Mosaic’s passive liquidity rebalancing module
Disclaimer: Information as of Feb 28, 2022. For the most recent updates, dive into our comprehensive documentation here
Mosaic, a transfer availability layer powering cross-chain-and-layer asset bridging, is transitioning into Phase 2. DeFi is currently multi-chain — applications have separate deployments on different chains — but not cross-chain. Moving assets across chains requires sufficient liquidity on both the source and destination chains, which cannot be guaranteed if the bridging solutions are not proactively monitoring or rebalancing. There is, therefore, a pressing need for a standard liquidity rebalancing procedure for cross-chain transfers.
As a next-generation bridging infrastructure, Mosaic facilitates the seamless transfer and proactive redirection of liquidity from its vaults on different chains. To do this, Mosaic taps into numerous existing bridging infrastructures for passive rebalancing, thereby guaranteeing the success of chain-and-layer agnostic transfers. This article delves deeper into how passive rebalancing works as a module on Mosaic.
What is Passive Liquidity Rebalancing?
As discussed in previous posts, Mosaic consists of several vaults on the chains it supports for LPs to deposit to and this liquidity is referred to as ‘passive liquidity’; rebalancing aims to maintain the LP vaults on different chains so that there is adequate liquidity on each chain. This is done through liquidity forecasting and moving assets across chains through external bridges (click here for a detailed explanation).
Mosaic uses external bridges like Hop, Connext, Multichain, etc., for this purpose. Mosaic is not to be mistaken for a bridge aggregator; it actively uses these bridges to rebalance its vaults to make sure there is sufficient liquidity across different chains so that transfers can go through. Simply put, Mosaic is a liquidity director for bridges.
How does Mosaic’s Passive Rebalancing work?
Mosaic implements a liquidity imbalance auto-detection mechanism for liquidity rebalancing, which taps into external liquidity vaults to algorithmically move liquidity across bridges based on forecasted demand. At a high level, there are two stages involved in passive rebalancing: firstly, forecasting and monitoring for liquidity shortfall and secondly, rebalancing across chains using external bridges.
Mosaic has a built-in forecasting mechanism that can predict in advance when a particular liquidity level will be reached for a certain layer and intelligently rebalances. To do this, Mosaic employs ML-enhanced forecasting with Gaussian Processes. This machine learning model can capture liquidity evolution across vaults and outperforms non-ML-based models, like ARIMA. The ability to predict liquidity values in vaults improves user experience and the overall system’s performance, as low liquidity vaults lead to failed transfers. The goal is to build a balanced network of vaults with sufficient liquidity where needed at all times.
The passive liquidity allocation module
When a liquidity shortfall is detected, Mosaic selects the best route to redirect necessary liquidity to a vault on that layer, using a range of bridging solutions. This ensures that liquidity is restored to a sufficient level so that transfers can go through.
Mosaic’s passive liquidity rebalancing allows for ‘single-sided staking’, where users earn yield by providing liquidity for one type of asset, in contrast to liquidity provisioning on AMMs which requires a pair of assets. Mosaic’s rebalancing mechanism can move liquidity across chains using bridges, thereby allowing users to gain from transfers outside of the chains they originally provide liquidity on.
While LP returns are maximized through rebalancing, the dynamic fee model works to keep transfer fees low for most of the retail transfers. While it’s common for operators to charge a 0.5% fee on all transfers, Mosaic takes into account the supply and demand dynamics of liquidity in each vault. Mosaic charges a 0.25% fee for small transfers and it starts ramping up when the transfer size becomes greater than 30% of the available liquidity on the destination layer, with a cap of 5%. Using data from Mosaic’s Proof of Concept (PoC) and Liquidity Simulation Environment (LSE), Composable discovered users incur an average transaction fee of 0.32%, while LPs earn an 8.5% yield across ETH and all-USDC-equivalent tokens.
Which bridges will Mosaic support at launch?
Mosaic is integrating several leading bridges, including the following, with more to be announced soon:
- Hop: a protocol enabling asset transfers across EVM-compatible scaling solutions (rollups, sidechains, etc.).
- Connext: a cross-chain liquidity protocol enabling non-custodial asset swaps across EVM-compatible chains, and L2 networks.
- Multichain: a cross chain router protocol that allows users to swap with no slippage between any two chains seamlessly.
In addition to external bridges, Mosaic is looking into other types of liquidity vaults to tap into for rebalancing, such as native bridges on L2s or bridges built by multi-chain DeFi protocols.
Summary
Mosaic unlocks DeFi’s significant potential with guaranteed transfers using passive liquidity rebalancing. By leveraging ML-enabled forecasting models and an existing network of bridges, Mosaic can predict liquidity shortfalls in vaults and ensure liquidity is routed to where it is needed. This drastically minimizes incidents of failed cross-chain transactions.
As an added layer of assurance, Mosaic’s active management module can act as a buffer to passive rebalancing, synergistically competing to ensure that liquidity will always be available, thereby facilitating chain-agnostic transfers regardless of size or volume.
For more information about Composable and how it is architecting the unified DeFi landscape of the future, check out our socials:
