In-depth Analysis of DeFi Innovation Models: From Lending, Trading to Liquidity Provisioning Methods and Second-layer Protocols

Written by | Jiang Haibo

The emergence of smart contracts on Ethereum has only been a few years, and now the funds locked in DeFi projects have exceeded $280 billion. The earliest DeFi projects like MakerDAO and Aave remain highly competitive and even dominate the market, but new innovative projects continue to emerge, hoping to carve out a place for themselves.

This article will detail the most representative innovative DeFi projects from different dimensions such as lending, trading, liquidity provision methods, and second-order protocols, exploring ways to reduce risks, improve trading efficiency, and enhance capital interest rates in DeFi development.

Lending

Among the leaders in lending, both Aave and Compound adjust lending rates based on the utilization rate of funds, which exposes users' deposits or borrowing interest to uncertain risks.

In projects incubated by Delphi Digital, the Mars Protocol in the Terra ecosystem optimizes interest rates in real-time through a self-designed PID dynamic interest rate model, while Apricot on Solana combines lending and leveraged trading to increase borrowing demand for lending projects. Both are micro-innovations to traditional lending protocols. More projects are innovating in areas such as interest rates and collateral.

Interest-Free Lending

Liquity is an interest-free lending protocol similar to MakerDAO. In Liquity, users can collateralize assets like ETH to borrow the stablecoin LUSD issued by the platform, but Liquity replaces the ongoing stability fee in MakerDAO with a one-time fee at the time of borrowing. As of November 17, Liquity's TVL was approximately $2.5 billion, and the circulating supply of the stablecoin LUSD was about $750 million.

In terms of the liquidation mechanism, Liquity also has notable innovations. Compared to other lending protocols, Liquity requires a minimum collateralization ratio of only 110%, leading to higher capital utilization, thanks to its efficient liquidation mechanism. Liquity uses a Stability Pool as the first line of defense for platform liquidations. Users can deposit LUSD into the Stability Pool, and when other users' collateralization ratios fall below 110%, the protocol directly uses the LUSD in the Stability Pool as a source of liquidity to repay the debts that need to be liquidated, and the liquidity pool receives the deposits of the liquidated users. This means that the assets in the Stability Pool will change; for example, initially, it only contains the LUSD deposited by users. When ETH drops, causing some users' collateralization ratios to fall below 110% and be liquidated, the liquidation pool pays LUSD to repay the protocol's debts and receives ETH, reducing LUSD and increasing ETH in the liquidation pool. When market fluctuations are small, the liquidation pool will gain liquidation profits.

Fixed Interest Rates

Traditional capital is more averse to the risks of interest rate fluctuations, and the fixed returns on blockchain are more attractive to traditional financial capital.

Yield, Notional, and Hifi achieve fixed returns through "zero-coupon bonds." Borrowers deposit collateral, mint bonds with a specific maturity date, and sell the bonds to obtain stablecoins; depositors purchase bonds with a maturity date using stablecoins or other assets; liquidity providers offer liquidity between the two to facilitate the entry and exit of deposits and loans at any time.

Taking the recently launched Notional V2 as an example, if a user wants to lend 100 DAI at a fixed interest rate, maturing on December 1, 2021, the depositor needs to convert 100 DAI into cDAI (the interest-bearing asset in Compound) and then trade cDAI in the liquidity pool to obtain 105 fDAI. On December 1, 2021, the depositor can exchange 105 fDAI for cDAI and then convert it back to 105 DAI.

Borrowers can also collateralize 1 ETH to mint a pair of (distinguishing positive and negative) 105 fDAI maturing on December 1, 2021 (requiring over-collateralization, with a minimum collateralization ratio of 150%), and sell the positive fDAI to the liquidity pool in exchange for 100 DAI. Liquidity providers offer liquidity between cDAI and fDAI, acting as counterparties for depositors and borrowers.

Liquidity providers can directly participate in minting nDAI with DAI or cDAI. nToken represents the user's share of liquidity in Notional, and holders can earn interest from Compound, transaction fees from deposits and loans in Notional, interest generated from fCash in Notional, and NOTE token rewards.

nToken can also be used as collateral for borrowing, so borrowers can first mint the corresponding nToken from their collateral before borrowing, leading to higher capital utilization.

From Yield to Notional V2, the capital utilization rate of this type of project is increasing. For example, when the underlying asset is converted into cToken in Compound, even if the deposit matures, cToken can continue to earn interest; liquidity tokens nToken can also be directly used as collateral for borrowing.

Products like Pendle, Swivel, and Element Finance, which tokenize yield, allow users to lock in floating rates in lending protocols like Compound and Aave to obtain fixed rates.

Taking Pendle as an example, depositors lock aTokens from Aave or cTokens from Compound in a smart contract to receive OT tokens representing ownership of the collateral and XYT tokens representing future yields. By selling XYT tokens, users can receive future deposit yields in Aave or Compound in advance.

Counterparties can also buy XYT tokens to gain the rights to future yields from the underlying assets in Aave or Compound. Borrowers in Aave or Compound can buy XYT tokens at the current market price to hedge against future interest rate fluctuations.

Uncollateralized Lending

TrueFi, Maple, and Clearpool can all provide partially collateralized or uncollateralized loans, allowing borrowers to avoid liquidation issues and only pay fixed interest. Anyone can deposit into the liquidity pool, but all three only conduct lending business with reputable institutional users.

TrueFi decides which institutions can receive loans through voting by TRU token holders, while Maple is managed centrally by professional agents, and Clearpool allows users to choose which institutions to lend funds to after voting ends.

These types of projects have some centralized characteristics, may require KYC, and may resort to legal means when borrowers cannot repay. The lack of collateral means that there won't be too many borrowing users. However, they can seek deposits based on borrowing situations, with high deposit utilization, and borrowers do not need to fully collateralize or worry about liquidation, making it the most suitable borrowing method for institutions, with significant growth potential in the future.

Cross-Chain Lending

Compound is temporarily absent from the multi-chain ecosystem of this round of DeFi, but it has proposed an effective cross-chain lending tool called Compound Chain (Gateway, a blockchain based on Substrate, using an authority proof mechanism, authorized by the Compound governance system). In its design, Compound Chain will support cross-chain collateral lending across different blockchain platforms and will issue a native token CASH as an interest payment tool, while the governance token remains COMP.

Taking cross-chain lending between Ethereum and BSC as an example, if a user wants to stake ETH and obtain a loan from BSC, they first need to deposit ETH into Binance Exchange, then withdraw it to BSC, and then use Venus or other collateral lending.

If Compound Chain supports BSC, users can directly deposit ETH on Ethereum and obtain loans on BSC through Compound Chain, without the need for centralized cross-chain methods or decentralized cross-chain bridges, completing one-stop multi-chain lending services. Currently, Gateway has launched a testnet.

Trading

Starting with the initial Bancor and Uniswap, AMMs have undergone various algorithmic iterations, leading to the emergence of Curve, which is specialized for Stableswap, Balancer, an automated portfolio manager supporting multiple assets, and DODO, which uses a proactive market maker (PMM) algorithm. The launch of Uniswap V3 this year is also an innovation of previous AMM algorithms, greatly improving capital utilization and triggering a wave of projects providing proactive market-making strategies specifically for Uniswap V3. Many articles have already introduced Uniswap V3, so I won't elaborate further here.

Liquidity Aggregation

Curve occupies a major market in stablecoin exchanges, allowing for nearly 1:1 exchanges between different stablecoins. Recently, Curve has expanded its business to multiple chains, and the launch of the V2 version tricrypto directly competes with Uniswap V3 in trading volatile assets.

Tricrypto includes BTC, ETH, and USDT, with each asset occupying nearly 1/3 of the pool, gathering more liquidity near market prices. This simultaneously achieves the liquidity aggregation of Uniswap V3 and the function of Balancer to maintain constant asset ratios.

Theoretically, the more concentrated the liquidity, the greater the change in the ratio of the two assets when prices fluctuate. For example, in Uniswap V3, concentrated liquidity also brings more impermanent loss.

Curve V2 addresses this issue through an internal oracle's Exponentially Moving Average price and dynamic fee adjustments. The internal oracle will use a weighted average of the current price and the last oracle price, and after the price update, it will adjust the exchange rate of concentrated liquidity when necessary to prevent excessive participation from arbitrageurs.

Dynamic fees will reference the state of the liquidity pool, and when the ratios of various assets deviate significantly from the balanced state, fees will increase. Currently, the ratio of USDT/WBTC/WETH in the tricrypto pool is 33.03%:33.43%:33.54%, with over $1 billion in liquidity and a fee rate of 0.074%.

Derivatives Trading

The leading derivatives trading platform dYdX only saw an explosion in trading volume after issuing its governance token. The main trading products of dYdX have shifted from leveraged trading to perpetual contracts, deployed on the Ethereum Layer 2 network. Under incentives like trading mining, dYdX's daily trading volume exceeds $1 billion, recently surpassing Uniswap V3, PancakeSwap, and other spot DEXs.

Perpetual and MCDEX have also updated to V2 and V3 versions, with the former using Uniswap V3's concentrated liquidity for market making, and the latter gathering liquidity near the index price through oracles.

In the options trading platform, Lyra hedges the delta risk of liquidity providers through on-chain spot trading, minimizing the risk for liquidity providers.

More products are also enabling permissionless trading through blockchain, such as the stablecoin Angle, which achieves slippage-free exchanges between different stablecoins through hedging agents.

Liquidity

Liquidity mining is at the core of "DeFi Summer," enabling projects to achieve cold starts with sufficient liquidity in the early stages. However, the drawbacks of liquidity mining are also becoming increasingly apparent, as project teams pay large amounts of tokens to obtain temporary liquidity, which shifts to the next place when incentives are insufficient, repeating the cycle. Whales exploiting liquidity to influence the market is also a common occurrence; projects with higher liquidity typically receive higher valuations, and when whales sell tokens and withdraw funds, many projects immediately face demise, such as ArbiNYAN on Arbitrum. Some projects specifically addressing liquidity issues have begun to emerge, helping project teams solve liquidity problems at a low cost.

Bond Market OlympusDAO

Setting aside the controversial launch method of Olympus, it is also the first project to use bonds to obtain "protocol-owned liquidity." By issuing native tokens OHM at a discount, users participate with OHM-LUSD LP, OHM-DAI LP, OHM-WETH LP, OHM-FRAX LP, and other tokens in exchange for OHM that unlocks after a certain period. The protocol attracts users to permanently sell their LP tokens to the protocol using newly issued tokens. So far, Olympus controls 99.89% of the OHM-DAI liquidity on SushiSwap, contributing millions of dollars in trading fee revenue to Olympus over the month, with OHM-DAI being the third-ranked trading pair in terms of liquidity on Sushi.

Nowadays, Olympus Pro promotes bond services for more DeFi ecosystems to achieve a win-win situation, allowing project teams to convert short-term liquidity obtained through incentives into permanent liquidity under their control. Olympus Pro will uniformly display channels for purchasing discounted tokens in its bond market, providing long-term holders with opportunities to purchase tokens at a discount. The more permanent liquidity held by project teams, the more it helps reduce price volatility of tokens and allows them to profit from trading fees. Olympus Pro charges a fee of 3.3% on newly issued bonds as its reserve.

Decentralized Market Maker Tokemak

Tokemak is a protocol designed specifically for liquidity provision, aiming to act as a decentralized market maker. The native token TOKE represents tokenized liquidity, used to influence liquidity direction in DeFi. Liquidity providers (individuals, project teams, etc.) deposit assets into Tokemak to earn TOKE token rewards, effectively leasing liquidity assets to the protocol.

Liquidity managers (LDs), who are holders of TOKE tokens, can stake TOKE in an asset reactor to support the other end of the aforementioned bidirectional liquidity, allocating more liquidity to that reactor. The protocol retains transaction fees while bearing impermanent loss.

In the long run, Tokemak hopes to guide the liquidity of the entire ecosystem solely through protocol-controlled value. Through transaction fees, Tokemak will gradually accumulate its assets to fund the asset side of reactor pools without needing to distribute TOKE tokens for these assets as it does now. When the accumulated liquidity is sufficient, Tokemak will no longer need to pay TOKE to lease liquidity, relying on its owned liquidity to meet traders' needs. This moment is referred to as the "singularity," after which the control of the protocol will be entirely transferred to the DAO.

As the number of Tokemak reactors is limited, competition among project teams for Tokemak's liquidity is bound to become increasingly fierce. In the competition for the second group of Tokemak liquidity reactors, 45 candidate protocols have already participated, including Terra, Perpetual Protocol, Posicle Finance, TempleDAO, APWine, GroDAO, IndexCoop, and others.

Currently, some protocols are helping participants compete for liquidity around Tokemak, such as Votemak, which is forked from Votium. Project teams competing for Tokemak reactors can bribe Tokemak voters to win in the competition for a group of liquidity reactors. Candidate protocols can deposit bribes into Votemak's smart contract, and TOKE holders can collect these bribes after voting.

Short-Term Liquidity Fei Protocol x Ondo Finance

Fei Protocol collaborates with Ondo Finance to provide short-term on-demand liquidity services for project teams. Projects seeking liquidity deposit their native tokens into Ondo's liquidity vault within a limited time, and Fei Protocol mints an equivalent value of FEI, then sends both tokens to AMM to provide liquidity.

This low-cost liquidity is particularly attractive for early projects. Project teams only need to pay a small fixed fee to Fei at maturity while earning trading fees and bearing impermanent loss themselves.

In the long term, decentralized market makers like Tokemak may change the current liquidity mining landscape, allowing project teams to obtain liquidity by paying only a small fee, enabling them to allocate more tokens for business development.

Second-Order Protocols

Chase Devens from Messari refers to projects that utilize the composability of DeFi and build on existing DeFi infrastructure as second-order protocols. From Yearn, which initially helped users aggregate mining, to Abracadabra, which now uses yield-bearing assets from other projects as collateral, they bring higher capital efficiency but also come with compounding risks. Some projects categorized under other categories may also belong to second-order protocols.

Curve One-Stop Platform Convex

Convex is a protocol built on Curve, which is the DEX with the largest on-chain locked value, exceeding $20 billion in TVL.

In Curve, users can lock CRV tokens for a certain period to gain governance rights over the project (such as deciding the CRV rewards for each liquidity pool; stablecoin or Bitcoin-pegged projects wanting to attract more liquidity on Curve need to stake CRV to obtain voting rights), earn up to 2.5 times CRV mining rewards, and share in the protocol's trading fee revenue. Given Curve's role in the entire DeFi ecosystem, the battle for control over Curve has long begun.

If users do not stake CRV tokens, they can directly mine through aggregators like Yearn, enjoying the increased mining speed brought by staking CRV in Yearn, with yields automatically reinvested.

If users stake CRV themselves, the mining speed multiplier will be dynamically adjusted based on the overall staking amount. Self-adjusting the staking amount can cause unnecessary trouble, and without mining, it can lead to waste of staked tokens.

Convex provides a channel for users to stake CRV. Users can stake CRV tokens through Convex to gain accelerated mining rights in Curve. The staked CRV will be irreversibly converted into cvxCRV to enhance mining rewards and earn Curve's trading fees. After mining ends, users can sell crvCRV on the secondary market to reduce capital waste.

We can see that the accumulated CRV in Convex quickly surpassed that in Yearn. The CRV staked by users through Convex will always remain in the protocol. Among the CRV token rewards obtained by users mining in Curve through Convex, 16% will be collected by Convex as platform fees, with 10% distributed to cvxCRV stakers, 5% distributed to CVX stakers in the form of cvxCRV, and 1% as gas fees for contract calls.

Future Yield Alchemix

Alchemix is a synthetic asset platform supported by "future yields." Users stake underlying collateral in the protocol, and the Vault first acts as a yield aggregator (user assets are deposited in Yearn) to help users mine.

After users deposit through Alchemix, the protocol allows them to borrow a portion of tokens in the form of synthetic assets. Depositing DAI allows borrowing 50% of alUSD, while depositing ETH allows borrowing 25% of alETH. For example, with DAI deposits, DAI and alUSD will be repaid at a 1:1 ratio, and users can use either alUSD or DAI to repay alUSD debts. After users borrow alSUD, as the underlying assets generate yields in the aggregator, the yield portion will be automatically used to repay users' debts, allowing users to repay less debt or borrow more alUSD.

Yield-Bearing Asset Reutilization Abracadabra

Abracadabra allows users to borrow stablecoin MIM from the protocol using various yield-bearing assets like yvUSDC and xSUSHI, employing collateralized debt positions for over-collateralization, similar to MakerDAO.

Previously, yield-bearing assets like yvUSDC were not well utilized, but Abracadabra enables these assets to borrow up to 90% of MIM. If MIM is exchanged for USDC and then deposited back into Yearn to obtain yvUSDC, the subsequent steps can achieve leveraged mining using stablecoins.

Functionally, Abracadabra seems similar to Alchemix; Alchemix integrates Yearn, while Abracadabra can use yield-bearing assets from Yearn. However, the borrowed asset in Abracadabra is the stablecoin MIM, which carries liquidation risks, while Alchemix can only borrow and deposit corresponding similar assets, avoiding liquidation risks.

Conclusion

Although many new projects currently directly fork from Ethereum to other emerging public chains, innovative projects continue to emerge, aiming to improve capital utilization, reduce risks, and lower costs.

Some projects build on existing infrastructure, simplifying user operations, optimizing yields, and increasing available leverage, such as Convex and Alchemix. Other projects attempt to change the existing fundamental gameplay, such as Tokemak, which may alter future liquidity supply methods.