Cross-Chain DEX Tour: How Zenlink Aggregates Liquidity within the Polkadot Ecosystem
Author: echo_z, Chain Teahouse
In the multi-chain landscape of blockchain, how DEX captures the liquidity from multiple chains is a long-standing topic. DEXs have showcased their unique capabilities and developed various models. Currently, cross-chain DEXs can be broadly categorized into three types:
1) Relying on various mature public chains to capture a large amount of liquidity within a single chain, such as Sushiswap deployed on multiple single chains;
2) Focusing on capturing cross-chain liquidity to meet users' needs for cross-chain trading assets, such as Thorchain supporting cross-chain trading of multiple leading native assets like BTC/ETH/DOGE. However, due to its reliance on the intermediary asset RUNE, its security is relatively low and has suffered multiple hacker attacks. Another example is the recent phenomenon product Stargate, which adopts a light client and relay design, offering higher security but currently only supports stablecoin cross-chain trading;
3) Focusing on capturing liquidity within a specific multi-chain ecosystem, which can be further divided into two categories:
One category captures liquidity through a combination of "cross-chain transfer + single-chain DEX" functionalities, such as Osmosis in the Cosmos ecosystem. In this scenario, assets are actually aggregated onto a single chain, but due to the relative convenience of cross-chain transfers within the ecosystem, it can integrate bridge + DEX functionalities within a single product interface, providing a decent user experience.
The other category is the approach taken by the protagonist of this article, Zenlink: deploying DEX protocols across multiple parachains within the Polkadot ecosystem, maintaining the independence of liquidity pools on each chain. When users trade assets, it automatically searches for the most optimal liquidity pool across all chains, completing the transaction cross-chain before returning to the original chain. If Polkadot's public chain infrastructure is akin to ETH, then the TVL of each public chain on Polkadot is similar to the TVL of various DEXs on ETH, so Zenlink to Polkadot is like 1inch to ETH.
In Zenlink's long-term vision, it ultimately aims to achieve direct cross-chain exchanges of assets from chain A to assets on chain B, similar to the combination of "Stargate + 1inch" functionalities within the Polkadot ecosystem.
Previously, Chain Teahouse has published research on Stargate and Osmosis. This article will conduct an in-depth study of Zenlink to present a more complete picture of the cross-chain DEX industry.
1: Product Mechanism
1.1 The Essence of Cross-Chain DEX: Smart Routing Aggregating Multi-Chain Liquidity
The core of Zenlink's realization of a cross-chain DEX is three functional modules: cross-chain transfers within the Polkadot ecosystem, DEXs on multiple single chains, and smart routing between multiple chains. The first two have been implemented, while smart routing has not yet gone live.
It is important to note that only after the smart routing goes live can Zenlink claim to have truly aggregated liquidity. The currently implemented functionality involves first aggregating liquidity within a single chain through cross-chain transfers, and then conducting asset trading within that single chain. The logic of cross-chain transfers and single-chain DEX is easy to understand and is implemented in the "Swap" interface on the official website.
Cross-chain transfers are achieved through the "Cross-chain" option under "X-Transfer" in the "Swap" interface, currently supporting asset transfers within the chains of Moonriver/Bifrost/Kusama/Karuka/Kintsugi/Statemine. It is important to note that cross-chain transfers rely on the establishment of channels between parachains, which depends on the deployment of XCM. Currently, XCM has only been deployed on Kusama, so the X-Transfer supports transfers only between Kusama parachains.
Zenlink Cross-Chain Transfer Function Interface
Another function under "X-Transfer," "Transfer," is used to support the transfer of assets within the Bifrost chain between different addresses. This function was established to compensate for the shortcomings of the Polkadot.js wallet and is unrelated to cross-chain functionality. However, for first-time users, having two types of functions side by side in the same interface can be somewhat confusing.
In the "Swap" interface, selecting the "Swap" function leads to a conventional DEX interface. Since the DEX functionality is deployed on Moonbeam/Moonriver/Bifrost, users can choose assets from these three chains during trading (as shown in the right image). Readers may notice a "Limit" function among these, which is similar to the limit order function of centralized exchanges and is also an innovation of Zenlink, currently still in testing.
Zenlink DEX Function Interface
The combination of cross-chain transfers + single-chain DEX functionalities makes Zenlink quite similar to Osmosis in the Cosmos ecosystem, but the difference lies in that Osmosis is deployed on a single chain, while Zenlink is deployed across multiple parachains.
Zenlink is currently the only DAPP in the Polkadot ecosystem that has implemented the aforementioned functionalities. Other competitors, such as Arthswap on the Astar network and Stellaswap on the Moonbeam network, although their front ends also integrate cross-chain transfers, require third-party transfer functionalities to achieve this, essentially making them single-chain DEXs.
With such a functional combination, liquidity is naturally dispersed across various chains. In the "Staking" interface on the official website, one can see that on different chains, the ZLK single-coin pools have independent TVL and APR.
Zenlink's "Staking" Interface
Liquidity dispersion can lead to shallow trading depth, harming user experience. So how can dispersed liquidity be aggregated? The key lies in the future "smart routing" functionality, which simply means finding the optimal price path among all chain liquidity pools. During this process, a single transaction may be split into multiple transactions across different chains, and after completion, the assets are transferred back to the chain where the user initially traded. The entire process is somewhat similar to how 1inch seeks the optimal path for trading across various DEXs within a single chain, but the difference is that 1inch borrows liquidity pools from other "DEXs," while Zenlink only borrows liquidity pools from Zenlink DEX on other chains.
The smart routing functionality has not yet gone live, and we can currently only see its design in the white paper, as illustrated below:
Source: https://wiki.zenlink.pro/zenlink-dex-dapp/aggregator
Zenlink's smart routing design positions it as a unique model in the spectrum of "cross-chain DEXs": unlike Osmosis/Arthswap/Stellaswap, which aggregate liquidity into a single chain's pool, it maintains the independence of liquidity across chains by seeking optimal paths among them to achieve "on-demand cross-chain".
Osmosis/Arthswap/Stellaswap are essentially single-chain DEXs. By attracting a large amount of liquidity to a single chain, they achieve better trading depth; however, the concentration of single-chain TVL inevitably sacrifices the TVL of other parachains within the ecosystem. In the Cosmos ecosystem, aside from Terra, which has high TVL due to capital support and the introduction of external assets, Osmosis can be said to dominate as a DEX.
Zenlink's positioning is more like serving the entire Polkadot ecosystem, encouraging each chain to have its mainstream assets and liquidity pools. When parachains compete for TVL without a concentrated winner like in Cosmos, Zenlink's value becomes more pronounced, as no chain would refuse Zenlink, and Zenlink can automatically find the best price for users. However, even so, Zenlink will still face competition from independent DEXs on each chain.
Moreover, the initial scenario for the smart routing to go live will still be asset trading on the same chain. Zenlink's ultimate goal is to achieve true cross-chain trading, that is, directly exchanging asset a from chain A for asset b on chain B. This process will depend on the utilization of parachains and XCM. How liquidity will be handled at that time—whether it will be unified on the source chain like Stargate and allocated to various target chains as needed, or whether independent liquidity pools will be created from the source chain to the target chain—will be another question.
1.2 Hybrid AMM: Meeting Various Price Curves
In Zenlink's design, two modes of AMM algorithms will be employed: Standard and Stable. Standard adopts the Uniswap V2 model, suitable for asset pairs of different values; Stable adopts the Curve model, suitable for stablecoin asset pairs.
Currently, the liquidity pools that have gone live all use the Standard mode, and the Stable mode will be launched soon. The team announced in a Medium post this April that they will soon launch a stablecoin four-pool composed of anyUSDT + anyUSDC + FRAX + UST on Moonbeam, followed by other stablecoin pools on Moonriver.
At the same time, Zenlink also provides a Bootstrap function, allowing project parties to achieve initial liquidity pools at low trading costs and complete price discovery.
Zenlink's "Bootstrap" Interface
2: Token Economics
Zenlink has a total token supply cap of 100 million. Due to the team's buyback mechanism, ZLK is overall slightly deflationary.
2.1 Token Functions
The main functions of the ZLK token are incentives and governance.
Zenlink does not use ZLK as a trading fee; instead, it uses the source token of the transaction as the fee currency. For example, when exchanging BUSD for USDC, BUSD will be charged as the fee. The trading fee is 0.3% per transaction, with 50% allocated to LP stakers, 30% periodically allocated to active users, 10% as trading mining rewards, and the remaining 10% reserved.
The incentive function of ZLK is reflected in the LP staking mining and trading mining functionalities.
LP staking mining does not differ much from other DEXs, with the amount of ZLK token rewards allocated based on the depth of each pool, and then distributed according to the user's share. This part of the reward comes from 25% of the total ZLK supply and 50% of the trading fees.
As mentioned earlier, liquidity on each chain is independent, requiring users to first choose which chain to provide liquidity on.
Zenlink's Staking Mining Interface
Trading mining rewards are allocated based on the user's trading volume, allowing users to earn without bearing the impermanent loss risk of providing LP. This part of the reward comes from 15% of the total ZLK supply and 10% of the trading fees.
Zenlink's Trading Mining Interface
In addition, ZLK will also serve as a governance token, allowing holders to vote on Zenlink's product functionalities, although this part has not yet gone live.
2.2 Token Distribution
The distribution of ZLK tokens is shown in the diagram below, with 50% allocated for community incentives, 26% for early investors, 20% for the team, and the remaining 4% for the Zenlink Foundation.
In the community incentive portion, 50% (i.e., 25% of the total) is allocated for LP staking mining incentives, 30% (i.e., 15% of the total) for trading mining incentives, and the rest is for the DAO treasury and community retention, used for future promotions, project grants, and rewards for community contributors.
Source: https://wiki.zenlink.pro/resources/tokenomics
The tokens will be fully released after 36 months, with deceleration points at the 6th, 15th, and 21st months.
Source: https://wiki.zenlink.pro/resources/tokenomics
Of these tokens, ~40% will be released on Kusama and ~60% on Polkadot, stimulating both ecosystems.
Additionally, the team will buy back a portion of ZLK each quarter and burn it, resulting in slight deflation. The funds for purchases currently mainly come from trading fees. So far, approximately 700,000 ZLK have been bought back, accounting for 0.7% of the total supply.
Source: https://dex.zenlink.pro/#/zlk/overview
Currently, the total circulating ZLK is about 20 million, accounting for ~20% of the total supply, with the remaining ~80% to be released over the next ~30 months, which is expected to exert some downward pressure on the token price.
Source: https://dex.zenlink.pro/#/zlk/overview
3: Operational Status
Zenlink launched in November 2021, with a peak TVL of around 100 million USD, recently hovering around 40 million USD, and an FDV of about 45 million USD.
Source: Defi Llama
Within the Polkadot ecosystem, there is a significant trend of homogeneous competition among multi-chains. As the only multi-chain DEX in the ecosystem, Zenlink, despite its unique positioning, inevitably faces competition from other single-chain DEXs. Arthswap on the Astar network launched in February this year and currently has a TVL of around 150 million USD; Stellaswap on the Moonbeam network launched in January this year and currently has a TVL of around 60 million USD, both leading Zenlink in terms of TVL.
The lower TVL compared to competitors may primarily be due to the DEX being deployed across Moonbeam/Moonriver/Bifrost, leading to more dispersed liquidity. Only two liquidity pools on Zenlink have a TVL exceeding 4 million, namely the madDAI/madUSDC pool on Moonbeam and the MOVR/xcKSM pool on Moonriver. The multi-chain strategy inherently leads to fragmented liquidity, which is a disadvantage for short-term competition. In contrast, the leading stablecoin pools on Arthswap and Stellaswap have achieved TVLs of 40 million and 10 million USD, respectively.
Left: Arthswap, Right: Stellaswap
Within Zenlink's planning, two changes are noteworthy.
First, Zenlink will launch on Astar and Acala in the future. Multi-chain deployment is a predetermined route, and its future development will also depend on the development of the Polkadot ecosystem itself. If multiple chains progress simultaneously, Zenlink will likely benefit from this landscape, as there may be overlapping asset pairs (such as stablecoins, which are frequently used) across chains, making inter-chain smart routing more useful.
Of course, Zenlink still faces competition from DEXs on each individual chain. Zenlink is fundamentally different from 1inch, as it does not aggregate liquidity from all other DEXs but can only aggregate the liquidity of Zenlink DEX across different chains. Therefore, it remains necessary for Zenlink to deepen liquidity within its DEX.
Thus, the second change is particularly important: after the launch of the 4pool stable pool, Zenlink will reduce some existing liquidity pools to deepen individual liquidity pools, optimizing trading prices and user experience. If single-chain DEXs (like Arthswap) siphon off most of the stablecoin liquidity within the Polkadot ecosystem, users may prefer to take additional cross-chain steps, transferring assets first to the Astar network, trading, and then transferring back to other networks, which would impact the value of Zenlink and all DEXs on other chains. Therefore, regardless of the circumstances, it is a necessary strategy for Zenlink to deepen liquidity within its single pools.
4: Team and Financing
The Zenlink team is primarily based in China and Singapore, with core members coming from imToken. Founder Guo Tao entered the blockchain industry at the end of 2016, initially working in product-related roles. In September 2017, he joined imToken, working in marketing and operations for the Tokenlon DEX business line, and founded Zenlink in 2020.
In September 2020, Zenlink raised 1.1 million USD in an angel round led by Hashkey, Continue Capital, IOSG Ventures, D1 Ventures, and Youbi Capital. In July 2021, Zenlink completed a Series A financing round led by Alameda Research, with participation from IOSG Ventures, Hashkey, OKEx Blockdream Ventures, Hypersphere Ventures, SNZ Holding, SevenX Ventures, DFG, TRG Capital, and PAKA, with the financing amount not disclosed.
5: Advantages and Risks
Chain Teahouse summarizes Zenlink's advantages as follows: it has a unique positioning as the only multi-chain DEX in the Polkadot ecosystem. When multiple chains thrive together and Zenlink has sufficiently deep liquidity across chains, its multi-chain value will become prominent, allowing users to enjoy cross-chain liquidity without needing to perform cross-chain operations.
The challenges Zenlink currently faces mainly include: multi-chain deployment leading to early liquidity dispersion, resulting in less depth compared to other single-chain DEXs, and the need for operational strategies to make trade-offs to focus on creating liquidity pools with sufficient depth.