DeFi's sudden popularity in 2020, excessive increases, constant security
concerns, and the entry of so much capital, where DeFi will be pushed? What future directions are more worthy of our attention?
concerns, and the entry of so much capital, where DeFi will be pushed? What future directions are more worthy of our attention?
LLab joined hands with the Bytom’s senior researcher Ma Lie to answer the questions for you, as well as a special guest Pan Chao came to the scene .
LLab: First of all, thank you very much for coming to our AMA, this AMA theme is "DeFi Watcher", the invited guest is the Bytom’s senior researcher - Ma Lie. Let's ask Ma Lie to introduce himself first.
Malie: Hello, I'm a researcher of the Bytom, Ma Lie, responsible for the
exploration and research of MOV business model. Thank you very much for LLab's invitation! I see a lot of bosses in the group whose knowledge to DeFi should be more than to me. The theme of this sharing is DeFi's Watcher. About the Watcher,everyone's perspective is different, and today I'm just offering my perspective and communicating with you.
exploration and research of MOV business model. Thank you very much for LLab's invitation! I see a lot of bosses in the group whose knowledge to DeFi should be more than to me. The theme of this sharing is DeFi's Watcher. About the Watcher,everyone's perspective is different, and today I'm just offering my perspective and communicating with you.
LLab: Thank you very much for coming to our community live, we've seen that the DeFi is hot this year. Do you think the concept of "DeFi" accurately sums up what it's doing? What do you think it should be like?
Male: The first question is very good. I also think that we should limit
the intension and extension of Defi before we discuss. Otherwise, we might say two different things.
the intension and extension of Defi before we discuss. Otherwise, we might say two different things.
DeFi is an abbreviation for Decentralized Finance. Let’s take it apart. Finance includes what such agreements are doing; in my opinion, Decentralized describes the way to implement such financial services.
If we limit the financial services associated with cryptocurrencies to
financial services based on smart contracts, then I think the concept of DeFi is accurate, like the financial services that include transactions, loans, and derivatives in Ethereum and Bytom are listed here.
financial services based on smart contracts, then I think the concept of DeFi is accurate, like the financial services that include transactions, loans, and derivatives in Ethereum and Bytom are listed here.
LLab: We all know a very hot project-COMP recently. According to CMC data, COMP current price is around ¥1,100, the COMP market value is around ¥2.9 billion. But as we know that COMP token is only governance token, and do not have any dividend authority. How does it support such a high currency price? Is there a thunderstorm risk?
Malie: Talking about DeFi, Compound can't be avoided now. COMP pre-circulation is low, it is relatively easy to increase the currency price. Now with the increase in the circulation of COMP, I think it is more difficult
to maintain its current market value.
COMP is obtained through liquidity mining. In fact, it has the problem of early high inflation like many tokens. People's capital also has opportunity
costs, and many of them are borrowed and bear the direct interest costs, so most participants might sell directly after mining COMP. We can also see that the market value of COMP has continued to decline since late June.
to maintain its current market value.
COMP is obtained through liquidity mining. In fact, it has the problem of early high inflation like many tokens. People's capital also has opportunity
costs, and many of them are borrowed and bear the direct interest costs, so most participants might sell directly after mining COMP. We can also see that the market value of COMP has continued to decline since late June.
The COMP mining revenue’s decline will also lead to a reduction in its
mining funds, which should be predictable. However, the mining of COMP will last 4 years, so the impact of mining will not disappear. If COMP can remove some high-risk leverage in this process, then the liquidity mining itself should have a positive impact on the agreement.
COMP does not pay dividends which avoids the risks that are regarded as
securities. At the same time, I don’t think that the temporary non-dividend is a big problem. We know that many technology companies' shares are also non-dividend. Since COMP has governance authority, it is possible to receive dividends through governance in the future, and of course it can take other means, such as "repo destruction".
mining funds, which should be predictable. However, the mining of COMP will last 4 years, so the impact of mining will not disappear. If COMP can remove some high-risk leverage in this process, then the liquidity mining itself should have a positive impact on the agreement.
COMP does not pay dividends which avoids the risks that are regarded as
securities. At the same time, I don’t think that the temporary non-dividend is a big problem. We know that many technology companies' shares are also non-dividend. Since COMP has governance authority, it is possible to receive dividends through governance in the future, and of course it can take other means, such as "repo destruction".
LLab: I'd like to ask about distributing governance tokens to protocol users in the hope of encouraging users to use the protocol and distributing ownership to active stakeholders, is such an incentive model sustainable? Do you think liquidity mining or "yield farming" will bring healthy governance to DeFi?
Malie: I think there are two problems: one is whether the decentralized
protocol needs governance tokens; the other is how to allocate governance
tokens;
protocol needs governance tokens; the other is how to allocate governance
tokens;
The first question: First of all, I think that the decentralized protocol should ultimately decentralize governance authority. MakerDAO has made a good example of this point. Secondly, the governance token is a choice for the implementation of decentralization governance. Whether there is a better option is unknow, we have very little exploration in this area. Actually, Compound has previously explored voting directly using the user's historical accumulated interest as votes. But in fact, I think that this method will have some drawbacks, so I won't go into details here. Issuing token is a better choice.
The second question: How to allocate governance tokens. Regarding to the allocation of tokens, we have seen several ways in the past – mining based on consensus mechanisms such us PoW/PoS, private sales, public sales, and as well as liquidity mining based on business incentive distribution.
The second question: How to allocate governance tokens. Regarding to the allocation of tokens, we have seen several ways in the past – mining based on consensus mechanisms such us PoW/PoS, private sales, public sales, and as well as liquidity mining based on business incentive distribution.
In this regard, we see that MakerDAO MKR is distributed to specific organizations or individuals for private sales, which have a positive effect on maintaining a high level of governance at an early stage; MakerDAO can use the raising funds to indirectly promote its business development.
Compound and many other DeFi agreements in liquidity mining actually use direct incentives or subsidies to their own business. Facts have proved that this method is very effective for short-term business improvement. In fact, we have already observed some examples showing that the high returns to partipants from such distribution methods cannot be sustained, but mining incentives will still work throughout the mining cycle.
Therefore, I think this model’s sustainability has yet to be verified which
brings some risks as compensation in the high-yield stage, and even affects
other agreements.
"Will liquidity mining or ‘yield farming’ bring healthy governance to DeFi?
"As I said earlier, liquidity mining is actually a problem of token allocation and cold start, and whether the agreement can be managed healthily is actually another complicated issue. The impact of liquidity mining on governance is mainly reflected in the votes distribution.
LLab: From the data, the growth rate of DeFi project tokens in 2020 has exceeded the mainstream tokens, like BTC and ETH, and there are still many institutions influx. In the face of DeFi projects that have already experienced high growth rates, do ordinary investors still have opportunities to buy? Do you have any investment advice if someone want to buy?
Ma Lie: In addition to the high gas fees, the threshold for ordinary investors to participate in DeFi is already very low. The current DeFi is not as invisible and intangible as the IC0 project in 2017. I suggest that users at least fully understand it before considering buying a certain DeFi token, including experiencing its entire business.
In addition, as mentioned earlier,liquidity mining project tokens may face the problem of early high inflation.High currency prices are difficult to sustain, and relevant risks need to be paid attention to. General investors can remain patient and wait for inflation to come down before participating. At this time, many projects have also passed the test of time. I personally invest in DeFi projects from a relatively long-term perspective.
LLab: At present, many DeFi projects have involved automatic market maker (AMM) in order to solve liquidity problems. The currently known AMMs include CSMM, CMMM and CFMM. What are the advantages and disadvantages of these types of AMM? Which AMM can better solve the current problem?
Ma Lie: Yes, there are already many use cases for AMM in DeFi, specifically CFMM. In response to this question, I consulted someone who is more suitable for answering this question than me-my colleague, the chief researcher of Bytom, Liu Qiushan, who is one of the main designers of MOV superconductivity and has a deep understanding of this question. I am sharing his answer here.
CSMM, CMMM, and CFMM, including the more famous CPMM (Uniswap), are typical representatives of AMM's landing on DeFi. Practice has also proved that these purely mathematical-based market maker mechanisms can make decentralization trading platform has an experience and just-needed scene comparable to DeFi.
From the perspective of principle,risk path, and engineering perspective, CSMM is the simplest and most straightforward which can maintain a constant exchange rate. The shortcomings are also obvious. There is a moment of exhaustion of liquidity. Once a certain currency is exchanged, it will be difficult to establish a liquidity pool spontaneously. Of course, it can be controlled by the so-called liquidity protection and manual intervention mechanism, which also greatly reduces the experience. At present, few projects directly adopt the CSMM market-making mechanism. The centralized stable currency exchange scenario will be partially introduced.
From low to high, the other three types of comprehensive complexity are CPMM <CMMM <CFMM. These three are often put together for comparison. From the perspective of application scenario overlap, CPMM and CMMM have a large overlap. CFMM is generally designed for specific exchange scenarios, such as stable assets exchange. The CFMM function construction process will also be much more complicated than the previous two, requiring the use of advanced mathematics knowledge and
calculation, but the corresponding system and algorithm complexity and risk will also be higher than CPMM and CMMM. Some advanced mathematical analysis and optimization methods, such as convex analysis/optimization, can be easily adapted to the simple mathematical formulas used by CPMM and CMMM, but difficult to directly act on CFMM.
From the perspective of principle,risk path, and engineering perspective, CSMM is the simplest and most straightforward which can maintain a constant exchange rate. The shortcomings are also obvious. There is a moment of exhaustion of liquidity. Once a certain currency is exchanged, it will be difficult to establish a liquidity pool spontaneously. Of course, it can be controlled by the so-called liquidity protection and manual intervention mechanism, which also greatly reduces the experience. At present, few projects directly adopt the CSMM market-making mechanism. The centralized stable currency exchange scenario will be partially introduced.
From low to high, the other three types of comprehensive complexity are CPMM <CMMM <CFMM. These three are often put together for comparison. From the perspective of application scenario overlap, CPMM and CMMM have a large overlap. CFMM is generally designed for specific exchange scenarios, such as stable assets exchange. The CFMM function construction process will also be much more complicated than the previous two, requiring the use of advanced mathematics knowledge and
calculation, but the corresponding system and algorithm complexity and risk will also be higher than CPMM and CMMM. Some advanced mathematical analysis and optimization methods, such as convex analysis/optimization, can be easily adapted to the simple mathematical formulas used by CPMM and CMMM, but difficult to directly act on CFMM.
The CFMM curve for stable currency exchange scenarios is a form between CSMM and CPMM, that is, it can have the advantage of CSMM to maintain a stable exchange rate, which is a rigid demand for stable currency exchange scenarios. It also can be used like Uniswap when the liquidity is about to dry up, the price will be continuously increased automatically in order to maintain endless supply and flexible exchange rate adjustment. The more the dry up, the greater the price. After all, there will be exchange rate deviation between stable currencies. CFMM needs to be able to automatically follow market fluctuations by attracting arbitrageurs to complete the exchange rate balance.
This CFMM should be the most efficient mechanism in the stable currency exchange scenario which is very competitive even compared to DeFi. The shortcomings are as we just said, its research and engineering have just started. For DeFi, it is a 0 to 1 process. There are very complex principles, engineering and risk. If the team does not have sufficient capacity, not only will not use it to bring benefits, but will bring disasters to itself. Because once failures and loopholes occur, there will be system-level financial risks. Another shortcoming is the scenario and benefits. This kind of CFMM is constantly pushed to more scenarios and users, so that it can bring considerable annual income to the liquidity provider and retain LP, which is also a problem.
Finally, we say that CPMM and CMMM are very similar, not only in the application scenario, that is, the exchange between any two assets, like the DeFi trading pair, but also the mathematical principles and mathematical formulas are almost the same. The CMMM just weights each item on the basis of the constant product of CPMM, so strictly speaking, it should be called a constant weighted geometric mean market maker, and the role of weighting is often more limited, especially in the exchange scenario. Generally, the weights of weighting are set to be the same, which is the same formula form as CPMM.
At present, CPMM and CMMM are almost the same, and the two are constantly grabbing each other's market share. They all maintain the most concise mathematical form, can provide endless liquidity, and can flexibly change the on-site exchange rate as the market exchange rate fluctuates. They are very suitable for the exchange scenarios of mainstream currencies and altcoins. And these exchanges also occupy a large share of the cryptocurrency trading market, so they can become the DEX platform with the largest trading volume. Their fully open currency listing and fee adjustment models also greatly shake DeFi's traditional status. Their shortcomings are also obvious. There will be obvious impermanence losses and slippage losses, especially when encountering extreme market conditions and lightning loan attacks, which not only bring system risks to themselves, but also bring the combine risk disasters to other DeFi products combined with them. Therefore, when building this trading platform, the most important thing is the construction of risk and
protection mechanisms, which will be far more important than the business itself.
Finally, we say that CPMM and CMMM are very similar, not only in the application scenario, that is, the exchange between any two assets, like the DeFi trading pair, but also the mathematical principles and mathematical formulas are almost the same. The CMMM just weights each item on the basis of the constant product of CPMM, so strictly speaking, it should be called a constant weighted geometric mean market maker, and the role of weighting is often more limited, especially in the exchange scenario. Generally, the weights of weighting are set to be the same, which is the same formula form as CPMM.
At present, CPMM and CMMM are almost the same, and the two are constantly grabbing each other's market share. They all maintain the most concise mathematical form, can provide endless liquidity, and can flexibly change the on-site exchange rate as the market exchange rate fluctuates. They are very suitable for the exchange scenarios of mainstream currencies and altcoins. And these exchanges also occupy a large share of the cryptocurrency trading market, so they can become the DEX platform with the largest trading volume. Their fully open currency listing and fee adjustment models also greatly shake DeFi's traditional status. Their shortcomings are also obvious. There will be obvious impermanence losses and slippage losses, especially when encountering extreme market conditions and lightning loan attacks, which not only bring system risks to themselves, but also bring the combine risk disasters to other DeFi products combined with them. Therefore, when building this trading platform, the most important thing is the construction of risk and
protection mechanisms, which will be far more important than the business itself.
LLab: Actually, I also want to hear what Pan Chao (MakerDAO Economic Researcher & Head of Chinese Community) thinks about this question?
Pan Chao: AMM (automatic market maker) has many misunderstandings in DeFi. AMM is associated with high-frequency trading in traditional financial transactions, requiring algorithms to be sufficiently complex, smart, and fast to respond, and heavily rely on calculations. In DEX, which is endogenous to Ethereum, AMM can only adopt a simple model due to the limitation of calculation cost, but the complexity of the financial market is obviously unable to be simulated by a linear function.
Either the constant sum market maker or the average value is the result of arbitrage simply simulating inventory changes, and price discovery cannot be achieved. I wrote a market maker article about automatic market making on DeFi, "Automatic Market Maker DEX Economics"
Pan Chao: AMM (automatic market maker) has many misunderstandings in DeFi. AMM is associated with high-frequency trading in traditional financial transactions, requiring algorithms to be sufficiently complex, smart, and fast to respond, and heavily rely on calculations. In DEX, which is endogenous to Ethereum, AMM can only adopt a simple model due to the limitation of calculation cost, but the complexity of the financial market is obviously unable to be simulated by a linear function.
Either the constant sum market maker or the average value is the result of arbitrage simply simulating inventory changes, and price discovery cannot be achieved. I wrote a market maker article about automatic market making on DeFi, "Automatic Market Maker DEX Economics"
Better AMM models need to be more complex, and complex models need to be built on networks with lower consensus costs. The settlement layer should be left in Ethereum, and transaction matching should be placed under the chain or the second layer.
LLab: Thanks for Ma Lie and Pan Chao’s sharing. We understand that most of the current DeFi "applications" still create liquidity through cryptocurrency mortgages for the digital world. When will we be able to access real world assets, and is it difficult to access real assets?
Ma Lie: Yes, the assets that DeFi relies on are still cryptocurrency, and
due to the demand for assets on the chain, we see that BTC is crossing the
chain to reach Ethereum and other chains.
The scale of assets on the chain limits the development of DeFi. Without the introduction of real-world assets, each protocol can only compete for a limited cake. I think the potential of DeFi is far more than what it shows now. In my opinion, the large-scale of real-world assets entry into DeFi , which will definitely happen, but I don't know when it will happen.
There will be some problems, the first is the ideological problem. The DeFi community is actually resistant to real-world assets, but USDT and USDC are both real-world assets in some extent,and they are currently being accepted more and more widely in DeFi-I think the “law of true fragrance” can work.
There will be some implementation problems when real assets enter DeFi. For example, many real world assets are different from cryptocurrency. They all have a certain period of time, and many specific issues need to be resolved. In addition, real assets enter DeFi which also need to deal with regulatory issues. It's really time for real assets to enter the DeFi protocol widely, and DeFi will have a richer extension, perhaps it is more appropriate to call it open finance.
LLab: Security is an issue that DeFi will always face. Under what circumstances the security of DeFi smart contracts can carry larger assets? Furthermore, if the security of DeFi is guaranteed in some extent, will this open a path for large-scale commercial use of public chains?
Ma Lie: Security is very important,but it is also extremely complex in the Ethereum DeFi ecosystem. In the Ethereum ecosystem, except the possible security issues of the DeFi protocol's own smart contracts, there are also complex coupling risks between the protocols.
These coupling risks include:
1. Lightning loan
double-edged sword problem, we have seen a number of incidents of 0 cost attacks on DeFi protocols through lightning loans, and I think this issue will coexist with Ethereum DeFi ecology for a long time;
double-edged sword problem, we have seen a number of incidents of 0 cost attacks on DeFi protocols through lightning loans, and I think this issue will coexist with Ethereum DeFi ecology for a long time;
2. With compatibility issues, we've seen some serious currency loss events on both Uniswap and Lendf platforms because of compatibility issues with the ERC-777 standard;
3. Conflicts between business models, for example, the current COMP mining are actually a liquidity black hole, affecting the business of other agreements and bring security concerns.
The coupling risk of Ethereum might only be eliminated through long-term
running-in between the protocols, but in this process, the new protocol might bring new risks; some certain standards to avoid coupling risks will gradually be formed in the later period. Due to the uncertainty in the Ethereum ecosystem, I am afraid that only after the security of its DeFi protocol has been tested over time, can it take on more assets.
"If the security of DeFi is guaranteed to some extent, will this open a path for the large-scale commercial use of public chains?" The security of DeFi is only part of the large-scale commercial use of public chains. Whether the public chain can be commercialized on a large scale also depends on the throughput and security; in addition, even if the DeFi security issues are not fully guaranteed, the public chain can be adopted from other fields.
LLab: Can you talk about the distribution layout of DeFi projects, such as the relationship between aggregators, oracles, DEX, and loan agreements, and the current distribution?
Ma Lie: I think this question is more intuitive to use the picture I posted earlier. We believe that DEX, lending, and derivatives are the three major tracks of DeFi applications, and the oracle plays a very important position among them. If there is no oracle, we see that many DeFi protocols, such as MakerDAO, Compound, Sythetix, etc., will not exist; even the current DEX needs the protection of the oracle.
I can talk about some of my views about the oracle.Smart contracts require decentralized oracles. Among today's decentralized oracles, Chainlink has the largest market value. In fact, I think that market value is a very important attribute of the oracle, and it represents the economic bandwidth of the oracle project. Why do I say that? Because in a decentralized world, there is no law and violence to hold accountability, only economic mechanism design can be used to prevent evil. The security issue is actually an economic issue here.
An oracle project with a larger market value can actually serve more valuable smart contracts, that is, it can carry more business, which will further promote the growth of its market value and complete the commercial closed loop. Relatively speaking, the development of oracle with small market value will be more difficult. However, Chainlink still lacks staking and slash mechanisms. From this perspective, it has not yet completed the commercial closed loop.
In addition, from the perspective of three tracks, except to the underlying protocol, I think composite applications can be a very important development area. In the current Internet (Web2), applications are almost completely isolated. Whether it is the public data generated by the user interacting with the application or the user's own data, which all controled by application provider. They use this to build their own moat.
In the open network, the competitiveness of applications does not come from the establishment of barriers, but from the degree of being "quoted." In an open network, all public data is publicly available, all private data is encrypted and can be shared on a peer-to-peer basis, and each application is not a closed box. This brings the foundation of composability and makes "composite applications" possible.
So far, the main development in the encryption field has been in DeFi, and it will continue to be DeFi in the short term, because its related protocols and tools already support the iteration of composite applications. We have been able to see that many DeFi applications that have gained more adoption are composite applications.
With the increase of applications in Web3, such as the improvement of infrastructure like identity and payment, the possibility of composite applications will greatly increase. This will be an important driving force for the development of Web3 and a further driving force for the development of DeFi. Killer applications can come from this. This is my overall view of the current DeFi landscape.
With the increase of applications in Web3, such as the improvement of infrastructure like identity and payment, the possibility of composite applications will greatly increase. This will be an important driving force for the development of Web3 and a further driving force for the development of DeFi. Killer applications can come from this. This is my overall view of the current DeFi landscape.