From Public To Private Transactions In Ethereum: The Flippening

Over the past year, Ethereum has seen a dramatic rise in private transaction order flow. New analysis using data from the Blocknative Data Archive shows that private transactions now consume more than 50% of all Ethereum L1 block space when measured by gas used. 

This "flippening" marks a crucial turning point in how we understand and analyze Ethereum's network dynamics. The associated loss in observability into demand signal impacts every on-chain actor in the ecosystem, including end-users, wallets, application developers, L2 providers, infrastructure operators, and more. 

A Clearer Measure of Private Transaction Activity

Historically, the ecosystem has measured private activity via confirmed transaction count. However, a more objective measure for network activity is gas used. 

Viewing the private transaction market through the lens of gas used reveals new insights that transaction count alone misses. Although approximately 30% of all transactions are private, today these account for more than 50% of all Ethereum L1 gas used. Why? Because users typically choose to transmit transactions privately for MEV protection, particularly when conducting more complex – and hence gas-intensive – on-chain actions such as swaps. These, in turn, consume more gas per transaction than non-MEV transactions. 

Gas Used: A More Accurate Measure of Economic Value

By shifting focus to the amount of gas used by private transactions, we can develop a more accurate understanding of the current network dynamics. In particular: 

1. Base Fee: The gas used in the current block determines the Base Fee for the next block, a critical UX component for dapps and their end users.
2. Resource Utilization: Gas consumption reflects the computational resources required to process transactions. A single complex transaction can use significantly more resources than multiple simple transactions.
3. Private Order Flow Disparity: Despite ongoing outreach efforts, not all Ethereum users know how to transmit a transaction privately, why they might want to do so, or what tools and applications they can use to enable this. This disparity has significant implications for user experience, network participation, fairness, and more.
4. Economic Value: Finally, gas used directly correlates with the economic value of block space. Each unit of gas represents a fraction of the block's capacity and, by extension, its economic worth.

The Flippening: A New Milestone Reached

Private transactions now account for over 50% of gas used in Ethereum. This flippening represents a fundamental shift in how the Ethereum network is being utilized – and by whom.

Key implications include:

• Increasing adoption of private transactions in the Ethereum ecosystem.
• New centralization vectors as private transaction order flow is accessible only to permissioned network participants.
• The opportunity to reevaluate network economics and the use of gas price optimization mechanisms.

Base Fee Volatility and Private Gas Usage

This flippening carries with it a noteworthy correlation: Base Fee volatility appears to be linearly correlated to private gas used. 

The Base Fee, central to the August 2021 London hard fork via EIP-1559, dynamically adapts to block space utilization. As private transactions increase and consume more gas, they interact with altruistic self-built blocks (also referred to as vanilla blocks) and act as a new source of base fee volatility, as outlined in our previous research post. More private transactions consuming more block space result in self-built blocks that are less full and thus exacerbate the base fee whiplash effect. 

However, this phenomenon is not only tied to private transactions' impact on self-built blocks, but also to the growing disparity of private order flow among different block builders. The graph below shows the trend in private gas used by block builders since the Dencun hard fork in March of 2024.

Here we can see the Private Gas Used by the top builders (Beaver, Titan, Rsync and Flashbots) versus all other builders (green). Other builders have grown their private order flow moderately from about 1.5MM private gas used to about 2.5MM, a ~70% increase. While the top builders have shown dramatic increases in private order flow since March:

• Titan has increased from ~3.5MM to 8.5MM private gas used (~140% increase)
• Beaver has increased from ~3MM to ~7.5MM private gas used (~150% increase)
• Flashbots has increased from ~3MM to 7MM private gas used (~130% increase)
• Rsync has increased from ~2.5MM to ~6MM private gas used (~140% increase)

Note that, even amongst these top four builders, there is a growing disparity in access to private order flow as Beaver and Titan consistently fill over 50% of their blocks with private gas used while Rsync and Flashbots are less than 50% private gas used. This impacts overall block space utilization – which affects the Base Fee – and hence every Ethereum user and application. 

The chart below shows that the top 4 builders typically have blocks that are over 50% full. Keep in mind that EIP-1559 targets 50% full blocks at 15MM gas used.

This suggests that the Rsync block builder has declined below the 50% threshold on average, while the Flashbots builder was declining but recently has recovered. Over June and July 2024:

• Rsync blocks declined from about 18MM gas used on average to 14.5MM gas used, a 20% decline.
• Flashbots blocks declined from about 17.5MM gas used on average to 15.5MM gas used, a ~10% decline, but appears to be growing once again. 
• Titan blocks have seen some variability, but have remained above the 15MM target.
• Beaver blocks have increased from about 15MM gas used on average to ~15.5MM gas used.

Perhaps most significantly we see that the “All Other Builders” blocks have declined from 12.5MM gas used on average to slightly below 10MM gas used, a 20% decline. The target block space of 15MM gas used now appears to be out of reach not just for self-built blocks but also for any of the ‘challenger’ builders that are not already in the top four. 

So does the Ethereum Base Fee now reflect the current spot demand for block space? Or does it increasingly reflect the intricacies of the block building market and private order flows? The current rate of >50% of all gas consumed originating with privately transmitted transactions suggests the answer is increasingly ‘both’. 

From a consumer perspective, this impacts both the UI and UX of Ethereum L1. There is reduced observability in the gas fees to get on-chain when the majority of block space is consumed by transactions that are sent privately. This means an increase in the likelihood that you will either price your transaction too low and end up with a stuck transaction or price your transaction too high to ensure your transaction lands on-chain. Additionally, the spikes in base fee due to the disparity between MEV-Boost block builders and self-built vanilla blocks already introduce increased volatility in gas pricing for users and builders alike.

The Shifting Landscape of Transaction Originators

Private order flow is a main centralizing force within the block building market. The more private order flow any given builder has, the more that builder will win a MEV-Boost block auction, and the more that builder attracts new private order flow. 

However, the transaction originator landscape is shifting at the same time: away from retail users towards centralized actors who can have an outsized impact on the block building market. In an EOA-only world, end-users interact with a dapp, generate a transaction, sign the transaction with their wallet, and then decide (via the wallet) if the transaction will be broadcast publicly or privately. But in today’s market, there are two new actors that may break up the wallet order flow monopoly: Telegram bots and intent protocols.

Telegram bots, such as the popular Banana Gun, allow users to conduct on-chain activities via the popular messaging app Telegram. However, the Telegram bot itself holds the private key – rather than the user – and therefore determines how to broadcast the transaction. This means that Telegram bots not only have full discretion on whether to send transactions publicly or privately, but also on which block builders to share that private transaction order flow. Given the significant private transaction volume that flows through Telegram bots today, they exert a significant influence on the block builder market. 

At the same time, many dapps are moving to intents: instead of signing a full transaction, the user signs an intent and hands it back to the dapp to transform into an on-chain transaction. The dapp works with a set of actors, typically known as solvers, to turn the signed intent into a signed transaction that can then be broadcast publicly or privately. In some dapps, the operator of the dapp itself is the entity that broadcasts the final transaction while other dapps allow the solvers to broadcast the final transaction. Intent-based dapps shift order flow away from wallets to either solvers or the dapp itself.  And the dapp or solver, like the Telegram bots, has full discretion on which block builders do or do not receive these transactions.

The dynamics we see playing out in the builder market are increasingly driven by the decisions of a smaller subset of actors – such as Telegram bots and intent-based dapps – instead of just the decisions of individual users via their wallet.

This raises several important questions:

1. What are the long-term implications for Ethereum's fee market if these trends continue?
2. Given that specific actors now have more information and control over private transaction flow and their relative fees than others, how might this information asymmetry give rise to additional centralizing forces?
3. If private transaction order flows increase during periods of high network activity, will this further exacerbate fee volatility?
4. How does the increased usage of private transactions impact overall network congestion?

Conclusion

By focusing on gas used rather than transaction count, we as an ecosystem can gain a clearer picture of Ethereum’s economic realities and resource allocation. As we move forward, we believe it is important for researchers, developers, and L2’s to consider these realities when analyzing network health and designing scaling solutions. 

The ‘flippening’ of private gas usage is not just a statistic: it is a call to action for the Ethereum community to ensure the network remains open, fair, and accessible to all users. If you and your team could use help navigating the increasingly complex world of gas, reach out – our team at Blocknative is building the tooling and infrastructure you need so that your users do not get lost or left behind.