Layer 2 Consolidation: Why the “More Chains Is Better” Era Is Ending
par LCX Team ·
For most of 2022 through 2024, the Ethereum scaling conversation had a simple rhythm: a new rollup would launch, run a points program, farm an airdrop cycle, and briefly show up near the top of the TVL leaderboards. Repeat. The implicit belief driving this cycle was that more chains meant more scaling, more experimentation, and more room for everyone to win. Ethereum leaned into this with its “rollup-centric roadmap,” and dozens of teams answered the call using the same handful of stacks, OP Stack, Arbitrum Orbit, ZK Stack, Polygon CDK.
By 2026, that belief has run into the data. The ecosystem didn’t get more evenly distributed as more chains launched, it got more concentrated. This piece looks at why fragmentation lost, who actually absorbed the liquidity and mindshare, and what the shift means for teams deciding where to build.
The numbers tell a power-law story, not a “rising tide” story
The rollup landscape now includes over 70 active networks securing tens of billions of dollars in value. But that value is not spread evenly, it never was, and the gap has widened. Depending on which measurement lens you use, two chains now account for the overwhelming majority of activity:
- On liquidity-secured metrics, Base and Arbitrum together represent roughly 80% of all value secured across the entire L2 ecosystem, a concentration no L2 pairing had reached before.
- On transaction volume, Arbitrum, Base, and Optimism together process around 90% of all L2 transactions, leaving dozens of other networks splitting the remainder.
- Base alone grew from about $3.1B in TVL in January 2025 to a peak above $5.6B by October, capturing close to half of all L2 DeFi TVL in the process, largely on the back of Coinbase’s distribution rather than any unique technical breakthrough.
Meanwhile, the long tail didn’t just grow slowly in many cases it evaporated. Blast is the clearest cautionary tale: its TVL collapsed from roughly $2.2B in mid-2024 to about $55M by the end of 2025, following a disappointing token launch and a broader exodus of both users and liquidity back toward Base and Arbitrum. Across the mid-tier of general-purpose rollups holding a few hundred million to a billion dollars in TVL, a similar pattern played out in early 2026: once liquidity-mining rewards expired, outflows followed almost immediately, because emissions had been the only reason capital was there in the first place.
Why fragmentation lost
A few structural forces explain why “just launch a chain” stopped being a viable growth strategy on its own.
- Liquidity is the actual product, and liquidity hates being spread thin. DeFi is a network-effects business: deeper liquidity means tighter spreads, better execution, and more composability between protocols. A brand-new chain starts every relationship from zero, new bridges, new oracle feeds, new liquidity pools that have to be bootstrapped from nothing. Incentive programs can fake this bootstrapping for a while, but once emissions stop, only organic usage keeps liquidity in place. Most new chains had none.
- Technical differentiation stopped being the deciding factor. In the early rollup era, chains competed on throughput, fees, and proof systems. By 2026, the base technology, OP Stack, Arbitrum’s Nitro, ZK Stack has converged enough that most general-purpose EVM rollups feel functionally similar to a builder deciding where to deploy. When the underlying tech is roughly interchangeable, distribution and existing user relationships become the deciding factor instead and that favors the chains that already have both.
- Distribution beat novelty. Base’s growth is the clearest illustration: it didn’t win by out-innovating other OP Stack chains technically, it won because Coinbase could funnel its existing user base directly onto the chain, with fiat access and regulatory grounding that consumer apps found attractive. Arbitrum’s staying power similarly rests less on any single technical edge and more on years of accumulated DeFi composability, deep liquidity across lending, perps, and DEXs that’s hard to replicate from a cold start.
- Fragmentation has a real UX cost that users increasingly notice. Holding ETH on Base but wanting to buy an NFT on Optimism, or needing to bridge before you can act, is friction that a single-chain experience never had. As users compared today’s fragmented landscape against the near-frictionless experience of a single chain, the tolerance for “just bridge it” dropped. That friction disproportionately punishes smaller, less-connected chains, because bridging in and out of a low-liquidity network is where users feel the pain most.
- Enterprises chose infrastructure, not experimentation. A wave of institutional and consumer platforms building their own L2s in 2025 chose to build on existing, proven stacks rather than commission bespoke chains, reinforcing exactly the networks that already had scale, rather than creating new independent competitors.
Who is actually absorbing liquidity and mindshare
Base is the standout of this cycle. Backed by Coinbase’s distribution, it leads on both raw activity (reportedly tens of millions of daily transactions and hundreds of thousands of daily active users in early 2026) and DeFi value secured, making it the default home for consumer-facing and payments-adjacent applications.
Arbitrum remains the deepest DeFi hub of the rollup world. Its combination of mature tooling, a long-established developer community, and a genuinely composable financial stack (perpetuals, lending, spot DEXs all interoperating on one chain) gives it retention that incentive programs alone can’t buy, a large share of its daily active users are repeat visitors, not one-off airdrop farmers.
Optimism, largely through the OP Stack and the “Superchain” concept, has positioned itself less as a single competing chain and more as the shared backbone that other chains, including Base are built on top of. That’s a different kind of win: rather than accumulating TVL directly, Optimism accumulates influence over how a large share of the ecosystem is built and (eventually) how it interoperates.
Everyone else is fighting for a shrinking remainder. ZK-rollups such as zkSync Era and Starknet, along with Linea, Ink, Katana, and a handful of others, are competing for institutional and specialized use cases, often justified less by “we’re a top-3 chain” and more by a specific value proposition (faster settlement finality, cryptographic guarantees, or a narrow vertical) that the big three don’t fully serve.
It’s not just consolidation, it’s an attempt to make the fragmentation invisible
The more interesting long-term story isn’t which chain “wins,” but that Ethereum’s own protocol roadmap has shifted toward treating fragmentation itself as the problem to solve, not the L2 count. Two efforts define this direction:
- Interoperability standards, like intent-based frameworks that let a user specify an outcome (“turn this asset into USDC on another chain”) and let solvers handle the routing, bridging, and execution behind the scenes, aiming to make cross-chain actions feel like a single transaction instead of a multi-step chore.
- Shared sequencing and native interop layers within ecosystems like the Superchain, which aim to let liquidity move between member chains without the overhead of traditional bridging, and eventually support atomic actions that span multiple chains in one flow.
If these efforts succeed, the practical difference between “one chain” and “a tightly integrated cluster of chains” narrows substantially, which raises a real question about whether today’s winners (Base, Arbitrum, the Superchain) end up as this cycle’s consolidation story, or whether tighter interop just changes what “winning” means without reducing the chain count at all. It’s also not without risk: a deeply integrated Superchain could become efficient internally while acting as a walled garden relative to non-member chains, effectively trading old-style fragmentation for a new, narrower kind.
What this means for builders choosing where to deploy
The practical takeaway isn’t “only deploy on Base or Arbitrum” it’s that the calculus for choosing a chain has changed shape.
- Distribution and existing liquidity now matter more than marginal technical advantages. If your application depends on deep liquidity (lending, perps, complex DeFi), deploying somewhere with genuine composability, largely Arbitrum today, outweighs a slightly cheaper or faster alternative with a thin ecosystem.
- Consumer and payments-focused products should weigh built-in distribution heavily. Base’s edge is largely about who’s already there and how easily they can onboard, not raw throughput.
- “New chain, big incentive program” is a much weaker signal than it used to be. TVL and user counts driven primarily by emissions tend to evaporate once the rewards stop; builders evaluating a chain (or considering launching an app-specific rollup of their own) should ask what happens to activity after incentives end, not just what the leaderboard looks like today.
- Application-specific rollups still have a real niche but the bar for justifying one is higher. A narrow, well-defined use case (a specific game, a specialized financial product, an enterprise deployment with particular compliance needs) can still make sense on its own chain. A general-purpose rollup competing head-on with Base or Arbitrum for the same undifferentiated DeFi and consumer traffic increasingly does not.
- Watch the interoperability roadmaps, not just today’s TVL charts. If shared sequencing and intent-based interop mature as planned, deploying within a well-connected cluster (like the Superchain) may offer much of the liquidity access of a top chain without requiring you to be the top chain yourself, a meaningfully different bet than picking the single biggest chain today.
The bottom line
“More chains” was never inherently better, it was a byproduct of how easy the tooling made it to launch one, combined with a funding environment that rewarded new token launches over sustainable usage. Once incentive-driven growth ran its course, the ecosystem reverted to something closer to how networks actually behave: liquidity and users concentrate where the network effects are strongest, and everyone else has to find a genuinely differentiated reason to exist. For builders, the lesson isn’t to chase the winners reflexively, it’s to be honest about whether your use case benefits from deep existing liquidity and distribution, or whether it has a narrow enough niche to justify swimming against the consolidation tide.
