Every financial institution in the world is, at its core, a trust settlement engine. Banks settle transactions. Exchanges settle trades. Clearinghouses settle derivatives. The entire superstructure of global finance is built on the problem of making two parties who don't know each other confident that value has changed hands reliably.
Ethereum was designed to solve this problem with math rather than institutions. A transaction on Ethereum settles with cryptographic finality — not because a bank guaranteed it, but because the network's consensus mechanism made any alternative economically irrational. That is the trust innovation.
The problem is throughput. Ethereum's base layer processes roughly 15 to 30 transactions per second. Visa processes around 24,000. The global financial system, if it were to run on Ethereum's base layer, would require something like 10,000 times its current capacity. This is not a theoretical constraint; it is a practical ceiling on how much of the world's financial activity can settle on Ethereum directly.
Layer 2 networks — rollups, in the technical vocabulary — were built to close that gap. They are the subject of what has become one of the most consequential infrastructure competitions in financial technology, and most of the people who will eventually use this infrastructure have never heard of it.
What Rollups Actually Do
A rollup is a blockchain that executes transactions off Ethereum's main chain — the base layer, often called Layer 1 — but periodically posts compressed proofs of those transactions back to Ethereum for final settlement. The base layer provides the security guarantee; the rollup provides the throughput.
The analogy that clarifies this: Ethereum's base layer is the central bank. It settles the final, authoritative record. Layer 2 rollups are the commercial banking system — they handle the vast majority of day-to-day transactions and periodically net everything out against the central record. The trust still ultimately flows from the base layer, but the volume is handled by the layer above it.
There are two main categories of rollup architecture. Optimistic rollups — Arbitrum and Optimism are the dominant examples — assume transactions are valid by default and only run verification computations if someone challenges a transaction. This approach is simpler to implement and has been in production longer. Zero-knowledge rollups, or ZK rollups — zkSync and StarkNet are the leading examples — use cryptographic proofs to verify every transaction batch mathematically before posting it to Ethereum. This is more computationally intensive but provides stronger guarantees and faster finality.
Both architectures process transactions at a fraction of Ethereum's base-layer cost — typically 90% to 99% cheaper per transaction — while inheriting Ethereum's security guarantees through the settlement mechanism.
The Major Competitors
The Layer 2 landscape has consolidated around four serious contenders, though the field is not static.
Arbitrum, built by Offchain Labs, launched in 2021 and currently holds the largest total value locked (TVL) among Layer 2 networks — consistently above $15 billion, with significant variation as market conditions shift. Arbitrum One, its flagship network, hosts a substantial ecosystem of DeFi protocols, and Arbitrum Orbit allows other projects to build their own chains that settle on Arbitrum rather than directly on Ethereum. The effect is a fractal structure: chains built on chains, all anchoring to Ethereum's trust.
Optimism, from OP Labs, launched its Optimism Mainnet in 2021 and has since developed the OP Stack — an open-source framework for building rollup chains. The OP Stack's most significant consequence is Base, a Layer 2 built by Coinbase using the OP Stack technology. Base launched in 2023 and has grown rapidly, processing over 4 million transactions per day at peak periods. When a company with Coinbase's distribution builds a Layer 2, the user acquisition dynamics change substantially.
zkSync Era, from Matter Labs, uses ZK rollup technology and has positioned itself on the argument that zero-knowledge proofs represent the architecturally superior approach for long-term security. The ZK approach requires more sophisticated cryptography, but the resulting proofs are verifiable without trusting the prover — a meaningful security property for applications where finality guarantees are critical.
StarkNet, from StarkWare, is the other major ZK rollup and has developed STARK proofs — a cryptographic approach that does not require a trusted setup (a potential vulnerability in some ZK proof systems). StarkWare has positioned StarkNet as the infrastructure layer for high-throughput applications and has attracted institutional attention from firms including Sequoia, Tiger Global, and Paradigm.
The competition between optimistic and ZK rollups is not purely technical — it is also a question of timing. Optimistic rollups are deployed, battle-tested, and processing real volume today. ZK rollups offer superior long-term security properties but are more complex and earlier in their maturity curve. The question is not which architecture is better in theory; it is which one can capture enough adoption to sustain its network effects through the transition period.
Why This Matters for Finance
The Layer 2 competition is not primarily about cryptocurrency speculation. It is about which networks become the settlement rails for programmable finance at scale.
BlackRock's BUIDL fund — a tokenized money market fund launched in 2024 — settled initially on Ethereum's base layer. As institutional tokenization of real-world assets scales, the economics of base-layer settlement become prohibitive for high-frequency transfers and small-denomination transactions. Layer 2 networks are the natural next step for institutional deployments that require both Ethereum's security guarantees and lower transaction costs.
PayPal's PYUSD stablecoin has expanded to Solana — a competitor to Ethereum — partly because Solana's throughput and cost structure are more favorable for payment-volume use cases. The countermove available to Ethereum is Layer 2 networks that can compete on throughput and cost while retaining the security and institutional credibility that Ethereum's base layer provides. Base, as a Coinbase product, is the clearest example of how that competition is playing out: a Layer 2 that inherits Ethereum's trust but operates at Solana-competitive speeds and costs.
The longer-term question is whether the financial system's eventual programmable infrastructure runs on Ethereum and its Layer 2 ecosystem, on Solana, on a central bank digital currency rail, or on something that does not yet have significant traction. There is no settled answer. What is clear is that the Layer 2 networks being built and hardened now are attempting to position themselves as essential infrastructure before that question is resolved.
The Fragmentation Problem
The proliferation of Layer 2 networks creates a problem that does not yet have a clean solution: liquidity and asset fragmentation.
USDC on Arbitrum and USDC on Base are both USDC, but they are not interchangeable without a bridging operation — transferring assets from one Layer 2 network to another. Bridging introduces cost, delay, and risk. Several hundred million dollars in assets have been lost to bridge exploits since 2021. The cross-chain bridge is currently one of the most attacked surfaces in the ecosystem.
The solution under active development is interoperability standards — protocols that allow assets and messages to move between Layer 2 networks without manual bridging. ERC-7281 (xERC-20) and various cross-chain messaging protocols represent the current state of this work. None has achieved the clean composability that exists within a single chain.
This fragmentation problem is not fatal to the Layer 2 model, but it is a real constraint on usability that will need to be resolved before programmable finance at scale can operate seamlessly across the ecosystem. The network that first solves cross-L2 composability cleanly will have a meaningful competitive advantage.
What to Watch
The meaningful signals in the Layer 2 competition over the next two years are not token prices. They are institutional deployment decisions — which networks BlackRock, JPMorgan, and large stablecoin issuers choose for their next real-world asset tokenization projects — and developer activity, measured by which network attracts the most sophisticated new protocol deployments.
Base's trajectory is the most interesting near-term data point. Coinbase's distribution reaches over 100 million verified users. If Base succeeds in converting a meaningful fraction of that user base into active Layer 2 participants, the network effects compound quickly. The rollup that owns Coinbase's distribution effectively owns a significant share of retail entry into programmable finance.
The infrastructure being built in the Layer 2 wars is not speculative in the way that most cryptocurrency commentary suggests. It is the plumbing for a financial system that moves value the way the internet moves information — permissionlessly, globally, cheaply, and with cryptographic guarantees that do not require trusting any particular institution. Whether that transition takes five years or twenty, the networks being hardened now are positioning to be its foundation.