The Rivalry Between EVM and L1s Will Shape the Future of DeFi (Opinion)

By Piers Ridyard, CEO of RDX Works

The 2018-19 bear market saw the development of the MetaMask wallet, Uniswap decentralized exchange, OpenSea NFT marketplace, and alternative Layer 1s such as Solana. Only with this kind of core infrastructure in place was the subsequent 2021 boom in DeFi and NFTs made possible.

A similar story is playing out today. Emerging from the rubble are two competing visions vying to become the core infrastructure of the next cycle:

The incumbent Ethereum and its ecosystem of Layer 2 (L2) scaling networks, such as Arbitrum and Polygon that run the Ethereum Virtual Machine (EVM).

A new cohort of Layer 1s (L1s) have purposefully avoided the EVM and aim for an enhanced wallet user experience, application development environment, and scalability, with networks such as Aptos, Radix, and Sui being the prime examples.

EVM Layer 2s: Scaling The Incumbent

The EVM is the dominant platform in Web3 today, accounting for ~95% of all DeFi assets under management (AUM), ~80% of active addresses, and ~40% of all Web3 developers.

This success has led to Ethereum’s congestion and high transaction fees. The almost universally accepted solution: L2 scaling networks.

L2s are separate networks, offering their own ledger, tokens, and decentralized applications (dApps.) Their defining feature is that they periodically post summaries of their transactions back to the L1, Ethereum, piggybacking on the L1 to guarantee that transactions won’t roll back.

These L2s offer the same application development environment as Ethereum, the EVM. This allows for any dApp built on Ethereum to be easily copied over to an L2. From DEXes to lending to NFTs, dApps copied over can benefit from a new network that has higher throughput and lower fees yet inherits some of the security of Ethereum itself.

But there are issues with this approach.

First, security and developer experience continues to be a major concern. From the original hack of The DAO in 2016 through the billions of dollars lost annually over 2021-2022, the EVM has proven time and again that dApps built with it cannot safeguard users’ funds.

Second, the UX is far from mainstream-ready. The EVM places a high technical burden on its users, including “blind signing” – equivalent to signing a blank check for every transaction; “seed phrases” – a password that must be kept secure, else you may lose all your assets; or the need to be wary of “malicious tokens” that could steal your assets.

The requirement to maintain backward compatibility means solutions tend to be additive, piling up more complexity and risk rather than making the deep-rooted changes needed to fix issues properly. A case in point, ERC-4337 Account Abstraction, which is Ethereum’s solution to seed phrases, proposes an entirely new “mempool” through which transactions must be routed.

Third, L2s only half-solve the problem of scalability as each new network is like a new island with its own dApps and liquidity, not “composable” with the Ethereum mainland or other L2s. For this reason, we shall continue to see projects prioritize being on Ethereum, or in the scenario that an L2 gains enough traction to provide a compelling alternative, it will ultimately itself become congested, taking us back to square one.

Non-EVM L1s: The Challengers

Rather than iterate on the EVM, a new batch of L1s are charting their own path, starting from scratch with their own custom stacks.

First, they differentiate by addressing the neverending hacks and exploits through an improved developer experience. To achieve this, some projects, for example, have turned smart contracts containing assets into physical objects that can be “moved” between owners, with features to improve the security of tokens and smart contracts.

At the same time, other protocols have taken the object model one step further, with all assets being natively governed by a “DeFi Engine.” Similar to how Game Engines reduced bugs and improved game developer productivity by natively governing behaviors such as physics and gravity, this same concept is now being applied to finance.

In fact, assets being native to the ledger isn’t just a benefit for developers. It is a prerequisite to an improved user experience. By natively understanding assets, these platforms can provide users with human-readable transactions that guarantee what the transaction is going to do.

This solves the blank check “blind signing” transactions that the EVM and its L2s are architecturally unable to fix, as they can’t offer guarantees on something they don’t natively understand.

On the subject of scalability (the very problem that L2s were built to solve), new approaches promise to offer “linear scalability” without compromising that all-important composability.

This includes “intra-validator sharding,” which allows for each computer that validates transactions to actually be composed of many different underlying computers, or “multi-shard consensus.”

This allows for parallelization of processing across multiple groupings of computers. In each of these cases, adding more computers to the network allows for more transactions to be processed, similar to how the internet itself scales.

The Fight Ahead

Despite the technical advantages offered by the latest L1s, decentralized networks are all about community and momentum. The EVM and its L2s hold a significant lead in public awareness, developer community, and general tooling and infrastructure.

Getting developers to learn a new language and for users to adopt a new chain amongst all the noise is not easy and depends on how well the value proposition of that new chain can be propagated.

But, taking a step back – DeFi and Web3 account for only 0.01% of global financial assets, 0.1% of internet users, and 0.1% of global developers. The journey ahead is long, and there is still ample opportunity for newer platforms with radically different approaches and significantly less technical debt to fight for the remaining 99.9%.

Author bio

Piers Ridyard is the CEO of RDX Works, a public protocol and ledger for DeFi. Piers started in crypto when he started mining on the genesis block of Ethereum in early 2015, investing in “The DAO” and going deep on everything from game theory to prediction markets. This eventually led him to build and exit Surematics, a YCombinator company that built decentralized dealroom software for insurance companies in 2017. Piers became CEO of RDX Works in 2017, joining the Founder, Dan Hughes, and building the team to over 75 people around the world. His background includes finance, law, electronics, and mathematics. He also has two degrees, one in Chinese and Business and a second in Law, as well as having achieved his level 1 Chartered Financial Analyst designation.