An interview with Liu Jie, Founder of MCDEX: “The Design Logic of MCDEX V3”
Crypto derivatives have been a hot topic this year. Mainstream investment institutions are accelerating the layout of this field. In the first installment of this column, we’ll talk to McDex, the trailblazer of the crypto derivatives market.
MCDEX, an early adopter of decentralized perpetual contracts, recently announced an update to the V3 along with the move to Layer2. In this session, chain catcher discussed the design concept of the MCDEX V3、the innovations and the current state of the derivatives industry with Liu Jie, the Founder and Product Manager of McDex, hoping to provide readers with a new way to think about the industry.
Q：Lately DeFi derivatives have caught lots of eyeballs. As a pioneer in this quite competitive market, where do you think we are at in terms of overall growth？
A: The crypto derivatives market currently includes options, futures, perpetual contracts, and heterogeneous derivatives. Since there is not a standard paradigm in this market, developers are still exploring their strategies. Take our product as an example, we had a hybrid model (order book+AMM) for MCDEX V2.
We’re at a stage where many of the products are evolving into their new versions, which are born from lessons learned and market feedback. Thus this industry, in general, is more likely to flourish. The market of derivatives is about to boom.
Q: The recent volatility has led to a surge in on-chain trading. How is MCDEX doing regarding the trading volume and other indicators now?
A：MCDEX V2 doesn’t have a high trading volume so it’s not really affected. However, lots of liquidations taking place on MCDEX V3 beta testnet is a good quality check for MCDEX’s settlement ability under extreme circumstances. LPs are also generating a profit from this because of the liquidation penalty.
Q：So can you elaborate more on this liquidation penalty ？
A: Suppose that a contract trade has a liquidation penalty of 0.5%, then 0.5% of the position value will go to the LP liquidity pool as liquidation penalty when the position is liquidated.
Q: From the perspective of traditional finance and centralized exchanges, it is still hard for derivatives to get huge traffic. What do you think is the key?
Compared to CeFi, Defi has an easier approach to obtain liquidity. Liquidity mining brings liquidity right away, so this is why we’re sticking with AMM for MCDEX V3. The AMM-only model works well in crowdfunding liquidity. As long as the design is reasonable and has a high capital efficiency, it is quite likely that liquidity will come along.
In the case of a centralized exchange with the order book model, there are only two ways to get liquidity. The first is to raise funds, get liquidity and bring market makers in. The second is to inflate assets, which is problematic. Once the liquidity problem occurs, the exchange will have to do a flash crash.
Q：Can you tell us more about MCDEX V3?
A：Mai Protocol V3 designed by MCDEX is an AMM-based decentralized perpetual swap protocol. Perpetual swap is one of the most popular derivatives that have no expiration date, supports leverages, and has its price soft pegged to index price. The goal of this protocol is to allow anyone to create and trade in any perpetual market.
To start with, anyone can create their own perpetual market with the price feed of the underlying asset and choose any ERC20 as collateral.
Secondly, we have designed an AMM for the perpetual market and this AMM also has better capital efficiency. Moreover, AMM solves the liquidity problem — anyone can provide liquidity to AMM by depositing assets in the pool and get a reasonable market-making profit.
At last, anyone can trade perpetual swaps permissionlessly. Traders’ assets are in the smart contract in a noncustodial way and the process of trading is conducted on-chain completely.
Q：What risks doesV3 have？
A：When AMM holds position, price change could cause a market-making risk. AMM holds position when the market is unbalanced, acting as the counterparty of the traders.
For example, if long position is bigger than short position, there will be a net long position. The difference will be countered by AMM acting as a short, and vice versa.
When AMM holds position, there is a market-making risk, aka inventory risk. Because a price change of the inventory could affect PnL for either longs or shorts. In order to balance this risk, we have designed a mechanism in which LP has a diverse source of income including trading fee, opening slippage, spread, and penalty.
On V3 testnet, the LP APY is already more than 100% without taking into account mining rewards. LP profit increases even more as people are getting liquidated. In such a way, the LP profit is actually the income of market makers and exchanges in a traditional order book model.
We have reshaped the market structure such that anyone can be empowered with the ability to make market and enjoy this profit.
Q：AMM is indeed revolutionary, but the impermanent loss is still a problem for providing liquidity on chain. How do you reduce the LP risk in your design?
A：There is no impermanent loss in V3. We have three ways to buffer the market-making risk of LP.
By charging a funding payment from the counterparty. If AMM holds position, the market is unbalanced. If AMM holds long position, there are more short positions in the market. Then the funding payment will be levied on short positions, that is, all short positions pay all long positions. A certain ratio of this payment goes to LP.
Spread. LP can avoid the problems caused by the inaccurate price feed and can earn a profit from spread, which is also a way to balance risks.
Automatic adjustment of the price feed. In our AMM design, the principle of price adjustment is to discourage trades that increase AMM risk and encourage trades that reduce AMM risk.
For example, AMM now holds long positions. If the trader wants to short, then AMM will have more long positions. AMM will offer a lower short price, which will make the trader unwilling to make further short positions or increase the cost for the trader to short, in order to dynamically protect LP.
In addition, the AMM will encourage the reduction of its own positions. In other words, AMM was selling its positions at a discount, encouraging traders to take them away and thus reducing AMM risk.
These risks are dynamically controlled by several risk parameters. We have a role called operator, which is authorized by LP to adjust these parameters at any time within a range based on market dynamics to help balance the situation.
If the LP risk is too low, the trader’s experience will be worse. Once the trader experience is too good, the risk of the LP increases. The two are in opposition, so it is necessary to find a balance between the LP’s benefit and the risks, while the trader experience is also good. This is a key consideration when we design the AMM.
Q：The design of custom perpetual contracts is like MCDEX providing a platform where anyone can open a store on. How does MCDEX ensure the quality of these stores?
A：We can’t ensure their quality. First, let the market choose the fittest. There are fake tokens in Uniswap and fake products on Taobao. The open platform can never guarantee the quality of the store, but we believe that this is a free market, which can eventually form a survival-of-the-fittest environment and let the good and real ones stay.
Second, we have a short number service. For example, the contract starting with 0 on the platform is verified by us, and the verified contract is more reliable. But certification is only a supplementary means, it still depends on the market to decide, the market will eventually leave the good ones.
Q：At the end of January, the tokenomics was revised and the total amount was reduced from 100 million to 10 million. Was there a problem with the previous one?
A：The total amount we originally designed to be 100 million and we have released 2 million. The community has decided to reduce the total amount to 10 million, which is equivalent to burning 90 million. There’s actually only 2 million in circulation, and that remains the same.
MCB previously adopted a linear vesting model for team, community, and foundation. However, the problem is that the issue is not related to the exchange’s performance, which may easily lead to an overissue without business growth. In the new tokenomics, MCDEX DAO decides the use of vault funds and the use of additional MCB. Any additional issue is restricted by the speed of value capture by the MCDEX DAO vault, that is, at a rate of 1MCB per $1 captured.
MCB holders have governance rights of MCDEX DAO. We let 0.150 ‰ of the transaction amount of the whole platform go to the DAO’s vault. In such way, the DAO vault is able to capture value. MCB holders can decide: the specific usage of the funds and the specific distribution way of the 8 million MCB yet to be issued.
Q：Both Uniswap and Sushiswap have mentioned order books in their V3, but your team seems to be on the other track, order book and then switch to AMM. doing the opposite, making order books first and AMM later.
A: The capital efficiency of their current constant product is very low and the liquidity is not sufficient. To increase liquidity, order book is better in their case. However, there are many ways for AMM to increase liquidity. For example, the capital efficiency V3 AMM is 1000 times higher than that of the constant product AMM.
Q: AMM is supposed to be a major innovation in DeFi, and many protocols are starting to make minor twists around it. Take vAMM as example, can you explain a bit?
A：A vAMM is virtual AMM, which is a mechanism that creates good liquidity out of nothing. However, a virtual AMM corresponds to virtual liquidity. In order to work, arbitrageurs or teams must constantly hedge market-making, that is, to bring in external liquidity.
Therefore, the vAMM liquidity is limited by the total liquidity of the robot (aks, capital scale). The vAMM looks pretty, but it doesn’t solve the liquidity problem, and it loses AMM’s ability to crowdfund liquidity.
Q: Out of so many L2 solutions, why Arbitrum?
A：There are three points. In order to support a platform that will be handling up to billions of dollars in transaction value, we need an L2 solution that is highly decentralized. Thus the side chain option (which is more centralized) is eliminated.
Secondly, we value the platform effect and the universal composability of the solution so that future DeFi projects can interact. Therefore, EVM compatibility is essential. All projects can be move over and the so-called island effect will be avoided. Arbitrum Rollup is very compatible. Some solutions require developers to rewrite the code to complete the migration. The migration cost is high for developers and the platform development is slow. For example, BSC is compatible with Ethereum EVM, so it develops very fast and the platform effect is very strong.
Finally, we also need technology maturity. We are definitely looking for a product that can be used in a short term and has a high maturity. Currently, there are many teams working on layer2 on Ethereum, but the ones that can really take things out and have a high maturity are Arbitrum and Optimistic. However, through testing, we find that the Gas consumption of Optimistic is 10 times that of Arbitrum, and there is a large limit on the size of the smart contract. Overall, Arbitrum is the best fit for us.
“Thinking about the industry”
Q: What will the DEFI market look like when layer2 boom?
A：It releases the DeFi effect. There are two reasons:
Gas costs are too high at this stage, which limits the development of the industry. From the popularity of BSC, we can see that once gas costs are reduced, users are willing to participate in the transaction.
No matter what industry it is, as long as the infrastructure conditions are improved, there will be new scenarios and market space, just like the arrival of 3G and 4G leads to the rise of short video applications. Each generation of infrastructure corresponds to a new generation of applications. When we look back, we can see why the DEX on layer1 uses the constant product formula AMM. The real reason is that the chain is very slow, so constant product is then the choice to compromise. Once the high TPS environment of Layer2 gets popular, there will be a lot of innovative scenarios.
Q：Community governance has been an issue. What is your philosophy on that?
A：I believe in the free market theory. Although macro-control is still needed, I believe the market is powerful in the blockchain environment.
There is a balance between democracy and centralization when it comes to community governance. Our approach is that teams or active contributors of the community provide governance proposals, which are discussed within the community and then voted on.
Q：With the recent spate of thefts, how exactly is the DeFi world responsible for the security of liquidity providers’ funds?
A：The only thing we can do is trying to avoid it. Make sure developers are experienced. The majority of problems can be avoided by finding a top audit institution to audit and check the problems. However, risks in the interaction of smart contracts cannot be completely avoided, because both the contract development and auditing institutions can only be responsible for their own contracts and cannot audit the contracts that may interact in the future.
Q. Some people believe that composability is the future of DEFI. What is your opinion？
A：Totoally agree. We will interact with other protocols in two directions.
One way is that the collateral for a V3 contract could be a token for another project since the AMM pool collateral could be any ERC20 token. Other projects, such as those with their own tokens, can set up their own perpetual contracts on MCDEX.
The second way is to use MCDEX perpetual protocol as an underlying asset for other projects. For example, MCDEX LP Token could be used as the underlying asset for projects like BarnBridge and Saffron.
Users can achieve risk stratification through these two types of projects: ordinary users who do not want to bear the market maker risk of AMM can enter low-risk tranches and receive fixed rate yields; users with higher risk tolerance can enter the high-risk tranches and earn higher profits.