Solidly Deep Dive Pt.2: Concentrated Liquidity and ve(3,3) vs. Uniswap
Preface:
In the first article of this series, we reviewed the Solidly ve(3,3) DEX flywheel, economics, and sustainability.
In this article, we will explore how the ve(3,3) exchange fits into the existing DEX landscape, and specifically why I believe the ve(3,3) model will dominate over Uniswap, with a particular focus on the much discussed concentrated liquidity pools.
My core thesis is that ve(3,3) exchanges are better positioned to be core DeFi infrastructure than Uniswap.
TLDR:
- Concentrated liquidity V3 pool challenges for long-tail assets create a market for traditional vAMM and sAMM pools, which the ve(3,3) model optimizes through incentive alignment and a bribe market
- Uniswap V3 liquidity provider (LP) competition challenges can be overcome with managed concentrated liquidity (CL) pools on ve(3,3) exchanges, allowing for more profitable LPs and overall total value locked (TVL)
- greater impermanent loss in Uniswap V3 pools can be offset by increased yields to LPs in the form of token incentives by ve(3,3) model
- with competition challenges and impermanent loss solved, ve(3,3) concentrated liquidity can gather more TVL and become the dominant source of liquidity and pricing for blue chip assets
- Uniswap has an existential issue with their fee switch which is overcome by ve(3,3) tokenomics, allowing for protocol sustainability
- ve(3,3) model is more resilient to competition risk than Uniswap
- each chain and ecosystem will have 1 dominant ve(3,3) exchange that will serve as a core infrastructure and liquidity layer.
- anything Uniswap can do, ve(3,3) can do better because the real innovation is the incentives alignment
I. Background
Uniswap V3 concentrated liquidity pools have been around for roughly 2 years now, and have accumulated a massive 2.83B in TVL.
We now have plenty of data to analyze concentrated liquidity efficacy as both an AMM operating through various (extreme) market conditions, and also as a source of yield of liquidity providers.
At the same time, the ve(3,3) Solidly model has also gained traction since the start of 2023, and has now amassed over $500M in TVL.
Interestingly the adoption of ve(3,3) has mapped out almost exactly according to “The Startup Curve”, with an initial enthusiasm around the original Andre Cronje implementation and then spending much of 2022 in the “Trough of Sorrow”. In 2023 the model has found product market fit and is currently scaling quickly.
As of April 1, 2023, Uniswap V3 concentrated liquidity AMM (clMM) Business Source License has expired and it is now open source. There is currently an arms race in the alternative DEX protocol space to fork and implement this technology.
ve(3,3) exchanges are no exception to this, and are also launching their own concentrated liquidity implementations. Most notably Thena with their “Fusion” managed liquidity product is set to launch imminently.
Now is a good moment to evaluate not only ve(3,3)s place in the market in competition with Uniswap, but also concentrated liquidity pools altogether.
II. Concentrated Liquidity Pool Mechanics
Uniswap V3 introduced the concept of concentrated liquidity, which allows liquidity providers (LPs) to allocate their capital within specific price ranges.
In Uniswap V2 pools, liquidity is scattered evenly along the invariant curve of xy=k. However, most trading activity occurs within a specific range at any given time, leaving liquidity in other parts of the xy=k curve underutilized.
By concentrating liquidity to specific and more active price ranges, LPs can increase capital efficiency and earn higher returns on their liquidity.
However, Uniswap’s v3 concentrated liquidity is not a one size fits all solution because it greatly increases the cost of managing on-chain liquidity and gives up its pricing power. This can lead to Uniswap becoming less competitive in the market compared to ve(3,3), which can utilize it’s pricing power and effective tokenomics to disrupt the incumbent.
III. Loss of Pricing Power:
One of the most significant factors contributing to ve(3,3)s competitive edge over Uniswap V3 is its pricing power.
What is pricing power? For any asset that is traded among several exchanges, only 1 can have pricing power.
An asset can be listed in multiple exchanges, both CEXs and DEXs, however, typically only 1 will have dominance (the one with the most uninformed order flow) in terms of pricing and other exchanges will be arbitraged accordingly.
Uniswap V3 has lost a considerable amount of its pricing power due to several factors:
Malfunctioning in Extreme Market Conditions:
Uniswap V3’s concentrated liquidity model may not function optimally during periods of high market volatility. This is a crucial disadvantage, as a reliable exchange must be able to operate effectively in all market scenarios.
In contrast, ve(3,3) utilizes constant product AMM for volatile pools and stableswap invariant pools for correlated assets that maintain more consistent functionality even in extreme market conditions, ensuring a seamless trading experience.
As an example of this condition, Uniswap V3 could not function properly during the UST crisis as LUNA price kept hitting “no mans land” on the concentrated liquidity ranges, meaning liquidity quickly dried up.
Here are the findings of “Exploring Price Accuracy on Uniswap V3 in Times of Distress” study:
“Our empirical analysis finds that the concentrated liquidity CPMM pioneered by Uniswap V3 is currently not ready to handle unexpected price drops. Not only did we observe significant price inaccuracies and financial loss on the liquidity provider side, but we also see the potential for TWAP attacks on Uniswap V3 oracles. For the success of Uniswap V3, it is thus imperative that liquidity providers will become more sophisticated and agile”
A DEX cannot maintain pricing power if its price information is not reliable in extreme market conditions.
Long Tail Asset Challenges:
There are a number of unique challenges introduced by concentrated liquidity to long-tail assets:
Active management: Uniswap V3’s concentrated liquidity model requires LPs to manage their positions actively, as they must adjust their price ranges to optimize their returns. As the price of new tokens fluctuates greatly, LPs need to adjust the price range very often. This creates a huge capital risk and increases the cost of managing liquidity which may not be affordable or practical.
Shallow initial liquidity: Uniswap V3’s concentrated liquidity model can make it challenging for new tokens with fluctuating price ranges to establish pools, requiring deeper initial liquidity to minimize slippage. Consequently, new projects may be deterred from using Uniswap V3 as their primary source of price information.
As just one example of this challenge, this is Saul's feedback from GMD Protocol:
The threshold of liquidity management with V3 becomes too high for new projects to launch long-tail asset pools. As a result, very, very few projects choose to launch new tokens on Uniswap V3.
Impermanent loss: Despite the potential for higher returns, concentrated liquidity also exposes LPs to greater risks associated with impermanent loss. When the market price of a token pair diverges significantly from the price range specified by an LP, its assets become less effective in facilitating trades, leading to potential losses.
This is a great thread on impermanent loss, specifically as it relates to V3 pools:
Concentrated liquidity acts as “leverage” by concentrating capital in a specific trading range. This means that potential profits and liquidity in that range are amplified, but so are the potential losses when assets trade outside the range.
Because liquidity is lower and volatility is higher in long-tail assets, the impermanent loss is also amplified.
These combined factors mean Uniswap v3 pools lose pricing power on long-tail assets.
Blue-chip Token Dominance:
Because of the above-mentioned challenges to long-tail assets, Uniswap V3 primarily hosts blue-chip tokens with lower price volatility. This further erodes its status as a primary source of price information for a wide range of assets.
ve(3,3) Opportunity:
All of the above limitations make Uniswap V3 less competitive compared to ve(3,3), which caters to both a broader range of low-volatility assets and high-volatility assets.
Because ve(3,3) provides an effective marketplace for new projects to source liquidity through the incentives and bribes market, it becomes the default pricing dominant exchange on these long-tail assets.
When it comes to blue-chip assets, ve(3,3) model is better able to incentivize liquidity providers within both a V2 pool and a concentrated liquidity pool. This means returns to LPs can be sustainable and high enough to offset the negative consequences of concentrated liquidity; more on this in the next section.
IV. Consequences of Losing Pricing Power
Because of all the reasons listed above, Uniswap V3 loses pricing power on the market. This exposes liquidity providers to several negative effects.
Toxic flow exposure:
LPs on DEXes without robust pricing power are exposed to increased toxic flow from arbitrage activities.
This is because the exchange experiences (i) less uninformed order flow (people trade mostly in the primary exchange) and (ii) increased toxic order flow (arbitrageurs take cues from the primary source of pricing information and can take advantage of LPs in the price discovery process of other AMMs).
Because AMM LPs act as the counterparty to trades, they suffer from adverse selection by being on the wrong end of trades with informed traders. Since an AMM does not rely on oracles, arbitrage orders are essential to the AMM price discovery process, but it is at the expense of LPs.
When an exchange loses pricing power, it is subject to a disproportionately high volume of toxic order flow.
According to this analysis by thiccythot (unfortunately the dune dashboard now has errors so we will go off of ,historical numbers):,
Approximately 43% of Uniswap V3’s trading volume comes from Miner Extractable Value (MEV) bots, which can cause consistent losses for LPs due to price discrepancies and arbitrage opportunities.
This has resulted in over $100 million dollars of LP losses in just the ETH/USDC pair AFTER fees:
LPs in AMMs profit when Fees > Impermanent Loss, and according to this analysis LPs on Uniswap V3 have overwhelming lost money.
This is also validated with analysis by CrocSwap that re-created similar calculations to find LP profitability:
Even in the best example of a blue-chip pair (ETH/USDC) which should receive by far the most nontoxic flow, passive LPs consistently lose money.
Another analysis by Frank Maseo shows ETH/USDC pool returns filtering out MEV bots:
Shows a cumulative realized IL of ~$240M for ETH / USDC LPs in Uniswap v3. This condition gets even worse in more volatile and less liquid assets.
Example / Explainer:
I recognize that this might be hard to understand for less experienced readers, so I will include a simple example to demonstrate toxic vs. non-toxic order flow and why that matters for LPs:
Let’s use a xy=k constant product pool that has 10 ETH at a price of $2,000, and evaluate it after ETH price moves to $4,000:
The permanent loss suffered by the LP during this movement is $3,431 (compared to what they would have if they just held the 10ETH and 20,000USDC during the same period).
Now to offset the impermanent loss, the pool must generates trade fees. The LPs are in profit when trade fees > impermanent loss.
In this example, the pool would have had to generate >$1.4M in trade volumes at 0.3% fee during this same period to offset this impermanent loss.
Now if we take two identical pools:
#1. This pool has pricing power and facilitates most uninformed and price discovery swap transactions. (predominantly non-toxic order flow)
#2. This pool has no organic trade volume, and serves only to arbitrage to pool #1. (strictly toxic order flow)
In the process of price discovery, pool #1 will have many swaps both buying and selling ETH the whole time the price moves, and is likely that this volume creates enough fees to offset the impermanent loss and be profitable.
However, pool #2 only receives order flow from arbitrageurs. Because we know arbitrageurs only perform trades when they are in profit net of gas + fees, they will inherently only arbitrage this pool to extract more value than it generates in fees.
As a result pool #2 might only receive ~500k in arb volume during this same time, and the net result is their LPs lose money on the impermanent loss.
Now this example is completely hypothetical, in reality all pools receive a mix of both toxic and non-toxic order flow. On-chain liquidity is dynamic. But there are pools which receive objectively greater toxic flow than others, and it is typically those that do not have pricing power.
Now because concentrated liquidity acts as a leveraged LP position, it also leverages up impermanent losses. So if you are receiving disproportionately high toxic order flow and are leveraged, this can add up to significant losses.
Algorithm-driven trading volume:
According to the Alastor “Uniswap Fee Switch” report: Over 70% of Uniswap V3’s trading volume is driven by algorithms, making it less “sticky” and more susceptible to shifts in pricing power.
Being that the majority of volumes come from algo trading, without maintaining pricing power and status, trade volumes can easily shift away to other primary exchanges. Because trade volumes are the driving force behind LP incentives, this can lead to a death spiral:
LPs less incentivized -> less TVL & liquidity -> higher slippage -> lower trades volume -> lower fees and LPs less incentivized
In contrast, ve(3,3) exchanges maintain its pricing power as the primary source of price information (at least for long tail assets), ensuring a more stable and sustainable ecosystem. As long as pricing power is maintained, it is less susceptible to shifting trade volumes.
Intensified competition among LPs:
Because of all these risk factors (impermanent loss from toxic order flow, increased management threshold, leveraged positioning), Uniswap V3’s concentrated liquidity model has led to increased competition among LPs, further eroding the platform’s pricing power over long-tail assets.
If the price falls outside of a range selected, LPs either (i) realize impermanent loss and select a new range, or (ii) wait until price falls back into the range (not guaranteed).
For (i), impermanent loss accumulates each time the LP rebalance the position, meaning during volatility LPs suffer death by a thousand cuts.
For (ii), LPs suffer lost opportunity cost in ADDITION to impermanent loss. they earn no transaction fees while the price is out of range as capital is idle, and they are force to realize impermanent loss to rebalance into a new range.
This means the threshold for liquidity management is lifted and creates immense competition amongst LPs, where passive LPs are outcompeted by active LPs.
If passive LPs provide liquidity in a greater range, active LPs operating in a tighter range will earn a disproportionately greater amount of trade fees, taking away from the passive LP profits.
Analysis by CrocSwap shows just 48.6% of liquidity providers of ETH/USDC V3 pools are in profit:
With a heavy distribution of positions generating no profits as the distribution is heavily skewed at 0.00.
This proves that while it may be profitable to LP in a concentrated liquidity position, most LPs either do not make money or lose money, and only a very small portion of most competitive and sophisticated LPers are making profits.
This competition not only puts passive LPs in jeopardy but also discourages them from providing liquidity in the first place. If active, professionally managed positions are the only way to be profitable, overall TVL will decrease.
One of Uniswaps main innovations in the AMM model was the democratization of access to liquidity provisioning to the general public, allowing for anyone to participate and increase overall liquidity in markets. With V3 pools, the model has regressed to just professional market makers being profitable, similar to orderbook style exchanges.
ve(3,3) by utilizing a pro-rata model, maintains a more balanced and sustainable competitive environment that fosters collaboration rather than rivalry amongst LPs.
ve(3,3) Solution:
Because the ve(3,3) model changes the incentive structure to LPs to a pro-rata model, it can effectively sustain larger TVL without some the negative consequences of concentrated liquidity.
Firstly, while V2 pools are less capital efficient than V3 pools, the optimized incentive structure of ve(3,3) model can make up for this with increased revenues to LPs that will better incentivize liquidity for long-tail assets in price discovery that may not be suited for a concentrated liquidity pool.
When it comes to concentrated liquidity and V3 pools, the ve(3,3) model through higher incentives to the LPs (token emissions) can help offset impermanent loss they may experience on Uniswap.
Further, managed concentrated liquidity pools can be introduced like Thena has done to benefit from both the capital efficiency and also incentives structure. Because these pools are managed, they take away the competition issues of Uniswap V3 pools.
This allows a greater amount of LPs to participate in concentrated liquidity pools, leading to potentially deeper liquidity.
Concentrated liquidity can be very powerful when both conditions are met:
- Pricing power is established.
- Managed concentrated liquidity solutions that solve for the management and competition threshold of Uniswap V3 pools.
Because ve(3,3) can solve for these issues with Uniswap V3 pools, it is possible that they also become the deepest sources of liquidity for blue chip short-tail assets, and establish pricing power.
In both cases, ve(3,3) is in a better position than Uniswap in incentivizing liquidity for both long-tail and short-tail liquid assets.
V. Profitability, Incentives, Competitive Advantage
A major factor that differentiates ve(3,3) from Uniswap is the unique combination of profitability, sustainability, and tokenomics.
The real innovation in the ve(3,3) model is not the swap technology, it is the incentives alignment between protocols, liquidity providers, traders, and users.
Profitability and Sustainability:
Uniswap has an existential issue when it comes to profitability. While Solidly ve(3,3) exchanges earn 100% of trade revenues, Uniswap earns 0%.
A business with no profits is not a good business. And so Uniswap governance has long considered taking a cut of profits from their LPs.
But this is not such a simple change, and Uniswap could be in serious trouble by doing so. Without pricing power, LPs suffer more from toxic flow, and are less incentivized to provide liquidity. If Uniswap now takes a cut on top of this, it further discourages LPs.
In web2 companies, it is normal for companies to not take profits in the initial years to build moats and a sticky user base, however, this is not the case with Uniswap. As previously mentioned, because Algorithm trading accounts for over 70% of volume, Uniswap does not actually have a “competitive moat” or “sticky volume”.
If you reduce LP incentives through a cut of the fee, this can lead to a death spiral:
LPs less incentivized -> less TVL & liquidity -> higher slippage -> lower trades volume -> lower fees and LPs less incentivized
Now on the other hand, if you raise fees across the board on pools, then you lose trade volume from algo routing (70% of Uniswap Volume) because trades will be less favorable. Again leading to the death spiral as lower trade volume reduces LP incentives.
As mentioned in my first article, the ve(3,3) model has a high number of utility to their DEX token:
- 100% revenue accrual (fees + bribes)
- governance value to direct emissions
- value as an LP pair
- future growth pricing
- speculative trading
- other use cases built on top
In contrast, UNI has no utility at all and is irrelevant to the Uniswap business. It also cannot effectively implement a fee cut.
This is the existential problem Uniswap is in, doomed to be an unprofitable business for the foreseeable future (at least with current products). Whereas ve(3,3) can sustain itself well into the future.
Competition Risk:
In the case of ve(3,3) DEX’s, its native token plays a crucial role in aligning the interests of liquidity providers (LPs), projects, and the platform itself. The ve-tokenomics model enables it to maintain high total value locked (TVL) without relying on leverage.
The ve(3,3) tokenomics, are designed to protect the platform against competition, maintain a balanced distribution of governance power, and incentivize continuous participation in the ecosystem.
In addition to becoming the de-facto market for newer projects, the Solidly wars in which protocols accumulate ownership and voting power in DEX’s provides a competitive moat to protect them from losing this liquidity.
While still early, we can see that the Solidly wars are already taking place with early mover protocols capitalizing on the opportunity to build large positions on these exchanges, such as Deus Finance:
Once these projects are embedded into ve(3,3) exchanges with the accumulation of voting power and bribes, it has an investment into that specific ecosystem. Taking their liquidity elsewhere will represent a loss of their investment.
Whereas with Uniswap, liquidity is transient and has no incentives to stay when higher yields can be realized elsewhere. Further without pricing power and sticky liquidity, trade volumes can free-follow away to other exchanges.
In contrast, the ve(3,3) model incentives are structured to capture new projects through inception (to build liquidity), and once liquidity and voting power is accumulated, the model protects that liquidity from leaving.
Maintaining it’s pricing power position means more non-toxic volume and fees will accumulate to the exchange, further benefiting its flywheel and solidifying it’s exchange dominance.
This also means that the ve(3,3) landscape will manifest differently than Uniswap, who has taken a multi-chain approach.
Due to the distinctive incentives and tokenomics of the Solidly model, efforts are better spent establishing a dominant presence within individual ecosystems (e.g., VELO on Optimism, THENA on BSC, EQUAL on Fantom). Once established, they become difficult to dislodge, but their reach in cross-chain operations may be limited.
Part of my thesis around ve(3,3) exchanges is that each chain and ecosystem will have 1 dominant exchange that will serve as a core infrastructure and liquidity layer.
VI. Conclusion:
In conclusion, I believe that the ve(3,3) model is better position to serve as core DeFi infrastructure than Uniswap.
A market opportunity is opened up because Uniswap V3 pools lose pricing power and are unfavorable to long-tail assets. The ve(3,3) model is ideal in addressing these challenges and market through it’s unique incentives optimizations:
- challenges associated with concentrated liquidity V3 pools for long-tail assets will provide a market for traditional vAMM and sAMM pools which the ve(3,3) model optimizes through incentive alignment and bribe market
- competition challenges for Uniswap V3 LPs can be overcome with managed CL pools on ve(3,3) exchanges, allowing for greater profitable LPs and overall TVL
- greater impermanent loss in Uniswap V3 pools can be offset by increased yields to LPs in the form of token incentives by ve(3,3) model
- with competition challenges and impermanent loss solved, ve(3,3) concentrated liquidity can gather more TVL and become the dominant source of liquidity and pricing for blue chip assets
- Uniswap has an existential issue with their fee switch which is overcome by ve(3,3) tokenomics, allowing for protocol sustainability
- ve(3,3) model is more resilient to competition risk than Uniswap
- each chain and ecosystem will have 1 dominant ve(3,3) exchange that will serve as a core infrastructure and liquidity layer.
- anything Uniswap can do, ve(3,3) can do better because the real innovation is the incentives alignment
As a final thought, when it comes to modified DEX structures such as Curve and Sushiswap, for all the reasons they are an improvement on Uniswap, ve(3,3) goes even a step further.
For example, where Curve provides 50% of trade fees to veCRV voters, ve(3,3) model gives 100%. Where Curve requires Convex to implement efficient boosted liquidity and bribe market, ve(3,3) has this native functionality built into veTOKEN voters. This integrated stack makes ve(3,3) more efficient because no fees have to paid to a third party provider like Convex. I may do some further write up on this in the future but feel it is unnecessary at this point.
Thanks for reading, I hope you have found this information as insightful as I have during my research. Please follow me for more insights on the ve(3,3) model:
