Every network effect faces cold-start problems, and the rollup landscape is no exception, currently grappling with sequencer centralization. Taiko, as a pioneer, has tackled the cold-start problem of rollup decentralization by enhancing L1 sequenced blocks—a technique known as “based rollups.”
Based rollups also means that MEV naturally flows to Ethereum, strengthening Ethereum’s economic security. There are some limitations to the revenue models of based rollups, as can be understood here. In this blog post, we’ll summarize the current economic landscape of rollups, and explore based rollups economy.
Rollups not only scale Ethereum, they also create their own economic models. Centralized rollups has their own transaction order mechanisms such as FCFS (First Come First Serve), PGA (Priority Gas Auction), Timeboost and more. Each mechanism has different economic opportunities for rollup operators.
The centralized sequencer’s exclusive rights to order transactions allow it to capture MEV and collect all priority fees associated with transaction inclusion—concentrating this power in a single entity.
2.1 OP Stack Chains Revenue, https://x.com/smyyguy/status/1831761080252539182
It’s important to separate the based rollup proposer and based rollup operator (note that operator can also be proposer) when discussing revenue. By “operator,” we’re referring to the based rollup team.
Based rollups are thought to be incapable of generating MEV due to their decentralized and permissionless nature of block building. This is because not all generated MEV (plus priority fee) is directed to the operator—since anyone can be a proposer, other proposers can also benefit from this MEV by proposing a block.
However, they will likely have to return most of this value to the L1. L2 Block proposers compete on priority fees to be included first in the L1 block. Consequently, the proposer willing to sacrifice the most profit wins.
3.1 Current Revenue Streams of Based Proposer
With its based and multi-proof structure, Taiko has a unique economic cycle.
Diagram 4.1 illustrates Taiko’s fee distribution and block building mechanism as a based rollup. When users submit transactions on Taiko L2, they pay fees split into two components: a priority tip and a base fee. The tip goes to the L2 block proposer, who builds and proposes new blocks. The base fee (currently 1 wei) is directed to the Taiko DAO Treasury. L2 block proposers must pay two fees: an L1 fee to the TaikoL1 contract on Ethereum, and a prover fee to the block prover. Block provers incur proof generation costs while proving the correctness of proposed blocks.
4.1 Taiko Block Lifecycle
Diagram 4.2 illustrates Taiko’s contestable multi-proof structure. The process begins when a proposer submits a new block, followed by a tier-1 prover who submits a proof with a TAIKO bond. During the 24-hour cooldown period, anyone can contest this proof by posting their own bond, as demonstrated by Cindy.
The system then supports two possible scenarios: If a higher-tier proof confirms the original proof was correct, the original prover receives back their bond plus a reward, while the contester loses their bond. Conversely, if the higher-tier proof shows the original was wrong, the contester receives back their bond plus a reward, and the original prover loses their stake.
4.2 Multi-proof Lifecycle (bond amounts may vary)
If the contester wins: The contester receives their contestation bond back plus 1/4 of the original prover’s validity bond. The new prover receives 1/4 of the original prover’s validity bond as a proving fee. The remaining 1/2 goes to the DAO treasury.
If the original prover wins: The original prover reclaims their validity bond and receives 1/4 of the contestation bond as a reward. The new prover (who may be the original prover) earns 1/4 of the contestation bond. The remaining 1/2 goes to the DAO treasury.
Currently, Taiko uses only SGX and Guardian (a multisig run by Taiko Labs) prover. We plan to integrate ZK proofs by the end of the year. After this integration, tiers will be as follows:
Tier 1 TEE
Tier 2 ZK
Tier 3 ZK+TEE
Tier 4 Guardian Minority
Tier 5 Guardian Major.
In the future, we plan to phase out the Guardian prover and move away from training wheels.
Taiko proposers generate revenue through transaction tips. For any given block, a Taiko proposer’s profitability is calculated as
[(gasPrice - baseFee) * gasUsed] - L1cost.
Let’s calculate the profitability for block #487373 as an example. By applying the above equation to each transaction in this block, we find that the proposer’s revenue amounts to 0.00337 ETH.
See the code; https://github.com/0xjunger/proposer_revenue/blob/main/revenue.py
If the fee paid for proposeBlock on L1 is 0.00128 ETH for a given block, the net profit is; 0.00337 ETH - 0.00128 ETH ≈ 0.00209 ETH (~$5.16).
https://etherscan.io/tx/0x8d44a9b62bec7c935dfdc25ff5e7a9b0d9953d0ca4419e36fa6e507fa94a5430
It’s worth noting that in based rollups, profitability is directly tied to the current L1 gas price and L2 gas used per block. In this example, the L1 gas price is 10 gwei, while the L2 gas used is approximately 20M gwei (~700 txs).
Taiko’s Current Profitability, growthepie.xyz/fundamentals/profit
The based operator can only capture the base fee from transactions. Our current implementation of EIP-1559 faces some challenges on the Taiko mainnet. Due to an issue in calculating the base fee in the smart contract, it’s difficult to increase this fee. The required protocol upgrade that will solve this issue will be implemented on November 15th.
After this upgrade, operators and proposers will be able to capture revenue from base fees. Currently, base fees accumulate in the TaikoL2 contract. However, due to the aforementioned issue (base fee = 1 wei), the balance is nearly zero for now. So the base fee will become a meaningful value after the upgrade.
At Taiko, we plan to gradually step back from block proposing and fully hand over this responsibility to the community in the future, under the control of the DAO. This is why we’re dedicated to boosting proposers’ profitability. As part of the Ontake fork, proposers will receive a significant portion of the base fee.
Preconfirmations will enhance user experience and increase profitability for based rollups. Profitability will rise due to batch proposing, user experience will improve thanks to instant transaction confirmations. However, this comes at a cost of slightly reduced liveness compared to a native based rollup. Proposers will need to register through a registry contract by staking collateral (can be TAIKO or ETH), which slightly raises the barrier to entry for proposing. It’s important to note that native based sequencing will always serve as a fallback option if a preconfirmer attempts to censor transactions. While they may seem similar to centralized sequencers, it’s crucial to remember that preconfirmers are L1 validators striving for maximum neutrality. Additionally, the implementation of the @mikeneuder/the-case-for-ilectra">inclusion list to the L1 will further enhance security and censorship resistance.
From revenue standpoint, preconfirmations create a new income stream for proposers. Once implemented, proposers—acting as preconfirmers—will offer users the option to pay an extra fee for quicker transaction confirmation. It’s crucial to reiterate that preconfirmations will also reduce L1 costs by enabling the proposal of multiple sequential blocks simultaneously.
We’ve shared our mental model on preconfirmations in depth here and here. We’re collaborating closely with Nethermind and Gattaca(Titan) to implement based preconfirmations. We have two working PoC internally. Additionally, we’ll soon unveil our final based preconfirmation design.
5.1 Based Proposer Revenue After Upgrade and Preconfirmations
An in-protocol market for buying and selling sequencing tickets can be built. Using a dynamic pricing mechanism, the based operator can target a specific number of tickets in circulation and adjust the price based on the existing supply. The original idea stems from L1 execution tickets.
Introduce PBS by allowing preconfirmers to outsource block building to specialized builders. This could potentially leverage the existing MEV-Boost PBS pipeline, enabling current L1 builders to construct preconfirmed L2 blocks as well. This concept is derived from Nethermind’s Taiko preconfirmation design.
and more.
As a new territory, based rollups need to be explored. We’re still experimenting with based sequencing from various perspectives. While based rollups are often thought to be incapable of generating MEV, this isn’t entirely accurate. Based proposers can capture value through transaction ordering, allowing them to earn from base fees, priority fees, and any residual MEV. The protocol design ensures these opportunities are equitably accessible due to its decentralized proposer structure.
Based rollups also offer additional revenue streams. After implementing preconfirmations, proposers will be able to charge a fee from users for instant transaction confirmations as credibly neutral L1 validators. Furthermore, various methods can be implemented to increase revenue, such as selling sequencing tickets and L2 PBS.
Every network effect faces cold-start problems, and the rollup landscape is no exception, currently grappling with sequencer centralization. Taiko, as a pioneer, has tackled the cold-start problem of rollup decentralization by enhancing L1 sequenced blocks—a technique known as “based rollups.”
Based rollups also means that MEV naturally flows to Ethereum, strengthening Ethereum’s economic security. There are some limitations to the revenue models of based rollups, as can be understood here. In this blog post, we’ll summarize the current economic landscape of rollups, and explore based rollups economy.
Rollups not only scale Ethereum, they also create their own economic models. Centralized rollups has their own transaction order mechanisms such as FCFS (First Come First Serve), PGA (Priority Gas Auction), Timeboost and more. Each mechanism has different economic opportunities for rollup operators.
The centralized sequencer’s exclusive rights to order transactions allow it to capture MEV and collect all priority fees associated with transaction inclusion—concentrating this power in a single entity.
2.1 OP Stack Chains Revenue, https://x.com/smyyguy/status/1831761080252539182
It’s important to separate the based rollup proposer and based rollup operator (note that operator can also be proposer) when discussing revenue. By “operator,” we’re referring to the based rollup team.
Based rollups are thought to be incapable of generating MEV due to their decentralized and permissionless nature of block building. This is because not all generated MEV (plus priority fee) is directed to the operator—since anyone can be a proposer, other proposers can also benefit from this MEV by proposing a block.
However, they will likely have to return most of this value to the L1. L2 Block proposers compete on priority fees to be included first in the L1 block. Consequently, the proposer willing to sacrifice the most profit wins.
3.1 Current Revenue Streams of Based Proposer
With its based and multi-proof structure, Taiko has a unique economic cycle.
Diagram 4.1 illustrates Taiko’s fee distribution and block building mechanism as a based rollup. When users submit transactions on Taiko L2, they pay fees split into two components: a priority tip and a base fee. The tip goes to the L2 block proposer, who builds and proposes new blocks. The base fee (currently 1 wei) is directed to the Taiko DAO Treasury. L2 block proposers must pay two fees: an L1 fee to the TaikoL1 contract on Ethereum, and a prover fee to the block prover. Block provers incur proof generation costs while proving the correctness of proposed blocks.
4.1 Taiko Block Lifecycle
Diagram 4.2 illustrates Taiko’s contestable multi-proof structure. The process begins when a proposer submits a new block, followed by a tier-1 prover who submits a proof with a TAIKO bond. During the 24-hour cooldown period, anyone can contest this proof by posting their own bond, as demonstrated by Cindy.
The system then supports two possible scenarios: If a higher-tier proof confirms the original proof was correct, the original prover receives back their bond plus a reward, while the contester loses their bond. Conversely, if the higher-tier proof shows the original was wrong, the contester receives back their bond plus a reward, and the original prover loses their stake.
4.2 Multi-proof Lifecycle (bond amounts may vary)
If the contester wins: The contester receives their contestation bond back plus 1/4 of the original prover’s validity bond. The new prover receives 1/4 of the original prover’s validity bond as a proving fee. The remaining 1/2 goes to the DAO treasury.
If the original prover wins: The original prover reclaims their validity bond and receives 1/4 of the contestation bond as a reward. The new prover (who may be the original prover) earns 1/4 of the contestation bond. The remaining 1/2 goes to the DAO treasury.
Currently, Taiko uses only SGX and Guardian (a multisig run by Taiko Labs) prover. We plan to integrate ZK proofs by the end of the year. After this integration, tiers will be as follows:
Tier 1 TEE
Tier 2 ZK
Tier 3 ZK+TEE
Tier 4 Guardian Minority
Tier 5 Guardian Major.
In the future, we plan to phase out the Guardian prover and move away from training wheels.
Taiko proposers generate revenue through transaction tips. For any given block, a Taiko proposer’s profitability is calculated as
[(gasPrice - baseFee) * gasUsed] - L1cost.
Let’s calculate the profitability for block #487373 as an example. By applying the above equation to each transaction in this block, we find that the proposer’s revenue amounts to 0.00337 ETH.
See the code; https://github.com/0xjunger/proposer_revenue/blob/main/revenue.py
If the fee paid for proposeBlock on L1 is 0.00128 ETH for a given block, the net profit is; 0.00337 ETH - 0.00128 ETH ≈ 0.00209 ETH (~$5.16).
https://etherscan.io/tx/0x8d44a9b62bec7c935dfdc25ff5e7a9b0d9953d0ca4419e36fa6e507fa94a5430
It’s worth noting that in based rollups, profitability is directly tied to the current L1 gas price and L2 gas used per block. In this example, the L1 gas price is 10 gwei, while the L2 gas used is approximately 20M gwei (~700 txs).
Taiko’s Current Profitability, growthepie.xyz/fundamentals/profit
The based operator can only capture the base fee from transactions. Our current implementation of EIP-1559 faces some challenges on the Taiko mainnet. Due to an issue in calculating the base fee in the smart contract, it’s difficult to increase this fee. The required protocol upgrade that will solve this issue will be implemented on November 15th.
After this upgrade, operators and proposers will be able to capture revenue from base fees. Currently, base fees accumulate in the TaikoL2 contract. However, due to the aforementioned issue (base fee = 1 wei), the balance is nearly zero for now. So the base fee will become a meaningful value after the upgrade.
At Taiko, we plan to gradually step back from block proposing and fully hand over this responsibility to the community in the future, under the control of the DAO. This is why we’re dedicated to boosting proposers’ profitability. As part of the Ontake fork, proposers will receive a significant portion of the base fee.
Preconfirmations will enhance user experience and increase profitability for based rollups. Profitability will rise due to batch proposing, user experience will improve thanks to instant transaction confirmations. However, this comes at a cost of slightly reduced liveness compared to a native based rollup. Proposers will need to register through a registry contract by staking collateral (can be TAIKO or ETH), which slightly raises the barrier to entry for proposing. It’s important to note that native based sequencing will always serve as a fallback option if a preconfirmer attempts to censor transactions. While they may seem similar to centralized sequencers, it’s crucial to remember that preconfirmers are L1 validators striving for maximum neutrality. Additionally, the implementation of the @mikeneuder/the-case-for-ilectra">inclusion list to the L1 will further enhance security and censorship resistance.
From revenue standpoint, preconfirmations create a new income stream for proposers. Once implemented, proposers—acting as preconfirmers—will offer users the option to pay an extra fee for quicker transaction confirmation. It’s crucial to reiterate that preconfirmations will also reduce L1 costs by enabling the proposal of multiple sequential blocks simultaneously.
We’ve shared our mental model on preconfirmations in depth here and here. We’re collaborating closely with Nethermind and Gattaca(Titan) to implement based preconfirmations. We have two working PoC internally. Additionally, we’ll soon unveil our final based preconfirmation design.
5.1 Based Proposer Revenue After Upgrade and Preconfirmations
An in-protocol market for buying and selling sequencing tickets can be built. Using a dynamic pricing mechanism, the based operator can target a specific number of tickets in circulation and adjust the price based on the existing supply. The original idea stems from L1 execution tickets.
Introduce PBS by allowing preconfirmers to outsource block building to specialized builders. This could potentially leverage the existing MEV-Boost PBS pipeline, enabling current L1 builders to construct preconfirmed L2 blocks as well. This concept is derived from Nethermind’s Taiko preconfirmation design.
and more.
As a new territory, based rollups need to be explored. We’re still experimenting with based sequencing from various perspectives. While based rollups are often thought to be incapable of generating MEV, this isn’t entirely accurate. Based proposers can capture value through transaction ordering, allowing them to earn from base fees, priority fees, and any residual MEV. The protocol design ensures these opportunities are equitably accessible due to its decentralized proposer structure.
Based rollups also offer additional revenue streams. After implementing preconfirmations, proposers will be able to charge a fee from users for instant transaction confirmations as credibly neutral L1 validators. Furthermore, various methods can be implemented to increase revenue, such as selling sequencing tickets and L2 PBS.