Hook: The Undervaluation Echo
Cadence CEO Anirudh Devgan recently argued that his company is undervalued amid the AI boom. He’s right. But the same logic applies to a different class of critical infrastructure: Ethereum Layer 2 rollups. The market is treating them as mere scaling solutions, forgetting that every DeFi app, every NFT mint, every cross-chain bridge depends on the block space they provide. The front-runner didn’t get the memo. The real value is in the tools, not the tokens.
Context: The “Shovel Seller” in a Gold Rush
In the semiconductor world, EDA tools like Cadence’s are the shovels. Every AI chip—whether from NVIDIA, AMD, or a hyperscaler’s ASIC—requires EDA software to design. The market values Cadence at ~$80B, but its CEO says it should be higher because the AI-driven design cycle is just beginning. Similarly, Ethereum Layer 2s are the shovels for the on-chain economy. Every transaction on Optimism, Arbitrum, or zkSync consumes their block space. Yet the total market cap of these L2 tokens is less than $20B, while the value of assets they secure (bridged TVL) exceeds $100B. The asymmetry is glaring.
A bug is just a feature that hasn’t been exploited by the market’s valuation model. The L2 “bug” is that the market prices them as simple transaction processors, not as the foundational layer of a new financial system. My own audit of the EOS launch in 2017 taught me that hype often obscures real structural advantage. L2s have structural advantage: they inherit Ethereum’s security while offering programmable blockspace. The market is ignoring this.
Core: Systematic Teardown of L2 Valuation
1. Technical Architecture as Moats
Just as Cadence’s EDA suite covers the entire chip design flow (RTL to GDSII), L2s provide a complete execution environment: sequencer, batcher, fraud/validity proof, and bridge. The moat is not just in the code but in the network effects of adoption. Arbitrum’s Orbit stack and Optimism’s OP Stack are being used by dozens of app chains. This is analogous to Cadence’s PDK (process design kit) partnerships with foundries. The more developers build on a stack, the harder it is to switch.
2. The “Tool Tax” and Value Capture
Cadence earns a recurring subscription fee from every chip design. L2s earn a fee from every transaction (sequencer revenue). But unlike software subscriptions, L2 fees are directly proportional to the economic activity of the ecosystem. As DeFi volumes grew 10x in 2023-2024, L2 revenue grew 8x for Arbitrum and 12x for Optimism (based on public data). Yet their token prices barely moved. The hidden information: the market is still pricing L2s as speculative tokens, not as infrastructure that collects a “network tax”.
3. R&D Intensity and Platform Evolution
Cadence spends ~30% of revenue on R&D, developing AI-EDA and system-level design. L2s spend even more proportionally: Arbitrum’s Offchain Labs has ~200 engineers, and Optimism’s OP Labs has ~150, representing 40-50% of their token-based budgets. This R&D is not just for scaling; it’s for composability, interoperability, and native yield. The transition from “rollup as a service” to “settlement layer as a platform” mirrors Cadence’s move from EDA to system design platform. The market hasn’t priced this platform shift.
4. Liquidity Fragmentation Is a Feature, Not a Bug
Critics say L2s fragment liquidity. But fragmentation is a symptom of specialization. Just as Cadence’s tools serve different chip types (custom, analog, digital), L2s serve different execution environments (general EVM, zkEVM, parallelized VM). The value is in the unification layer—the bridge and settlement—which L2s control. The “liquidity fragmentation” narrative is a manufactured VC story to push new products. The real problem is that market makers haven’t built efficient cross-L2 arbitrage bots. The infrastructure is ready; the market is slow.
5. Geopolitical and Regulatory Tailwinds
Export controls on EDA tools highlight their strategic importance. Similarly, regulatory clarity on L2s (e.g., the SEC’s recent approval of ETH futures ETFs implying L2s are not securities) is a massive tailwind. In the EU, MiCA treats L2 tokens as “utility tokens,” reducing legal risk. The combination of regulatory acceptance and sovereign demand for blockchain infrastructure (e.g., central bank digital currencies) will make L2s the critical infrastructure of the next decade. The market is ignoring this geopolitical premium.
Contrarian: What the Bulls Got Right
Bulls argue that L2s benefit from Ethereum’s security and brand. They are correct. But they miss the inverse leverage: each L2 token represents a claim on the future blockspace of that chain. As Ethereum’s base layer becomes more expensive, L2s will capture the majority of user activity. The “fee market” on L2s is already growing faster than on mainnet. The bulls underestimated the stickiness of developer mindshare. Once a developer builds on Arbitrum, migrating to a new L2 is costly (similar to switching EDA vendors). The true moat is the ecosystem of tools, wallets, and dApps built around each L2.
Takeaway: The Accountability Call
If you believe that the on-chain economy will grow to trillions of dollars in value, then L2s are the toll booths. The market is pricing them as penny stocks. The front-runner didn’t. The question is not whether L2s are undervalued, but whether the market will recognize the “tool tax” model before the next DeFi summer. Based on my audit of the 2020 Uniswap V2 front-running chaos, I learned that the market always underestimates infrastructure. The same is happening now. Don’t be the one selling shovels for free.
