2s. Future roadmap work around Glamsterdam and ePBS also points to a chain that wants to scale while improving proposer-builder design and settlement-layer quality.

This matters because Ethereum's counterattack is not simply "make L1 expensive again." Expensive gas is not a business model. The more important goal is to make Ethereum settlement more valuable, more scalable, and more credible. If L2s can post more data, settle more activity, and connect more valuable applications back to Ethereum, then cheap blobs can still become a strategic advantage for ETH.

But the current data does not yet prove that the cure has arrived. In the latest Growthepie settlement-cost series, the 7-day averages were about $1.4K in L1 costs, $205 in blob costs, and $1.6K in aggregate rent paid as of June 9, 2026. Ethereum's roadmap creates the conditions for a stronger settlement-layer premium, but the market still needs proof that more blob capacity can translate into meaningful rent, burn, validator economics, or ETH-denominated demand.

The strongest counterargument is that thin rent may be acceptable. ETH may not need to capture value mainly through fee share if it becomes the neutral settlement asset, liquidity base, and institutional collateral layer for a much larger modular economy. In that view, Ethereum is closer to a settlement network than an app chain, and ETH's premium comes from credible neutrality rather than high take rates. CoinEx Research treats that argument seriously. The problem is that markets still need measurable proof: ETH-denominated liquidity, staking demand, institutional allocation, and settlement premium must grow enough to offset the visible decline in direct fee capture.

The risk for weak L2s is that pure low-cost execution becomes less defensible. If Ethereum makes data availability more scalable and the strongest L2s capture most distribution, then a small chain cannot survive just by saying it is cheap. It needs a reason to exist: a unique app market, a strong distribution channel, a credible ETH-alignment model, or a technical architecture that changes what developers can build.

This is where the misfortune can become fortune. The bear market exposed weak token economics, zombie chains, and value leakage. Ethereum upgrades are making the settlement layer more scalable. Together, they can clean the field. The next cycle may reward fewer L2s, but those survivors could be stronger if they help convert L2 usage into visible ETH value.

From Vampires To Value Allies: The L2 Models That Still Have A Chance

CoinEx Research would treat potential survivors as watch buckets, not a leaderboard.

The first bucket is distribution-led leaders. Base is the clearest example. It has Coinbase distribution, high transaction activity, and strong consumer-app visibility. The important update is that Base is no longer only a "no-token" story. At BaseCamp 2025, Base said it had begun exploring a network token. That would make Base the most important real-world test of this article's value-capture thesis: can a high-activity L2 turn operator economics and distribution into a token model that benefits the broader network rather than repeating the governance-token problem?

The opportunity is obvious. A Base token could turn the largest L2 transaction franchise into direct token exposure and may attract attention from airdrop, listing, governance, and fee-capture narratives. Market estimates have already tried to size that optionality; a JPMorgan estimate reported by The Block put potential value as high as $34 billion. The risk is just as important. Base has not finalized timing, design, or regulatory structure; Coinbase has emphasized the need to do it carefully. A large implied valuation could also create expectations, unlock pressure, or a sell-the-news setup. For ETH, the key question remains whether Base activity creates durable ETH-denominated demand, app liquidity, blob demand, and public-goods alignment rather than only operator economics.

The second bucket is DeFi and liquidity leaders. Arbitrum still has deep TVL, mature DeFi infrastructure, and a large developer ecosystem. That gives it a stronger moat than most long-tail L2s. The open question is ARB value capture. If Arbitrum's DAO, Orbit ecosystem, sequencer economics, or governance design creates a clearer path from network usage to token value, ARB may remain a major L2 attention proxy. Without that, strong chain fundamentals may not fully translate into token performance.

The third bucket is stack and network-effect leaders. Optimism and the Superchain are not only an L2; they are an attempt to make OP Stack chains share standards, revenue logic, and ecosystem coordination. The opportunity is broad distribution through many chains. The risk is value diffusion. If activity spreads across the Superchain without clear OP-level capture, the network may grow while the token still needs stronger proof.

The fourth bucket should be split into three subtypes. ETH-aligned or based-rollup candidates, such as Taiko and potentially Linea if its value-return mechanics are verified, matter because they try to route sequencing or economic value closer to Ethereum itself. Their risk is scale: alignment is strategically valuable, but it must show production usage.

App-specific L2s are a different case. Unichain is relevant because an app-specific chain backed by Uniswap could test whether a major application can bring liquidity to L2 without fully disconnecting from Ethereum. Its risk is that early TVL peaks and liquidity migration can exaggerate both upside and drawdown; the market should separate post-launch volatility from durable app economics.

Distribution, treasury, and performance-led L2s form a third subtype. Mantle and Ink may benefit from exchange, treasury, or institutional distribution, but their ETH-alliance case depends on actual settlement/rent contribution and token utility. MegaETH remains a differentiated-execution candidate, but it should not be treated like a proven production winner until live data shows that high-performance execution creates sticky applications rather than only attention.

L2 Survivor Matrix

Source: CoinEx Research. Review date: June 10, 2026.

None of these categories guarantees token performance. They are screening buckets. The market may pay attention to them in a bull cycle, but each needs confirmation from real usage, token economics, and Ethereum value feedback.

The Six-Factor Checklist For Screening Ethereum L2s In The Next Cycle

The most useful question for L2 token holders is not "which chain has the best story?" It is "which chain has the least fragile mechanism?"

First, check value capture. Does network usage create any real path to the token through fee sharing, burn, buybacks, staking, validator rewards, or treasury accrual? If the answer is no, the token may remain mostly a governance and incentive asset.

Second, check distribution. Does the chain have an exchange, wallet, super-app, consumer product, DeFi venue, or application funnel that can bring users without permanent incentives? In the current L2 market, distribution may matter more than small technical differences.

Third, check network effects. TVL/TVS, active users, stablecoin liquidity, DEX volume, bridges, and developer activity all matter. A chain with liquidity can recover faster. A chain with only incentives can fade quickly.

Fourth, check differentiation. If Ethereum becomes cheaper and faster at the settlement layer, an L2 needs more than "low fees." It needs a category: payments, real-time execution, RWA, gaming, consumer apps, appchain liquidity, privacy, or institutional workflows.

Fifth, check token economics. FDV, circulating supply, unlock schedules, emissions, and market-maker liquidity can overwhelm improving fundamentals. A good chain can still have a weak token setup.

Sixth, check Ethereum alignment. Does the L2 increase ETH-denominated liquidity, blob demand, L1 settlement revenue, validator economics, or Ethereum's settlement premium? Or does it mainly extract users and attention while returning little economic value? This factor matters technically, but it also matters socially. Ethereum-native users may reward L2s that look like allies and punish those that look like pure extractors.

What L2 Token Holders Should Watch Before The Next Rotation

The next L2 rotation, if it comes, should be evaluated through catalysts rather than emotion. The first catalyst is Base's token exploration: any official detail on timing, distribution, governance, fee capture, regulatory structure, or listing path would become a market-wide reference point for L2 token design. The second is governance: ARB, OP, and other L2 tokens need concrete fee-sharing, buyback, staking, or sequencer-revenue proposals, not only broad decentralization language.

The third catalyst is unlock pressure. A token can improve its fundamentals and still underperform if circulating supply expands faster than demand. The fourth is post-Fusaka settlement evidence: blob demand, L1 rent, and ETH-denominated liquidity need to rise enough to show that Ethereum's scaling roadmap is beginning to matter to ETH again. The fifth is app retention: users and fees should remain after incentives fade, especially for app-specific or high-performance L2s.

The "vampire" model worked in the short run because low fees, high incentives, and cheap settlement helped L2s grow quickly. But the same model can become a long-term liability if it leaves ETH holders, L2 token holders, and app users fighting over a thin value pool.

That is the real meaning of fortune and misfortune moving together. The bear market exposed the weak chains. Ethereum upgrades are raising the bar. The next bull market may still bring L2 speculation, but it is unlikely to forgive every L2 equally. The L2s with the best chance are those that can stop being seen as vampires and start proving they are ETH value allies.

Disclaimer: This content is for reference only and does not constitute investment advice. Information may be incomplete or inaccurate. Please do your own research; the author assumes no responsibility for losses.