From the perspective of MEV and value capture, why will Uniswap eventually turn to UNIchain?

Written by: Nascent co-founder Dan Elitzer

Compiled by: The Way of DeFi

In recent months, there has been growing interest in application-specific blockchains (or rollups), which we often refer to as application chains. The basic idea here is to have an independent blockchain that has its own validators/aggregators, focusing on specific protocols and applications.

This is different from the philosophy of general-purpose blockchains that have dominated the market so far, from Ethereum to BSC, from Polygon to Solana, not to mention general-purpose L2s like Optimism, Arbitrum, or Starknet. It has long been believed that the "interoperability" and "composability" gained by having a public platform for storing and updating state is crucial.

However, the view that "various applications ultimately want their own chains" is not new. For years, teams dedicated to Cosmos, Polkadot, and Avalanche have been working towards this future vision. Compound took a step in this direction in 2020 when they announced the Compound chain built on Polkadot's Substrate, but ultimately abandoned the effort when they realized the infrastructure was not mature enough (at least for now).

Earlier this year, dYdX announced it would leverage the Cosmos SDK to migrate to its own application chain, reigniting interest in application chains. Recently, the Delphi Labs team shared their incredible comprehensive analysis on how they view the evolution of the crypto and DeFi ecosystem, ultimately focusing on Cosmos as their platform of choice.

At Nascent, over the past two years, we have increasingly believed that some form of application chain-centric (or application rollup-centric) world is a possible state for the future of the crypto ecosystem.

I will not attempt a comprehensive analysis of why and how this will happen, but instead provide a single case study that I hope will help builders who have not yet considered application chains to understand why they ultimately must consider them.

As the dominant DEX with a strong brand, Uniswap is a major candidate for eventually migrating to its own application chain or rollup. Among all contracts/applications, Uniswap has the strongest capability and the greatest motivation to transition to its own chain.

In the past year, Uniswap V3 had a trading volume of just under $700 billion on Ethereum L1, with an average trade size of about $17,870 and over 38.8 million trades. What costs do traders incur when making these trades?

Cost 1: Swap Trading Fees

The first and most obvious fee is the swap trading fee. Most centralized exchanges charge fees to both makers (those who place quotes on the order book) and takers (those who accept quotes posted on the order book). Since exchanges want to encourage market makers to provide deep orders, the swap fees paid by makers are typically lower than those paid by takers, and some high-volume traders can even receive rebates.

In the case of Uniswap, liquidity providers (LPs) act as makers, while traders act as takers. Historically, traders' swap fees have been 0.30% (30 basis points), with all these fees going to LPs, resulting in LPs receiving -0.30% (negative 30 basis points) in fees. In UniV3, different pools can have different fees, ranging from 0.01% to 1.00%, but they all share the same dynamic, where LPs receive 100% of the fees paid by traders. Therefore, in the past year, traders paid nearly $1.2 billion in swap trading fees, averaging 0.171% per trade, all of which went to LPs.

That's a significant amount of fees! Compared to the fees paid by retail traders on centralized exchanges, this is quite reasonable.

Unfortunately for traders, this is not their biggest cost when trading on Uniswap.

Cost 2: Transaction Gas Fees

So where are these Uniswap trades settled? Of course, on Ethereum. Each transaction has associated gas costs, which are paid to Ethereum validators (directly) and ETH holders (indirectly). Note that this is not a fee that traders on centralized exchanges must pay. Therefore, in the pursuit of becoming the best trading venue in the world, Uniswap's requirement for its traders to pay a fixed (though volatile) fee for each transaction is far from ideal.

Like gasoline in the real world, gas on Ethereum is not cheap: Uniswap traders spent $1.63 billion on gas fees in the past year. This adds an additional 0.235% (23.5 basis points) to each transaction.

This means that, on average, traders pay 17.1 basis points to LPs and 23.5 basis points to Ethereum validators. The fees paid to validators are 37% higher than those paid to LPs! Clearly, a significant amount of value is leaking.

Cost 3: MEV

Earlier this year, 0x Labs obtained data on 700,000 trades entering AMMs through the 0x API. They found that, on average, the slippage of these trades exceeded 20 basis points compared to the initial quotes… and slippage increased with trade size! Overall, in that month (June), MEV losses through the 0x API reached $27 million, prompting developers to launch a new slippage protection feature to address this issue.

Due to the lack of more direct data on Uniswap-related MEV, we can make a very rough estimate of the MEV captured on Uniswap over the past year based on the following assumptions:

• The MEV impact on trades conducted on Uniswap is roughly the same as that on trades conducted through the old 0x API;

• 70% of AMM 0x API trade flow goes to Uniswap V3 (roughly corresponding to UniV3's AMM market share);

• The proportion of MEV in trading volume in June is consistent with the rest of the year;

• 16% of Uniswap's trading volume comes from the 0x API;

Of course, these assumptions are almost certainly wrong, but based on estimates, the MEV cost for Uniswap traders over the past year is about $1.76 billion. The average cost per trade increases by another 25.4 basis points. Adding this to the swap trading fees and gas trading fees results in a total cost of 66 basis points per trade.

How should UNI holders try to capture value?

If traders on Uniswap are paying an average of 66 basis points in fees, and Uniswap is poised to become the market leader among DEXs, then UNI holders should have sufficient capacity to capture a portion of that.

Currently, the only way for UNI holders to capture value is to use governance to activate the fee switch and take a share of the swap fees currently flowing to LPs. If the fee switch were activated and set to the level used in UniV2, it would have generated about $193 million (approximately 16.67%) in revenue for UNI holders over the past year.

Of course, these potential profits do not take into account the negative impact of activating the fee switch on LP profitability, which would almost certainly lead to reduced liquidity and trading volume, thereby decreasing fee revenue. This is the core issue with using swap fees as a source of value capture for UNI holders: it introduces a spread in the market, inevitably making it a less efficient trading venue.

What about the other two costs, gas trading fees and MEV? As long as Uniswap continues to rely on Ethereum or other general-purpose chains and rollups, UNI holders currently have no good way to leverage either.

Enter UNIchain.

If Uniswap were to migrate most of its activity to its own chain, where UNI holders are validators, it would immediately provide a powerful option for lowering trading costs and capturing value.

First, rather than paying a large fixed gas fee to Ethereum validators, these costs could be reduced by an order of magnitude or more, immediately lowering the costs of trades (especially smaller trades) significantly, thus improving overall trading efficiency. Surprisingly, Uniswap has so far succeeded, with traders spending more on fixed gas fees than on volume-based swap fees. Making fixed gas fees a small part of trading costs would be a huge win for traders, and allowing UNI holders rather than ETH holders to capture these fees would be a significant improvement over the existing "worthless governance token" model of UNI.

Second, controlling the validators/aggregators means that UNI holders would be able to take measures to minimize traders' MEV costs, such as implementing threshold encryption algorithms (like Osmosis) or batch swaps to execute all trades between assets in a block at the same price (like CoW Protocol). Reducing MEV would make Uniswap a more efficient trading venue, but in cases where MEV cannot be completely eliminated, at least a Flashbots auction-style system would ensure that most MEV can be internalized to UNI holders/validators rather than given to ETH holders/validators. This ultimately resembles a form of payment for order flow (PFOF).

While the best value capture model for UNI resembling the idea of PFOF may sound traditional, it may be inevitable, and at least in this form, access to it is democratized and transparent.

It is also important to note that not all MEV is harmful to traders. While it is crucial to minimize forms of MEV that directly negatively impact trader prices, such as front-running and sandwich attacks, other forms of MEV (such as back-running and arbitrage across pools or venues) are harmless to traders and represent a huge potential source of profit. How large is the volume (and profit) of MEV trading on Uniswap?

About two-thirds of total trading volume.

Similarly, since most MEV experts agree that MEV can only be minimized, not completely eliminated, it is better to allow MEV to occur on-chain or in a rollup where the incentives for validators align with minimizing the harmful types of MEV for Uniswap traders. If back-running and arbitrage-based MEV can be internalized, it could significantly increase value for traders, LPs, and UNI holders.

On general-purpose chains/rollups, a wide range of use cases and stakeholders must be considered when attempting to address MEV.

However, on UNIchain, solutions can be tailored specifically to the needs of traders and LPs. For inspiration, take a look at the recent proposal by Skip Protocol to convert MEV revenue into Osmosis protocol revenue.

Assumptions of Application Chain Theory

When I present this theory to those who have not yet embraced application chains, the main obstacle I encounter is the belief that migrating to an application chain would undermine composability. (Trust me, I am well aware of the benefits of composability!) My response is that we have entered a radically multi-chain world, where much of the economic activity in crypto will clearly not all happen on a single chain.

This means we need to do better in terms of trustless and minimally trusted cross-chain bridges, just as we need (and still need) to write secure contracts better. Compared to any traditional financial rails or databases, almost any two or more public, permissionless blockchains have greater potential composability. Application chains may not provide the same level of atomic composability that we have become accustomed to in the early days of DeFi, but as the ecosystem matures, we will find the benefits of user experience and security when making more explicit design decisions about the types of composability to prioritize.

Another driving factor for an application chain-centric future is that many applications need more direct control over their costs and access to block space. While creators of Solana and various other high-throughput blockchains have managed to achieve impressive capacity growth, ultimately, there are physical limits to the volume of transactions that can be processed on any decentralized network that values meaningful decentralization. If you, like me, are optimistic about the future of our industry, you must believe that if costs are low enough, the demand to write data to a globally accessible and censorship-resistant database is, in fact, infinite. Therefore, some form of market pricing for block space is needed.

On a permissionless general-purpose chain, this creates a problem: each application has slightly different limits on the transaction fees its users are willing to accept; if you make block space too cheap, some will always do things like put entire games on-chain, consuming all the space until they push costs above what other application users are willing to pay.

Thus, once an application reaches a certain scale, it becomes increasingly important to have more direct control over its access and the costs of users accessing block space. The solution to this problem is to create a chain or rollup where access to block space is managed by validators, whose primary focus is the success of a single application (their own application).

Conclusion

To recap:

1. Currently, traders using Uniswap need to pay three types of fees:

   (1) Swap trading fees (variable, paid to LPs)

   (2) Gas trading fees (fixed, paid to ETH validators)

   (3) MEV (variable, paid to ETH validators and market makers)

2. Swap trading fees are currently a small part of the costs traders pay and the only place UNI can potentially capture value;

3. Transitioning to an application-specific rollup, where UNI is staked by validators (i.e., UNIchain), will reduce gas trading fees and MEV;

4. Unavoidable MEV will largely be internalized by UNI holders, which may be the best place for UNI to capture value in the long term;

Any application that grows to a sufficient size may have similar reasons to consider migrating its main application to its own application chain or rollup at some point.

It is clear that while we (Nascent) believe application chains will ultimately become a major force in the crypto ecosystem, we think the timeline for this to become a reality will be more than three years away. For application chains to become a major force in the crypto ecosystem, significant progress needs to be made in development tools, validator infrastructure, rollups, cross-chain bridges, wallets, and more.

This does not mean it is not worth considering the application chain path early on, and Nascent has already invested in multiple teams/projects based on this theory, including Osmosis, SX Network, and Ribbon, as well as their upcoming options exchange Aevo.

[1] Technically, the correct way to simulate this "swap fee" is not as a fee but as a 60 basis point bid-ask spread, as it is paid to LPs rather than a third party. However, since Uniswap has always treated this as a fee, even accumulating it separately in V3, it seems fair to frame it this way.

[2] In fact, this could be an Ethereum-based rollup where validator/aggregator rights are managed through staked UNI; it is hard to imagine the Uniswap Labs team choosing a Cosmos zone or other non-rollup alternative to completely exit the Ethereum ecosystem.

A special thanks to Will Price, Elle, Matteo Leibowitz, Josh Felker, and Brock Elmore for their valuable feedback and editorial work.