Exploring the Evolution of Spot DEX: Aggregators, veToken, and Uni V3 Model
TechFlow Selected TechFlow Selected
Exploring the Evolution of Spot DEX: Aggregators, veToken, and Uni V3 Model
This article mainly discusses the development history of DEX and its potential future directions.
Navigating Web3 tides with focused insightsAuthor: Kylo, Foresight Research
Tips:
-
AMM and RFQ actually represent the difference between DeFi and TradFi mindsets;
-
AMM improves capital efficiency through LP leverage;
-
RFQ model has a natural advantage for cross-chain transactions;
-
The introduction of Core Pools will significantly change Balancer's revenue structure;
-
Liquidity incentives based on price ranges can reshape the form of Liquidity Positions.
This article mainly discusses the development history and potential future directions of DEXs. Research on Perpetual Trading DEXs has already been covered in detail in "Perpetual DEX: The March Toward LP Productization." This piece instead focuses on spot DEXs, covering areas such as aggregators, veToken models, and Uni V3-style DEXs.
I. Spot Trading Aggregators
1) AMM vs. RFQ
Before diving into the main content, it’s important to review the historical evolution of AMM vs. RFQ. Today, with various foundational frameworks in DeFi relatively mature, we’ve grown accustomed to AMM-based DEXs. However, the original form of DEXs was not what we see today—order books and RFQ were the earliest methods for on-chain asset trading. In the evolution of DEXs, RFQ was eventually overtaken by AMM. The competition between AMM and RFQ actually reflects the fundamental difference between DeFi and TradFi thinking patterns. From a TradFi perspective, finance is fundamentally about improving capital efficiency. Compared to RFQ, AMM mechanisms are extremely inefficient in capital utilization and often expose assets to price volatility when pricing. This suggests that AMM should be inferior to RFQ. Yet, in reality, the opposite happened—AMM has advanced further along the DeFi path, even influencing perpetual trading models.
This indicates that approaching DeFi development from a TradFi logic leads to misjudgment. As a blockchain-native financial system, DeFi must have its own underlying logic that determines its trajectory. We can summarize this logic into several key points:
-
Democratization of LP market making
-
Transparent pricing power
-
LP leveraged capital
Among these three characteristics, the first two stem from Web3 ideology shaping DeFi, while LP leverage can align with TradFi logic. In TradFi, enhancing capital efficiency is considered the essence of finance. Since TradFi systems are already highly developed, composability between different applications is limited. Thus, for TradFi, increasing capital turnover equals improving capital efficiency. The RFQ model exemplifies this approach—achieving higher capital efficiency via high turnover rates. In contrast, DeFi enhances capital efficiency in two ways: increasing capital turnover within individual projects and enhancing asset composability across multiple projects. This means DeFi doesn't rely solely on turnover; it can also reuse capital through composability by stacking asset leverage via LP positions.
Understanding this logic makes it easy to deduce that P2P pools in perpetual trading will become a dominant model alongside order books—GMX and GNS are simply extensions of Uniswap within the DeFi development framework.
Put simply, AMM and RFQ achieve efficient capital use from different angles—the former through composability, the latter through high turnover. In early-stage DeFi, where the internal architecture wasn’t fully built out, boosting capital efficiency via composability was more impactful than merely increasing turnover. This explains why AMM outperformed RFQ in early DeFi development. But development follows a winding path—AMM dominance in trading isn’t absolute. Currently, reusing capital through composability appears to have reached its leverage limit, prompting AMM DEXs to refocus on improving capital turnover. Now, RFQ—a model representing high capital efficiency—is being reconsidered, with the narrative again centering on capital utilization.
2) RFQ as a Complement to AMM
As mentioned earlier, RFQ excels in high capital turnover and lower slippage. Today, most trading aggregators like 1inch and Paraswap incorporate RFQ components. RFQ is introduced as a complement to AMM trading, and it also contributes to increasing the complexity of aggregator tokenomics. Designing economic models for trading aggregators is challenging—if they charge fees, competitors could undercut them on price, so most aggregators don’t take fees. The integration of RFQ provides utility for protocol tokens. For example, 1inch requires market makers to stake a certain amount of $1INCH before executing RFQ functions. Paraswap takes this further by using RFQ as a mechanism for token distribution. Multiple market makers fulfill RFQ roles within Paraswap, and newly issued $PARA tokens are distributed weekly among them according to voting weights determined by $PARA holdings.
3) RFQ’s Role in Cross-Chain Transactions
Currently, common cross-chain transaction models still rely on AMM, but RFQ has inherent advantages for asset bridging. AMM-based cross-chain liquidity solutions suffer from fragmented liquidity and inefficiency: each new chain requires duplicating liquidity pools, and exchange rates are highly sensitive to pool depth. In contrast, using RFQ with market makers as intermediaries avoids such fragmentation. A practical solution involves deploying Brokers on each chain, managed by a shared ledger. When users need to bridge assets or perform cross-chain swaps, only the relevant Broker on each chain needs to coordinate.
Of course, this description of Broker-enabled cross-chain trading is simplified—the actual implementation is more complex. Projects currently adopting similar approaches include WooFi DEX and UxUy. WooFi DEX already supports multiple networks and enables cross-chain swaps between various tokens, while UxUy remains in testnet phase.
II. Changing the Revenue Structure of veToken Models — Introducing Core Pools
veToken model refers to a class of DEXs where token emissions across pools are determined via Gauge Voting, including Curve, Balancer, and various Ve(3,3)-style DEXs. Using Curve as an example, the revenue loop in the veToken model lies in veCRV holders capturing discounted bribes paid by B-side protocols. Continuous $CRV emissions are essentially borrowing against future bribe income. However, as DEXs, Curve and Balancer seem to underutilize their core trading functionality—they primarily function as liquidity protocols, accumulating liquidity through token subsidies, which are then reinforced by B-side bribes. This results in liquidity waste, as most funds “idle” in Curve without active circulation.
This is a major criticism of the veToken model: it attracts large amounts of liquidity through heavy subsidies, yet this liquidity fails to circulate rapidly or generate external yield. Therefore, for the entire Curve system, external revenue sources beyond minimal swap fees consist almost entirely of bribes from B-side protocols to veCRV holders.
In economic terms, token subsidies can be seen as debt—systems borrow against future external revenue to pay users, and this debt must be repaid through actual external earnings. In Curve’s case, the debt consists of periodic $CRV rewards, while external revenue comes from B-side bribes and minor swap fees. If Curve’s economic design could incorporate additional sources of external revenue, the overall debt burden would improve significantly.
Thus, the current state of Curve-style DEXs can be summarized as follows:
- Liquidity within Curve is underutilized
- Adding external revenue streams to Curve would significantly reduce systemic debt
The solution becomes clear: leverage underutilized liquidity within Curve-model DEXs to generate new external revenue streams.
1) Introduction of Boosted Pools
Balancer was the first DEX to propose this solution, introducing Boosted Pools to increase LP returns. Boosted Pools combine Composable Stable Pools, Weighted Pools, and Linear Pools into a more sophisticated liquidity structure. Weighted Pools are AMM pools with customizable weightings—while standard AMMs use 50/50 splits, Weighted Pools allow custom allocations (e.g., Radiant Capital’s dLP uses an 80/20 RDNT/ETH split). Linear Pools are designed specifically for an asset and its derivative, such as DAI and aDAI. Typically, Weighted Pools work with Linear Pools to create Nested Linear Pools. For instance, a Nested Linear DAI Pool might use a 20/80 DAI–aDAI weighting, issuing LP tokens labeled bb-a-DAI. Similarly, there are bb-a-USDC and bb-a-USDT. Combining bb-a-DAI, bb-a-USDT, and bb-a-USDC into one pool creates a Composable Stable Pool, whose LP token is called bb-a-USD.
Although bb-a-USD is an LP token, it adheres to the ERC-20 standard and can thus be paired with other assets in AMM pools. A notable example is the bb-a-USD/ETH pool, characterized by:
- Enabling swaps between ETH, DAI, USDT, USDC, aDAI, aUSDT, and aUSDC
- Generating continuous interest due to aToken accrual
Such AMM pools that continuously earn interest are defined by Balancer as Boosted Pools.
The high capital efficiency of Boosted Pools stems from leveraging assets through LP token issuance. Traditional AMM pools require deep liquidity to maintain price stability during swaps, meaning much of the capital serves price stabilization rather than facilitating trades. Boosted Pools innovate by separating these functions—one portion handles swapping, while another portion used for price anchoring is deployed in external protocols to earn yield, with the resulting LP tokens fulfilling the original price-stabilizing role.
However, the high capital efficiency of Weighted Pools benefits only LPs, not the protocol itself. To address this limitation, Balancer proposed Core Pools via BIP-19, aiming to ensure the Balancer protocol also benefits from Boosted Pools.
2) Introduction of Core Pools
Key elements of BIP-19 include:
- Selecting certain Weighted Pools to form Core Pools
- Interest earned within Core Pools is distributed as follows: 50% to LPs, 17.5% to BalancerDAO, and 32.5% as Bribes to incentivize veBAL holders to vote for Core Pools, thereby increasing their TVL
- 25% of Bribe funds are directly allocated to Aura Finance’s bribe marketplace
- Remaining bribe funds go to Redacted Cartel’s Hidden Hands
The introduction of Core Pools guides Balancer’s future direction, encouraging other pools to transition toward Boosted Pool structures. Since the weekly $BAL released for liquidity incentives remains constant, official bribes directing veBAL votes toward Core Pools cause most $BAL emissions to flow into Core Pools, leaving less incentive allocation for other pools. Hence, joining Core Pools becomes the optimal choice for maximizing returns.
Pre-BIP-19, protocol revenue distribution was as follows:
- 50% of trading fees to LPs, 17.5% to BalancerDAO, 32.5% to veBAL holders
- 100% of bribes to veBAL holders
Post-BIP-19, the distribution becomes:
- 50% of trading fees to LPs, 17.5% to BalancerDAO, 32.5% to veBAL holders
- 100% of bribes to veBAL holders
- Interest from Core Pools: 50% to LPs, 17.5% to BalancerDAO, 32.5% as bribes to encourage veBAL voting for Core Pools
Core Pool yield sources vary, including lending protocols like Aave and LSD assets like stETH. From an asset swap perspective, Core Pool LPs trade 50% of yield for $BAL incentives funded by 32.5% of yield. As long as the value of received $BAL exceeds the forgone 50% yield, LPs profit overall. For veBAL holders, since $BAL’s intrinsic value equals the discounted sum of all future bribes to veBAL, Core Pools introduce additional real external revenue, theoretically increasing $BAL’s fundamental value.
Overall, Core Pools transform Balancer’s revenue structure by adding external yield-bearing assets as a new source of protocol income. Compared to other veToken-model DEXs, this better reduces systemic debt.
III. Uni V3 with Liquidity Incentives — Reshaping Liquidity Position Morphology
During the Arbitrum airdrop, TraderJoe captured significant $ARB-related trading volume due to its unique V2 algorithm, making Uni V3 a short-term DeFi spotlight and pushing Timeless Finance—a platform featuring Uni V3-style liquidity incentives—into the limelight. Consequently, claims emerged that “Uni V3 models with liquidity incentives will lead the next DEX narrative.” Currently, there are two forms of liquidity incentives and bribing on-chain: one targets overall LP depth regardless of price, the other ties incentives to specific price ranges, requiring price guidance through differential rewards across price zones. These two models differ significantly—the former aims to deepen liquidity universally, while the latter focuses on targeted price zone optimization.
From a market demand perspective, deepening liquidity is a universal need across protocols, whereas price-guided liquidity is limited to specific pegged assets like ETH-LSD pairs or stablecoins. Therefore, demand for deeper liquidity is broader, while price-guided incentives serve a narrower set of use cases. Based on this classification, Curve and Timeless Finance (for Uni V3) fall into the first category—focused on deeper liquidity—while Maverick Protocol belongs to the second, offering more precise and efficient liquidity placement.
Two pain points currently exist in AMM liquidity incentives:
- In Uni V3 AMMs, only liquidity near the current price range is active; liquidity outside earns little usage, yet both types receive equal rewards under depth-based incentive schemes
- For Curve-like models, incentivizing liquidity depth for pegged assets may worsen de-pegging
The second point may be hard to grasp. Take stETH as an example: before the Shanghai upgrade, ETH deposited into the beacon chain couldn’t be withdrawn, causing stETH to trade at a discount (~0.98 ETH). To earn stETH–ETH Curve LP rewards, users had to deposit liquidity at this 0.98 rate. This added deeper liquidity at a mispriced level, reinforcing the discount rather than helping stETH return to parity.
Maverick Protocol offers a solution to both issues, calling it Boosted Position. The team uses the term “surgical” to vividly describe how Boosted Position precisely guides liquidity. Using stETH again as an example, its “surgical” features include:
- Extra incentives for LPs whose positions automatically rebalance following stETH’s price movement
- Additional rewards for LPs providing liquidity in the 0.98–1 price range
Thus, Maverick’s liquidity guidance operates on two fronts: directional price alignment and targeted liquidity zone incentives. Through selective yield boosts, it reshapes Liquidity Position morphology, concentrating incentivized liquidity around the spot price and improving capital efficiency.
The next DEX likely to adopt price-range-specific liquidity incentives is TraderJoe, thanks to its unique V2 mechanism that seamlessly integrates such incentives. Key aspects of TraderJoe’s design include:
- Price bins are pre-defined; LPs can only add liquidity within fixed intervals
- Each adjacent interval forms a “bin,” representing a single-price liquidity point. Price remains unchanged until liquidity within a bin is exhausted
Due to TraderJoe’s clear price binning and vertical aggregation of LP liquidity within each bin, distributing liquidity incentives becomes significantly simpler. Simplified incentive distribution implies that adding bribe functionality on TraderJoe may require only integrating a third-party module.
Conclusion
The DEX examples discussed above provide a macro-level view of DEX evolution and future possibilities. In practice, there are also intriguing micro-innovations—such as Crocswap’s volume-adjusted swap fee mechanism, Cowswap’s MEV-resistant batch auctions, Cow Protocol’s Coincidence of Wants system, and 1inch’s MEV-protecting rabbit hole function in collaboration with MetaMask. These small innovations serve as optional tools users can leverage during swaps. Additionally, studying monetization models of fee-free protocols like 1inch and WOOFi presents another fascinating research direction. “We all thought DeFi had peaked, but it keeps exploring new mechanisms and business models.” Whether RFQ, new Balancer revenue designs, or Uni V3-style bribing represent hype or the future—only time will tell.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
Foresight Ventures
