What's next in 2023 for account abstraction?

Authored by Kiwi Yao, researcher @OKX Ventures

2023 was a big year for account abstraction (AA), as the arrival of the ERC-4337 Ethereum standard brought new opportunities for users of the network. ERC-4337 gave access to smart wallet contracts, and its progress since launch has brought new capabilities.

So, what's next for account abstraction in 2023?

Who are the main account abstraction market participants?

Today there are two types of participants in the AA market.

The first is smart contract wallets, with features such as social login, social recovery, gas abstraction, transaction batching, integration, and aggregation of third-party services (such as fiat on and off-ramp, and decentralized finance (DeFi) protocols).

Currently, most smart contract wallets are in the early stages of transitioning to EIP-4337. For example, Unipass has achieved EIP-4337 compatibility and is actively developing an AA solution with Keystone. Visa, as a leading Web2 player, supports automatic payments on StarkNet and is also actively exploring the AA theme. Many early stage projects are also emerging in this area.

The second kind of participant is third-party providers of infrastructure — such as bundlers or paymasters — that focus on modular design. StackUp and Biconomy are two of the more mature solutions on the market with their EIP-4337 software development kits (SDKs), while other projects are mostly in the early demo stage.

Pimlico, an infrastructure layer that's building the next generation of smart wallets, has a leading market share of 52% of user operations (user ops) in all-chain bundlers. The entity led the way for profitability on Polygon in July 2023 (earning 637 MATIC), and also showed strong profitability on Optimism.

StackUp, meanwhile, has bundler businesses on multiple chains, and had unmatched profitability on Arbitrum in July and August 2023, earning 0.4 and 5 ETH respectively. Although many bundlers fail today, significant profits are still possible.

What's StackUp?

StackUp is a provider of bundlers and paymasters. Its bundler is implemented in the Go language, meets all the requirements of a EIP-4377 test suite, and is completely open source and free to use.

The StackUp bundler supports different modes:

• Private mode: User ops enter a private mempool for slower transaction execution speed in exchange for better privacy. User ops will not be displayed in the public mempool, thereby avoiding maximal extractable value attacks.

• Searcher mode: Used by bot operators (also known as searchers, such as DeFi arbitrage bots) in the Ethereum ecosystem, integrated with block builders like Flashbot, to send UserOperations through MEV-Boost, allowing searchers to bid for specific transaction ordering for block builders to select the maximum MEV transaction ordering to build blocks.

The bundler also supports two types of Paymaster:

• Verification paymaster: Provides gas abstraction for off-chain transactions (such as fiat currency deposits and withdrawals). For example, users can choose to use a credit card subscription paymaster service to pay for gas fees. Users can also customize the logic of gas payments. StackUp will charge users on a pay as you go basis.

• Deposit paymaster: This allows users to pay gas fees with ERC20 tokens.

According to data from Dune, StackUp's bundle accounts for 90% of failed user ops (address 0x25Df024637d4e56c1aE9563987Bf3e92C9f534c0), but it's still the most profitable bundle address, earning 1,057 USD in July 2023.

What's more, StackUp has a bundler business for each chain, and its user op quantity accounts for 9.1% of the total bundle. It has reached profit in Optimism, Avalanche, and Polygon, alongside it's unmatched Arbitrum profitability in July and August 2023 as referenced above.

What's Pimlico?

Pimlico is an encrypted infrastructure for account abstraction that provides plug-and-play application programming interfaces (APIs) for wallets to sponsor user transactions.

Pimlico uses the Alto ERC-4337 bundle, which is written in TypeScript. Meanwhile, its paymaster supports ERC20 tokens. It also uses Chainlink for price oracles and calculates the price of ERC20 tokens in native tokens based on ERC20-to-USD and native token-to-USD prices. It can be deployed on any token-supported chain with Chainlink support, but is subject to Chainlink's security constraints.

Pimilico's user ops account for 52% of the entire bundle. Bundlers are mostly deployed on Polygon and Optimism — with profit gained on each — with Polygon being the most profitable, earning 637 MATIC in July 2023.

What's Biconomy?

Biconomy provides SDK functionality for developers using the ERC-4337 standard. The developer simplifies the process of creating and sending user ops, optimizing the development and management of decentralized applications (DApps).

With Biconomy's SDK, users can estimate Gas, send user operations, and retrieve user operation receipts. Basic paymaster services are also available, including obtaining quotes or paymaster data based on the provided user operation and paymasterServiceData.

Currently, Biconomy holds a market share of 15.07% in the user op market, with over 90% of user ops deployed on Polygon and Avalanche. The developer's share of the user op count in the entire chain bundle stands at 7.9% as of September 2023, with bundlers deployed on Polygon and Optimism operating at a loss. A 427 MATIC loss was recorded on Polygon in June 2023, before an 0.3 ETH loss on Optimism in August. Profit of 11 AVAX was seen on Avalanche in September 2023.

What are account abstraction related EIPs?

Ethereum improvement proposals (EIPs) related to ERC-4337 mainly revolve around optimizing AA deployment, because the official direction for ERC-4337 has been specified.

One auxiliary product derived from ERC-4337 is ERC-6551, an NFT tokenization protocol supporting the popularization of AAs, with similar underlying logic to ERC-4337.

Various other proposals and standards also exist to help optimize AAs:

• EIP 6662: Helps to improve the practical use cases of AA accounts, supporting a more user-friendly authentication model where users can use their wallets as login accounts for managing assets in DApps.

• ERC 6900: Supports modular account abstraction to enhance flexibility and ease of development for smart contracts.

• ERC 1271 and ERC 6492: Each primarily addresses signature verification issues in the ERC-4337 domain. ERC-6492 makes some improvements over ERC-1271, allowing users to verify signatures without deploying contracts.

• EIP-7204 and EIP-7197: Enable AA wallets to define token standards for each other.

Providing account migration assistance is ERC 7377, which helps facilitate the future migration of externally owned account wallets to smart contract wallets.

The roadmap for ERC-4337

In the short term, the focus for ERC-4337 is reportedly to expand the market and form a mutually beneficial relationship with layer 2. In the medium term, the focus is on implementing modular bundlers and paymasters, as well as the deployment of SDKs, while also optimizing the user experience regarding details (such as reducing gas costs and adding optional EOA-to-ERC-4337 conversion). In the long term, the consideration is reportedly to enforce the conversion of EOA wallets.

The future of EIP-4337 infrastructure?

As EIP-4337 infrastructure continues to evolve, many in the industry anticipate myriad changes relating to account abstraction.

Market changes

• The account abstraction market is still in its early stages, and many believe its future development depends on the adoption of EIP-4337 by the ecosystem. If the number of EOAs and smart contract wallets becomes equal, the market size could grow by 310 times.

• Currently, many smart contract wallets in the market have built their own infrastructure, but the adoption rate is low, potentially due to the incomplete status of EIP-4337. The infrastructure deployment process has yet to reach a permissionless state.

• Layer 2 hasn't implemented all the EIP-4337 interfaces, and the implementation of its account abstraction infrastructure is more complex than layer 1. Currently, the overall market is relatively quiet.

Product changes

• The account abstraction market requires much innovation, such as permissionless modular infrastructure, integration with existing fiat currencies and DeFi services, DApp SDK, and potential independent account layers. Each of the aforementioned are potential business opportunities in the future.

• The advantage of AA lies in its ability to provide more convenience for DApps in complex on-chain interactions. Business development is the key in this fierce competition, as AA can provide more benefits for DApps, so most wallets want to support ecosystems such as DeFi, GameFi, and persuade large-scale DApps.

• The monetization model of AA still needs further exploration. The industry leans towards the 'to customer' model to find high-value and sustainable application scenarios and profit through quantity.

What might the endgame be for infrastructure?

The future of account abstraction depends on the adoption of EIP-4337 by the ecosystem, and this market is still in its relative infancy. Assuming that the number of EOAs and smart contract wallets is equal, many predict that the market size of smart contract wallets has the potential to grow by 310%. It's therefore necessary to keep a close eye on adoption.

According to the latest data from Coin Metrics, there are approximately 100 million EOA addresses with non-zero balances in Ethereum. However, the total number of smart contract wallets, estimated based on the sum of Gnosis Safe (250,000) and Argent (76,000) accounts (which are the most widely used products), is only around 320,000.

Currently, many smart contract wallets have built their own infrastructure, but this infrastructure is not tailored to third-party use cases. Third-party infrastructure providers, such as Stackup, are developing modular bundlers and paymasters, but their deployment processes have not yet reached a permissionless state. Since EIP-4337 isn't finalized, the modular capabilities of this infrastructure hasn't been fully defined. Additionally, due to the low volume of user op transactions, wallets that have enabled signature aggregation (such as the BLS wallet) are still not mainstream.

Business development is the key to competition, and the advantage of AA lies not only in seamless logins (as Web3 authentication can be achieved through custodial means), but also in providing more benefits to DApps in complex and customized on-chain interactions. Therefore, most wallets want to support ecosystems such as DeFi and GameFi, and convince large DApps to find breakthroughs.

The monetization model of AA also needs further exploration. The 'to business' model may not generate much profit or accumulate its own user base, while the to customer model requires finding high-value use cases to profit based on quantity. Integrating derivatives and bridging functionalities can be profitable, but finding a sustainable model is crucial.

As account abstraction continues to evolve, many believe that during 2023 and into 2024, innovation is required across the key areas of permissionless modular infrastructure, integration with existing fiat and DeFi services, DApp SDKs, and potential independent account layers.