PBS in Danger? Our Latest MEV Supply Chain Data Tell You Why

More validators start to do block building.

Jul 18, 2023

One grave concern emerging from our research on the MEV supply chain: an increasing number of validators choose to construct their own MEV block, which is indicated by the bigger share of their MEV incomes generated from non-MEV-Boost sources. It's worth considering the implication for the future of PBS.

We calculated the data ranging from June 11 to July 11. Out of the $16.43M revenue for validators, $7.14M comes from MEV generated by non-MEV-Boost sources, showing as No Direct Transfer EOB in the chart below, accounting for 45%.

Our previous study on the data of Jan and Feb of 2023 painted a different picture: 12.2%, $4.21M, of the total $34.51M came from MEV generated by non-MEV-Boost sources.

It's worth noting that $4.21M was generated in 2 months, while this time, validators reaped $7.14M in just one month. We will share more data from this time's research later. However, this major red flag is worth the community's attention.

How do we define and calculate “non-MEV-Boost” data?

According to bertcmiller, the convention of MEV-Boost about “fee recipient” is:

look at who the last tx in the block is paying to figure out who the validator actually is.

So we identify MEV-Boost-related blocks using two conditions:

The block builder is the fee recipient address.
Validators accept the transfer at the end of the block.

However, we have recently noticed some new trends and classified them as below:

Type 1: The builder sets the fee recipient's address as the validator's address and leaves a note in the block's extra data.
Type 2: The builder does not pay the validator fee through the last transaction in the block, and there's no extra data in the block indicating the builder directly sets the fee recipient address as the validator's address.

These 2 types come from 2 situations:

Situation 1: The fee recipient is set as a proposer/validator's address by the block builder. In this situation, all the block's income(including priority fee and builder payment) belongs to the proposer/validator's address.
Situation 2: The block is built by the validator, which certainly is a non-MEV-Boost scenario. (Btw, we just met a big validator who is doing block-building during ETHCC. )

Here is the full chart.

In situation 1 for type 1, we can determine whether the fee recipient has been changed by the builder through the extracted data provided by the builder. We have identified approximately 44,525 blocks where the fee recipient address was modified. Those blocks contribute to the validator's income of $6.4M out of $7.14M. This s a significant change in the revenue of builders who directly set the validator address as the fee recipient. This portion of revenue directly transferred to validators has increased from $2.67M in 2 months to about $6.4M in just one month. Further research is required to understand the relationship between builders and validators and the underlying factors driving this trend.

This indicates the issue of PBS: the builder and proposer are not separated since there is a strong suspicion that builders and validators have close relationships.

Regarding type 2 and situation 2, comparing the latest data from January and February with June, we can observe that there is no significant change in the revenue of validators who directly participate in the MEV market as builders. (considering that the calculation period for the new data, $0.74M, is only covering 30 days, multiplying by 2 will yield a result close to the previous data of $1.54M for Jan and Feb).

We will share more specific data regarding MEV-Boost/Direct Transfer EOB, and the other 2 types that are No Direct Transfer EOBs.

Our data scientist Dr. Sophie Liu gave a talk during the ongoing EthCC. Here is the video:

You can find out the slides used in the video, along with more about how to use EigenTx to parse and visualize transactions: bit.ly/eigentx

Wisdom of DeFi by EigenPhi

Discussion about this post