Last week, a founder pitched me his new Layer 2 for NFT derivatives. He had a deck, a testnet, and a valuation of $200 million. I asked him one question: 'How many unique wallets are you expecting in month one?' He paused. 'We hope to onboard the Arbitrum users.'
That's the problem. There are only so many Arbitrum users. And Optimism users. And Base users. And Scroll users. And every other L2 that launched this year. The number of Ethereum L2s has crossed 50. Yet, the average daily active users across all of them barely matches what Arbitrum alone saw in October 2023.
This isn't scaling. This is atomization. We are slicing the same small user base into finer and finer fragments, and calling it progress.
I've been in this industry long enough to recognize a recurring pattern. In 2017, it was ICOs—hundreds of tokens chasing the same pool of retail capital. In 2021, it was L1s—Solana, Avalanche, Terra—each promising to be the next Ethereum. Now, in 2024, the narrative has shifted to L2s. But the arithmetic hasn't changed: when you increase supply without increasing demand, prices—and attention—dilute.
History doesn't repeat, but it often rhymes. This is the same verse, just remixed for a bull market.
Let's look at the data. According to L2Beat, as of January 2025, the total value locked across all L2s is roughly $38 billion. Sounds impressive. But strip out the double-counting from canonical bridges and multi-chain tokens, and the real TVL closer to $25 billion. Compare that to Ethereum L1's $45 billion. The L2s collectively hold barely half of the base layer.
More telling is user activity. Dune Analytics shows that the top five L2s—Arbitrum, Optimism, Base, Blast, and zkSync—capture 92% of all L2 transactions. The remaining 45 L2s fight over the crumbs. And those top five are themselves losing share to new entrants like Taiko and Linea, which are siphoning users through incentive programs.
The problem is structural. Each L2 creates its own isolated liquidity pool, its own bridge, its own user onboarding flow. Moving assets between L2s is not frictionless—it's a cross-chain puzzle.
During the DeFi summer of 2020, Uniswap proved that concentrated liquidity in a single venue creates network effects. The more traders, the better pricing. The better pricing, the more traders. L2s break that loop. They fragment the order book, increase slippage, and force users to make 10-hop journeys just to access the best yield.
Alpha isn't extracted; it's dissipated across fragmented cash registers.
I remember a conversation from late 2023 with a quant friend at a market-making firm. He told me they had to maintain inventory on eight different L2s to capture basic arbitrage. The operational overhead ate 40% of their margins. They reduced to three L2s by mid-2024. Now they're back to just Ethereum L1 and Arbitrum. The others weren't worth the ledger entries.
The illusion of value in digital scarcity—that's what this L2 proliferation reminds me of. Each new rollup claims to be 'the home of the next billion users.' But those users aren't coming. Not because crypto is dead, but because the onboarding experience is now more fragmented than ever.
Let me illustrate with a mental model. Imagine the early internet circa 1999. Hundreds of ISPs, each with its own proprietary browser, email client, and content portal. Users had to choose AOL, CompuServe, or Prodigy—and then could only access content within that walled garden. The market didn't scale. It was a mess. Then consolidation happened: HTTP, TCP/IP, and the browser made everything interoperable. The walled gardens collapsed.
Crypto is currently in the 'ISP era' of L2s. Each rollup is a walled garden with its own bridge, its own token, its own governance. Interoperability solutions like LayerZero and Chainlink CCIP are the equivalent of early dial-up modems—they work, but they add latency and cost. The user experience is a far cry from the unified, single-chain world that retail investors dream of.
And here's the contrarian twist: the market is pricing this fragmentation as a feature, but it's a bug. Venture capitalists are pouring billions into L2 infrastructure because they believe that 'many L2s' is the natural end state. They point to the Linux kernel spawning hundreds of distributions. But Linux distros don't compete for network effects or liquidity. They compete for user preference in functionality. In crypto, liquidity is the oxygen. When you split liquidity, you suffocate the applications.
Surviving the winter to harvest the spring—that was my mantra during the 2022 bear market. I spent those months auditing bridge audits and governance failures. I wrote a post-mortem series on 20 failed protocols. The common thread wasn't bad code; it was liquidity fragmentation. Projects died because they couldn't sustain the activity on their own isolated chain. The same will happen to most L2s in this bull market.
My contrarian take: By 2026, 80% of today's L2s will be zombie chains with near-zero activity. The value will concentrate back to Ethereum L1 and one or two 'winning' L2s that achieve critical mass.
Why? Because users are lazy. They want to click one button and access all of DeFi. They don't want to bridge, unwrap, approve, and manage gas on five different chains. The recent success of Base is instructive: it benefited from Coinbase's massive distribution and direct fiat on-ramp. That's the kind of network effect that can break the fragmentation loop.
Decoding the signal from the blockchain noise—the signal is that value accrues to aggregators and unified liquidity layers, not to individual L2s. Projects like Across Protocol and Stargate that offer cross-chain swaps are seeing usage spike. That's where the fee revenue flows. The L2s themselves become commoditized settlement layers, like cloud providers. Do you care which AWS region your server runs on? No. You care about the service. Similarly, users will care about the application, not the rollup it runs on.
Let me ground this in numbers. I pulled data on L2 revenue over the last six months. Only Arbitrum and Optimism generate meaningful sequencer fees—about $2 million per month each. The rest are burning cash on sequencer costs and incentive programs. zkSync Era, despite its hype, is barely breaking even. Blast is hemorrhaging TVL as its point system expires.
The bull market euphoria masks this. Everyone is celebrating the 'L2 ecosystem' while ignoring the fundamentals. New L2s launch every week, each raising $20–50 million at billion-dollar valuations. They live on marketing hype and temporary incentives. But once the liquidity mining ends, users leave. It's the same playbook from 2021's DEX wars, just repackaged for scaling.
Chasing the ghost of 2017's fever dream—back then, we chased ICO tokens. Now we chase L2 point programs. The mechanics are the same: issue a token, attract liquidity, boost metrics, dump on retail. The difference is the complexity. Back then, you just bought ETH and sent it to a smart contract. Now you need to bridge, wrap, stake, and interact with 15 dApps to qualify for airdrops. The friction alone filters out 99% of potential users.
We are building the financial equivalent of a Rube Goldberg machine. It's impressive to watch, but it's not efficient.
I've been writing about this since 2020. Back then, I published a report on Uniswap's AMM model and how it democratized liquidity. That report reached 50,000 readers. The lesson was: simplicity wins. Uniswap beat order book DEXes because it was one-click. Now, we've gone backwards. We've re-introduced complexity.
Structuring chaos into profitable narratives—that's my job. I see the L2 chaos and I look for the alpha. Where is the opportunity? In the aggregation layer. In tools that abstract away the L2 fragmentation. In wallets that auto-route through the cheapest bridge. In indexes that give exposure to the 'L2 market' without forcing users to pick winners.
The market hasn't priced this yet. Most investors are still betting on individual L2 tokens. They think 'more L2s = more total value.' But the data says otherwise. The total value across all L2s is growing slower than L1 alone. The L2 sector is cannibalizing itself.
Look at the stablecoin flows. USDC and USDT supply on L2s has been flat since August 2024. Meanwhile, the total supply on Ethereum L1 has grown 15%. The new money is staying on base layer. Why? Because institutions that entered via the ETF and direct fiat ramps prefer simplicity. They buy ETH on Coinbase, use a hardware wallet, and hold. They don't want to bridge to Arbitrum to farm 5% yield.
The illusion of value in digital scarcity—stablecoins are the ultimate test. If L2s were truly scaling adoption, we'd see stablecoin supply shift disproportionately. We don't. The data shows L2s are mostly recycling the same capital through airdrop farming loops.
So what's the takeaway for this bull market?
Don't chase every new L2 token. The winners will be the ones with the strongest distribution (like Base, OP Stack) or the ones that become modularly integrated (like Arbitrum's Orbit). The rest will be dust. The narrative will shift from 'L2 proliferation' to 'L2 consolidation' by 2026. Watch for acquisitions, merges, and finality.
Next cycle. Same game. Better odds.—but only if you understand the structure.
In my experience, the best trades are contrarian. When everyone is bullish on L2s, sell the fragmentation. When everyone has given up on L1s because 'everything is going to L2,' buy ETH. The base layer captures the settlement fees of all those L2s. It's the ultimate zero-fragmentation asset.
The market is currently mispricing the cost of fragmentation. It's treating each new L2 as an independent network. In reality, they are tenants on the same landlord's property. The landlord — Ethereum — will collect the rent.
I've seen this movie before. In 2021, everyone said 'sidechains will kill Ethereum.' Then Polygon's fees were usurped by Arbitrum. In 2022, everyone said 'move to Solana.' Then FTX collapsed and liquidity flowed back to ETH. In 2023, the mantra was 'rollup-centric roadmap.' Now, in 2025, we're seeing the cracks.
Alpha extracted. Noise filtered.
The signal is clear: the next narrative isn't L2 vs L1. It's unified liquidity. Protocols that can bridge the fragmentation—intelligently, programmatically, seamlessly—will capture the next wave of value. Think of it as the 'HTTP of crypto.'
But that's a story for another article.
For now, I'll leave you with a framework: when evaluating any new L2, ask three questions:
1. What unique distribution advantage does this L2 have that others cannot replicate? If the answer is 'just a token incentive,' walk away.
2. How long can it sustain operations without incentive budgets? If the answer is less than 18 months, it's a zombie in waiting.
3. Does it meaningfully reduce user friction compared to existing L2s? If the answer is 'no' or 'we have better ZK proofs,' the user doesn't care.
The vast majority will fail these tests. That's fine. The market needs to learn through failure. But you don't have to be the one paying the tuition.
Surviving the winter to harvest the spring—and in a bull market, the winter is just a metaphor for the correction that will follow the hype. Be ready.
Decoding the signal from the blockchain noise — the signal is consolidation. The noise is the endless parade of new L2 announcements. You know where to focus.