Block Builder Profitability | November 2023
This report offers insight into block-building profitability for November 2023.
TL;DR
- We introduce the concept of adjusted profit to explain a builder’s on-chain net profit.
- Total adjusted profit in the month of November is 1,246 ETH
- Five builders dominate the market (with 92% of total adjusted profit)
- 14 builders each exceeded adjusted profit of 1 ETH in the month of November (on-chain profit only)
- There are still examples of successful, lesser-known builders, example: Anon:0x83bee
- Currently available leaderboards do not show accurate profits
- Impostor builders continue to spoof Flashbots builder
Most Profitable Builders - Earned at least 1 ETH:
Rank | Builder | Adj. Profit | MEV-Boost Value | Blocks | Profitable Blocks | Subsidy Blocks | Zero Profit Blocks | Total Subsidy |
---|---|---|---|---|---|---|---|---|
1. | beaverbuild |
712.299 | 8,322.352 | 57,713 | 30,546 | 0 | 27,167 | 0 |
2. | rsync-builder |
230.291 | 8,451.581 | 58,046 | 16,811 | 9,626 | 31,609 | -21.375 |
3. | Anon:0x83bee |
77.413 | 207.993 | 293 | 293 | 0 | 0 | 0 |
4. | builder0x69 |
69.569 | 1,081.899 | 9,567 | 1,849 | 7,718 | 0 | -2.436 |
5. | Titan Builder |
59.053 | 5,165.841 | 43,996 | 12,094 | 5,449 | 26,453 | -1.77 |
6. | Anon:0xb3a6d |
54.068 | 472.634 | 176 | 176 | 0 | 0 | 0 |
7. | BuildAI |
24.253 | 134.329 | 469 | 370 | 99 | 0 | -74 |
8. | Gambit Labs |
19.822 | 180.223 | 2,608 | 575 | 1,754 | 279 | -6.763 |
9. | f1b-test.io |
5.824 | 5.878 | 2 | 2 | 0 | 0 | 0 |
10. | payload.de |
4.223 | 8.475 | 62 | 46 | 16 | 0 | -9 |
11. | f1b.io |
3.371 | 31.967 | 396 | 165 | 231 | 0 | -113 |
12. | https://eth-builder.com |
2.423 | 133.912 | 293 | 208 | 85 | 0 | -122 |
13. | lokibuilder.xyz |
1.320 | 66.106 | 1,536 | 183 | 1,283 | 70 | -7.136 |
14. | nfactorial.xyz |
1.233 | 72.589 | 109 | 109 | 0 | 0 | 0 |
15-35. | Others, with profit |
2.370 | 1,979.781 | 20,503 | 360 | 96 | 20,047 | -0.0047 |
Notes:
- Official builder pubkeys were used when available (published on docs or websites of the builder), for other builders extra_data was used to identify the builder
- Two anonymous builders (tagged as
Anon:0x...
) were tagged as they were significantly profitable, compared to other builders that did not use any extra data - The category ‘Others’ is grouped by all block builders that still made some profit or were balance neutral; all builders with negative profit were excluded
- Query for generating this results can be found here: BigQuery stored query
Methodology For Profit Calculations and Their Adjustments
Block builder profit is defined as a difference between total value received by the builder address and value sent. The simple formula for block builder profit is:
(miner_rewards + builder_payments) - sent_from_builder
This simple method calculates the balance differences on block builder addresses (similar to the method relayscan uses), but we apply specific adjustments to arrive at an adjusted profit value.
Why Adjustments Are Needed
The main reason for adjustments is the fact that some builders deviate from standard practices of building blocks and paying MEV-Boost rewards to the validators resulting in inaccurate calculations and leaderboards.
Summary of Adjustments
In the process of writing and researching for this report, five adjustments were considered:
Account Linked to Builder Profit
Account Linked to Builder Costs
High-fee Transactions from Account Linked to Builder
Withdrawals Considered in Block Value
Impostor Builders
Refer to the appendix for more detail about each adjustment.
Impact of Adjustments on Report
- Total profit with adjustments: 1246.20 ETH
- Total profit without adjustments: 1189.75 ETH
Note: Actual impact is higher as some adjustments are done in both directions, positive and negative, gross impact is closer to 70 ETH.
Most Affected Builders
- PenguinBuild
- Profit with adjustments: -13.46 ETH (without: -6.92 ETH)
High-fee Transactions from Account Linked to Builder
Withdrawals Considered in Block Value
- Profit with adjustments: -13.46 ETH (without: -6.92 ETH)
- Flashbots:
- Profit with adjustments: 0.009 ETH (without: 6.15 ETH)
Imposter Builders
- Profit with adjustments: 0.009 ETH (without: 6.15 ETH)
- Rsync
- Profit with adjustments: 230.29 ETH (without: 154.98 ETH)
Account Linked to Builder Profit
Account Linked to Builder Costs
- Profit with adjustments: 230.29 ETH (without: 154.98 ETH)
Most Profitable Blocks (Adjusted)
Builder | Block Number | Adj. Profit | MEV-Boost Value | Builder Received | Profit Margin |
---|---|---|---|---|---|
Anon:0xb3a6d | 18552552 | 17.246 | 9.413 | 26.660 | 64.69% |
beaverbuild | 18545397 | 11.933 | 21.065 | 32.998 | 36.16% |
builder0x69 | 18680805 | 11.775 | 34.357 | 46.133 | 25.52% |
beaverbuild | 18621443 | 11.491 | 9.260 | 20.752 | 55.37% |
builder0x69 | 18686685 | 11.276 | 16.038 | 27.315 | 41.28% |
Anon:0x83bee | 18622526 | 10.439 | 13.59 | 24.029 | 43.44% |
beaverbuild | 18531923 | 10.296 | 3.890 | 14.187 | 72.57% |
Gambit Labs | 18502213 | 10.087 | 0.050 | 10.137 | 99.5% |
beaverbuild | 18535507 | 9.188 | 12.422 | 21.614 | 42.51% |
rsync-builder | 18535505 | 8.631 | 7.227 | 15.862 | 54.41% |
Notes:
- Profit Margin is calculated with the following formula:
100 * (profit / (sent_to_builder + miner_rewards))
- Builders are identified where possible with pubkey tags and
extra_data
as a fallback - Query for generating this results can be found here: BigQuery stored query
Conclusions
- Five builders dominated the market
- Adjusted profit methodology is needed that goes beyond simple balance changes on the builder’s coinbase address
- Currently available dashboards do not accurately display profits for some builders
- Impostor builders can also affect calculations, use of pubkey mapping table helps avoid this
Next Steps
There are a few improvements that could make this report better and more detailed:
- Consideration of integrated Builder-Searcher profit
- Estimation of off-chain profits (as part of cross-domain activity)
- Sources of block profit (exclusive/private searchers, MEV-Share/MEV Blocker, bundle analysis, etc.)
Appendix
Methodology For Calculating Block Builder Profit
Block builder receivables
Block builders on Ethereum have two ways of accumulating block value (from where they extract some profit):
- Priority fees
- Transfers to block coinbase address
Priority fees are calculated using this formula:
(receipt_effective_gas_price - base_fee_per_gas) * receipt_gas_used
Transfers to coinbase are calculated with data from the traces and transactions table.
We sum all of the transfers to builder address (usually coinbase, with some exceptions)
Block builder outgoing transfers
A block builder has one main outgoing transfer on its blocks, mev-boost payment to proposer_fee_recipient
(address of the validator, chosen for proposing the block).
To capture that transaction and all other outgoing transfers from the builder are summed and grouped by block_number
.
proposer_fee_receipent
is sourced from eden-data-public.mev_boost.payloads
To just find the value of transfers not being transferred to proposer_fee_receipent (important for finding adjustments in the logic for calculating profit),recipient address is used as filter and all other trace values are summed.
Methodology for Adjustments
Account Linked to Builder Profit
If the builder decides to cash out part of its profit in their own blocks, this results in underestimation of their total profit as this transfer is categorized as outgoing transfer. To avoid this, the Account Linked to Builder Profit
must be omitted in the sum of outgoing transfers.
Based on the timeframe of this research (November 2023) only rsync-builder used an Account Linked to Builder Profit
:
0x6d12cfdf3929eeabe0a73e4674d43864388e6ee1
In order to correctly calculate profits and include the profits that were sent to linked addresses, we added an adjustment to the methodology, where all transfers to this address are considered in the adjusted profit.
This resulted in a more accurate profit calculation for rsync-builder and the logic can also be reused for other builders if they decide to use Accounts Linked to Builder Profit
for cashing out their profits.
Account Linked to Builder Costs
Block builders must deliver the promised value of the block to the proposer fee recipient (in other words, pay their costs), but there is no rule that the reward must come from the block coinbase/miner address.
Based on the timeframe of this research (November 2023) only rsync-builder used Account Linked to Builder Costs
:
0xb4e29628cacea3567cf00ca0721234bd37fa0c10
High-fee Transactions from Account Linked to Builder
If an account linked to the builder is sending transactions with unusually high priority fees, the simple profit calculation considers those fees receivable.
Based on the timeframe of this research (November 2023) only jetbldr used High-fee Transactions from Account Linked to Builder
:
0xf0d96e4648bfe7322c146bf77f0b8d043e9f2a6a
We found that this adjustment considerably changes the perception of the block builder as profitable.
Withdrawals Considered in Block Value
Some builders use the proposer fee recipient as the block coinbase address rather than using their own distinct coinbase address. As a result, withdrawals to the proposer fee recipient counts towards MEV-Boost value.
Because there is no way to collect profit without a transfer from the coinbase address (which is owned by the proposer), the simple profit methodology overestimates gross revenues generated by the builder.
Based on the timeframe of this research (November 2023) only PenguinBuild was affected by this adjustment.
Impostor Builders
Three Impostor Builders
have been identified in the dataset, all impersonating Flashbots builders. What makes them special is that they use the Flashbots coinbase address, but send MEV-Boost payments to the proposer fee recipient from an address not associated with Flashbots.
As Impostor Builders
do not profit from these blocks, but rather pay for from their own funds, profits have been subtracted as part of the adjusted methodology.
Based on the timeframe of this research (November 2023) , here is the list of Impostor Builders
pubkeys:
0xa95b3a3cfc35a77663d6a5a9ac133bf1b68b4118f7f7a6f4ec43b298211441d1ebd1a1063446fea18138e7ef6c1379b6
0xb61a17407826a0c7a20ce8a0e9c848350bb94bf258be9c40da0dafd5be83be240c3d24c901e1d4423cc2eb90703ff0bc
0xa003117a3befd6d4f4f5a6db633caf7a2038d3f195c97a6b83ce6760cbbb1c0d09c11c33286fb14eb64c33ffb47f93cf
Data Sources
Data sources used in this research:
eden-data-public.ethereum_auxiliary.tags_pubkey
eden-data-public.mev_boost.payloads
bigquery-public-data.crypto_ethereum.transactions
bigquery-public-data.crypto_ethereum.blocks
bigquery-public-data.crypto_ethereum.traces