Ethereum: Brief History of MEV

Subscribe to our NewsletterStake ETH with Confidence

Pre-Flashbots

In the early days of MEV, extraction bots were created to find profitable opportunities and then would compete with each other for blockspace via Priority Gas Auctions (PGAs) - bots competitively bidding up transaction fees in order to obtain priority ordering. The results were: periods of higher gas prices, bloated pending transaction pool, and wasted gas - attempted transactions still needed to pay some gas for operations (for more see here).

In this example of an observed bot PGA, notice the number of bids between two bots over the course of fourteen seconds (85 total bids):

(Flash Boys 2.0: Frontrunning, Transaction Reordering, and Consensus Instability in Decentralized Exchanges by Daian et al) 

“Maximal extractable value (MEV) refers to the maximum value that can be extracted from block production in excess of the standard block rewards and gas fees by including, excluding, and changing the order of transactions in a block” - Ethereum

Post-Flashbots (pre-merge)

Flashbots was created to mitigate some of the negative effects of MEV; as noted on their website:

“Our primary focus is to enable a permissionless, transparent, and fair ecosystem for MEV extraction. It falls under three goals: Democratizing Access to MEV Revenue, Bringing Transparency to MEV Activity, and Redistributing MEV Revenue.”

Beyond these stated goals, by creating a marketplace that connected searchers (i.e., MEV extracting bots) and miners, Flashbots was able to eliminate most of the problems mentioned above, namely a bloated pending transaction pool and wasted gas.

Flashbots became a significant player on Ethereum, as shown here in the percentage of bundles that landed on chain - around 50%: 

Lastly, here is a look at miner rewards related to MEV pre-merge, giving an indication of the average reward miners could receive from including bundles in a block:

Source

Lastly, here is a look at miner rewards related to MEV pre-merge, giving an indication of the average reward miners could receive from including bundles in a block: 

Source

MEV Today - Post Merge

The post-merge architecture created by Flashbots represents another leap forward:

Source

MEV Boost is referred to as a sidecar - a piece of software run in addition to the typical Ethereum validator stack. This is compared to pre-merge - when miners ran a forked version of geth called geth-mev; in other words, geth-mev would replace a miner’s Eth1 node.

Two immediate benefits of running MEV through a sidecar architecture are liveness and client diversity. From the liveness perspective - it is now possible for validators to revert to local production in the event that MEV Boost is not available. On the client diversity front, MEV Boost is not client specific - validators can choose one of the major consensus clients and run MEV Boost as well, unlike pre-merge when miners ran a fork of geth.

Another big benefit is that anyone from the proposer side (i.e., validators) can now participate, as opposed to pre- merge, where only trusted miners were able to participate in the Flashbots’ marketplace. This was a requirement because, pre-merge, miners could view the transaction bundles they received and potentially “steal” the transactions (for more see here). Post-merge, validators are not able to view the transaction payloads (i.e., the actual transactions), until after they have signed the block header.

Mev-Boost Proposal Block

Source

The process is shown above; read from top to bottom and follow the arrows; specifically, look for getHeader and then later getPayload.

The result is that validators do not need to be trusted because they do not have access to the transactions before they sign the block header and, so, cannot “steal” the transactions. In other words, this allows any validator to run MEV Boost and greatly levels the playing field.

One interesting aspect to The Merge is not the comparison of pre-merge to post-merge but the actual period of transition. Blockspace was relatively abundant for searchers pre-merge, with around 85% of network hashrate using geth-mev. However, at the merge there was roughly 0% of blockspace available to searchers. That number has climbed to around half today (see below), but the result was that in some cases searchers had to fall back to PGAs again, that is, open competition for blockspace in the mempool.

Post-merge MEV by the numbers

So far Flashbots is the dominant relay:

Source

There is some dispersion in average block value, but the numbers have converged recently. That said, registrations have slowed down in the last few days:

Source

On a network basis, 47% of validators have registered with the Flashbots relay (see here).

Below is another helpful graphic showing adoption of Mev Boost from Elias Simos at Rated:

Source

Notice in the above chart that around 50% of blocks are coming from MEV Boost.

On a performance basis, there is a significant difference in performance between those running MEV Boost and those not:

Source

As can be seen, running MEV Boost increases execution rewards between 1.7 and 6.1 times. 

In terms of APR (note MEV Boost rewards are captured in ‘EL APR%’):

Source

A few points are worth bearing in mind. First, it is still early days; although the numbers above give a sense of rewards, it is hard to say what longer-term rewards will look like. Additionally, in connecting the numbers from the first chart to the second, it is important to recognize that MEV rewards are only relevant when proposing a block - something that happens relatively infrequently.

It is also very important to pay attention to the column: “mev-boost % of proposed.” This column shows the percentage of blocks that came from MEV Boost. Note the relationship between EL APR% (that is, execution layer APR%) and the “mev-boost % of proposed” - as the percentage of blocks proposed from MEV Boost increases, so too, generally, does the EL APR%.

Figment

Figment was an early adopter of MEV Boost and has added about 771 ETH between September 15th and October 9th.

As can be seen in the chart below, there was a noticeable increase in rewards following The Merge.

Priority fees explain the network-wide increase as all validators began to receive priority fees. Figment’s outperformance is explained, in part, by its early adoption of MEV Boost, contributing to its roughly 12% gain vs the average Ethereum validator.

It should be noted that The Merge happened mid-month, so the pickup in rewards network-wide will be more pronounced for a full month. Also, the data for the month of October is up to October 13th.

Due to the structure of Ethereum, rewards optimization opportunities are rare. This is why engaging in MEV Boost is especially important. For more on Figment’s MEV policy see here.

To read more about Figment’s performance on Ethereum see here.

Stay in Touch

Subscribe to receive Figment and Web 3 ecosystem updates.
Get Updates
Light blue dots