Dialogue with DAOSquare: How to Systematically Build a DeFi Cognitive System

Recently, the 20th session of the Catcher Academy, hosted by Chain Catcher, invited DAOSquare researcher Dashuo to give a thematic presentation on "How to Systematically Build a DeFi Cognitive System."

Dashuo is also a co-founder of the Mixed Rice Research Institute, which is a research organization under DAOSquare dedicated to research and sharing across various dimensions such as technology, trends, finance, and entrepreneurship, helping more people grow. DAOSquare was born out of the MetaCartel community and aims to promote the concept of DAOs.

During the presentation, Dashuo shared with community users the logic behind the recent collective rise of leading DeFi projects, common misconceptions about DeFi in the market, and how to grasp certainty amid market uncertainty, among many other insights. The full text is organized below, hoping to inspire readers.

Chain Catcher: You have a very deep understanding of the DeFi field. How did you initially realize the huge potential of DeFi and get involved? Has the current development exceeded your expectations?

Dashuo: The biggest difference in this bull market compared to previous ones is that value investment has started to dominate, with confidence primarily coming from data indicators: TVL, user numbers, and growth, rather than purely relying on faith.

Through data indicators, we can perceive that DeFi is developing rapidly. Money is the smartest; the market is hot simply because the confidence is strong enough. The fact that DeFi is the most resilient during market downturns and rebounds first after the market stabilizes reflects this point.

In addition to general data indicators, the degree of necessity is also a key factor in selecting investment targets. To judge necessity, you first need to see which project you use the most. Borrowing a saying from stock investment, "buy the brands you are familiar with."

If you use Uniswap the most, then buy UNI; if you use 1inch the most, then buy 1INCH; in lending, if you use Aave or Compound the most, then buy whichever you use more.

Similarly, for other projects, even if you are not a user, you can speculate, but very few projects can outperform those commonly used leaders.

Tracing back to the origins of DeFi, during the bear market of 2018, many investors in the market had little money. They needed to borrow to buy more coins they believed were at the bottom. At that time, the main providers of lending were MakerDAO and Compound, but borrowing money from there still left the question of where to buy coins.

At that time, order book exchanges like IDEX and 0x had poor depth, until the launch of Uniswap at the end of 2018, which formed a closed loop for borrowing and buying coins.

However, during this stage, DeFi had not yet heated up. You could only see data steadily growing. Each major market movement was driven by innovation, which fermented for a period before eventually exploding, giving you enough time to observe and think. Once the data reached a certain level, you could make some assumptions: "If certain conditions are met, could it become the catalyst for a bull market?"

For example, if new projects are issued on Uniswap, will people FOMO in due to the profit effect, thereby driving the lending market to pledge mainstream coins to buy altcoins?

Lending and synthetic assets are both innovations, but these two do not play a decisive role for two reasons. First, without a large number of speculative assets on Uniswap, the demand for lending would not increase.

Second, without Uniswap attracting a large amount of stablecoins, the synthetic asset and derivatives market would not become active.

Chain Catcher: From which angles do you usually assess the value of a DeFi project? Can you analyze it based on your most successful investment?

Dashuo: In addition to TVL and user numbers, there are also monopoly power and innovation. Monopoly power comes from network effects and brand reputation, such as Uniswap and Aave.

Although Sushi has siphoned off some liquidity from LPs, most projects still launch on Uniswap because it has the largest user base and the lowest gas fees, which are the two strongest defenses.

Aave is similar; it was the first to issue tokens (even before Compound), and it has a strong community. A large TVL ensures it has enough borrowable assets, and the longer it operates safely, the more it can increase user confidence and reduce insurance costs, which is hard to replicate.

Of course, it is precisely because the accumulation of safety and reputation is hard to replicate that the migration of users to v2 has been relatively slow.

I think my most successful investment was in Kyber at the end of 2019 when DeFi was not yet booming. My friends and I often discussed how interesting Uniswap was, and we referred to it as the foundation of "the prosperity of all tokens."

But at that time, Uniswap had no token; it only had Kyber, which had decent data. Of course, in terms of profit scale, it was not the most successful, but Kyber is most relevant to today's theme of "building a cognitive system," as it was forward-looking.

Additionally, regarding DeFi investments, I have a very knowledgeable mentor who advises to layout the entire sector, which I strongly agree with, as decision-making costs are too high in new fields.

Chain Catcher: Recently, the market capitalization of leading DeFi projects has been rising, breaking many people's perceptions and being regarded as "the strong get stronger." What do you think is the reason behind this upward trend?

Dashuo: This question can be summarized in one sentence: "Large funds pursue certainty."

After the explosive growth of DeFi in 2020, many investors gained huge profits, creating a cumulative effect with newly entering institutional funds. Everyone will only invest in projects that have data, users, and sufficient moats.

Rather than believing in the success of a new project, it is better to bet on an already successful project, which is the fundamental reason for the strong getting stronger.

At the same time, DeFi is still in a competitive state within the crypto circle. If expanded to the global economy, it is still just a small boat. Since the ceiling for simple foundational business is already high enough, large funds naturally prefer these leading projects.

Chain Catcher: New projects are continuously emerging. What channels do you usually use to discover new DeFi projects? What tools do you use to query or assess the value of these early-stage projects?

Dashuo: The channels I use are Twitter, WeChat groups, and Telegram groups. For popular new coins, there’s also a tool called Dextool. I have friends who track large addresses to catch new projects, but I mainly focus my energy on technical research, so I can only pay attention to second-hand information from groups and friends.

But this also has an advantage; it helps me filter out some bad projects. From my observation, the success rate of new projects emerging in a bull market is actually not high.

From a speculative perspective, I mainly look at the investment institutions and partners of the project. Partners are more important than investors because DeFi is combinatorial; it needs support from other mature DeFi modules to heat up.

Additionally, I look at whether there are innovations that can match market expectations, whether they can solve necessity issues, and whether the product is already developed and quickly launched on the mainnet. This has been particularly evident this year, so it’s best to avoid buying air coins.

Chain Catcher: Among the various tracks in DeFi such as lending, DEX, and derivatives, which tracks do you think have the most potential? Which tracks are currently undervalued? Why?

Dashuo: Actually, all of them have great potential. However, DEX (AMM) has a greater entry advantage, derivatives have better profitability, and lending is the core foundational business of DeFi, but it is currently only played within the circle and has not yet expanded outward. There is still a lot of room for imagination in the future.

In the financial industry, we look at how large the cash flow is, then the transaction fees, and how high the transaction fees can go mainly depends on how much certainty value the services provided by the protocol can help users obtain.

For example, YFI has a very high fee, and products like Lido that provide staking liquidity also have high fees, emphasizing certainty returns. Based on this, it can be inferred that fixed-income product protocol fees will also be higher. These products with expected returns have very low trading frequency.

Only DEXs, which do not have fixed return expectations, are high-frequency and have high turnover. The two different fee structures show that the market clearly recognizes high-frequency protocols; only when they are expensive enough will it turn to low-frequency, high-value scenarios.

Chain Catcher: The middleware you mentioned in your recent article is also very interesting. Besides categorizing DeFi projects horizontally, how do you think we can categorize different levels of DeFi projects vertically? How can each level capture value?

Dashuo: Middleware and aggregators are relatively special tracks. There’s no need to elaborate on middleware; the value of oracles should be felt by users and investors of DeFi projects. Without a safe and reliable oracle to feed prices, your assets could be subject to price attacks (flash loans) at any time.

Many people still do not fully understand the value of aggregators. If Uniswap were the only player, it would be meaningless; only when liquidity is fragmented does it benefit them more. By splitting Uniswap's liquidity, Sushi has clarified the position of aggregators, introducing the logic of layering and fragmentation.

Fragmentation: Sushi and Uniswap are the two largest liquidity fragments.

Layering: 1inch is the top-level aggregation interface, with Uniswap, Sushi, and 41 other protocols below it.

Thus, the significance of Sushi forking Uniswap is profound. It not only facilitated Uniswap to issue tokens earlier but also reduced slippage for users after liquidity fragmentation.

Imagine taking water from one tank versus taking the same amount from two tanks; the water level will definitely drop more evenly. The same principle applies in AMM pools. This fork is historically significant, but it does not mean they will necessarily exist long-term.

In 2019, I proposed a concept that "Web3 is first decentralized, then aggregated," which is also validated here. This means that not only blockchain networks will have fragmentation and layering (Layer 1, Layer 2), but liquidity pools will also have fragmentation and layering, and many other unknown innovations will develop this way.

Regarding how each level captures value, I believe it comes down to service value. Whether the aggregator layer is necessary depends on the granularity of liquidity segmentation. Given that Uniswap and Sushi currently occupy 90% of the market, the value is already evident, not to mention liquidity optimization solutions like Balancer, Bancor, DODO, etc.

In this situation, aggregators like 1inch, Matcha, or DODO have become essential for users, at least for me. Although the gas fees are slightly higher, buying or selling through multiple pools is a more cost-effective trading method.

Another situation is that I might not even know which AMM a certain token's pool is built on. For example, once I accidentally bought a token from Uniswap, and it was 30% more expensive than on Sushi. Therefore, while aggregators like 1inch may be costly, they can actually help users save money.

Another essential function of aggregators is private trading, which prevents front-running. Users who frequently use AMMs should be able to relate to this; front-running by bots is very frustrating, as your set slippage tolerance can be completely consumed by bots.

This is even more evident when rushing to buy new coins. With the same 3% slippage, my transaction failure rate on Uniswap is extremely high, while using 1inch's "on-chain pre-trade privacy protection" option rarely results in failure. If users trade through 1inch, capturing value becomes a natural outcome.

Aggregator revenue sources include self-built pools and value-added services. Introducing lending (leverage) and derivatives is also a potential future business. Of course, lending is not a core trading function; it is just one of the strategies to amplify trading volume. In short, the entry point is fiercely contested.

I recently saw the DeFi Pooling concept proposed by Starkware, which formally established the layering logic. Bitcoin represents a distributed personal private key bank, while Ethereum is a distributed smart cooperative bank.

One is aimed at individuals, while the other is aimed at organizations and groups, representing a distributed autonomous organization (DAO). We are just beginning to enter this state. Recently, have you noticed that many project names have the DAO suffix, such as Lido DAO, Badger DAO, PieDAO, etc.? Regardless of whether they end with DAO, they are all DAOs.

The AMM represented by Uniswap is a type of public pool that allows the community to participate in liquidity contributions. This behavior is DAO-like, which order books do not possess. Order books are fully market-competitive and only encourage competition.

AMMs encourage cooperation, promoting the idea of growing the pool together and making money together, which is the true source of value in blockchain. This is also a key criterion when selecting projects: whether the project's goals and economic model encourage cooperation.

Chain Catcher: A few days ago, Chain Catcher published an article titled "Rethinking Decentralized Governance," which raised the question that the more token holders there are and the higher the community voting frequency, the degree of decentralization of the project does not necessarily increase. This is because the stability of the mechanism is also an important factor in measuring a project's decentralization. In this sense, Maker and Yearn are not decentralized enough. What is your view on this perspective?

Dashuo: In the Chinese context, we translate decentralized as 去中心化, but it also has the meaning of "devolution of power" in the dictionary.

Due to historical reasons (ICO regulatory compliance issues), many projects did not directly distribute power (tokens) to their real users, as there were no users in the early stages of the project.

Taking MakerDAO as an example, its community is divided into two parts: lending providers and DAI stablecoin users. MKR is only distributed to investors and not to DAI users.

YFI's distribution method has always been praised, but it has recently encountered financial issues and needs to issue more tokens to sustain project development. Both of these projects have greatly promoted industry progress, but their token structures also have certain problems.

In the early days, MakerDAO needed the foundation's efforts to expand its use cases, and now many projects actively use DAI. In the future, it could try to involve these ecosystem members in governance rather than just relying on token-holder governance. The amount of DAI locked in applications could also serve as voting rights and dividend rights.

The goal of devolution of power is to give power to enough stakeholders. What constitutes "enough" varies from project to project, as long as it does not affect its business growth and community expansion. In the future, there will be more solutions to make governance fairer.

The answer to the previous question has actually led to DAOs. Regarding governance, regardless of its goals, its essence is to rebalance the protocol, such as adjusting protocol tax amounts, adding new coins to liquidity mining, or voting to decide whether to develop a new feature, etc.

These are all processes of breaking and then rebalancing when the protocol expands.

Once we want to grow, there will inevitably be this expansion process, re-coordinating the interests of all parties in the project. The goal of a DAO is to attract more participation and form the largest consensus range through autonomous protocols. Aside from the underlying public chains, more and more applications are beginning to choose delegated governance models, which is very helpful in improving community decision-making efficiency during the early development process.

From our observations, many investors are very interested in proposal content, but due to language barriers and insufficient governance tools, high voting costs, and other issues, there is significant resistance to participation.

Of course, it does not mean that solving technical issues will lead to 90% of people voting for governance. The ultimate goal of governance is to reduce voting, replacing governance with "non-cooperative game" individual behaviors in the community, retaining only the minimum voting requirements. The more formalized governance is, the more inefficient the system becomes.

For example, does listing a token on a DEX require voting governance? A permissionless protocol should not need governance in this regard and can freely list tokens. Uniswap's token list is a great attempt to reduce governance; anyone can create their own token list.

Chain Catcher: The rise of Sushi this year has also been very impressive, rapidly transitioning from a copycat project to a leading project, with significant contributions from market KOLs like SBF. Does this mean that regardless of a project's technology or team background, as long as community consensus is strong enough, it is sufficient to become a leading project? Is this phenomenon reasonable?

Dashuo: There are many factors for Sushi's success, but the most critical factor is that it is reasonable and valuable. Sushi forked Uniswap, splitting the standard AMM into two independent pools, which benefits user experience and decentralization by avoiding single points of failure.

Another fork, YFI/YFII, has not shown any special significance so far. Of course, such products may have localized community demands, and if that’s the case, they can exist independently.

Even if they merge again, it will not result in a 1+1>2 effect; it may instead reduce the motivation for innovation. I believe that having at least two competitive projects in a track is more beneficial than harmful.

Chain Catcher: What do you think are the common misconceptions about DeFi in the market currently? Have you had any significant judgment or cognitive errors in the past?

Dashuo: One common misconception we discussed earlier is that many investors in the market are overly enthusiastic about new DeFi projects, sacrificing positions in leading DeFi projects. In fact, looking back, not many new projects have outperformed UNI or AAVE. Even if there are, it’s unlikely you would heavily invest in them.

Insurance is considered one of the main tracks in DeFi, but due to slow business growth, this sector has performed poorly overall. So, in summary, it still comes down to looking at data indicators. Projects without data support may pump quickly but will crash even faster, making it difficult to become a stable opportunity for wealth accumulation.

Not heavily investing in YFI during the DeFi boom in July-August last year was one of my significant mistakes. The reason for this was that during that time, some projects made quick profits, causing me to lose sensitivity to innovative projects.

Chain Catcher: The disconnect between knowledge and action is also the biggest pain point for ordinary users in investing. Why do you think it is difficult for people to achieve alignment between knowledge and action in their investments? What methods can solve this problem?

Dashuo: List out the fundamentals, important processes, and collaborations you are optimistic about, and during panic moments, review these fundamentals to see if there have been any changes, trying not to be influenced by short-term negative emotions.

Holding coins during a bull market and not looking around is also crucial. It’s not that you can’t adjust your positions, but the premise for adjustment is that the fundamental conditions at the time of purchase have changed.

Achieving alignment between knowledge and action is quite difficult; sometimes I can’t fully do it either. Investing is a process of filling gaps; as long as there is still a gap, the failure rate will increase. Therefore, diversifying investment targets is better. Going all-in for quick wealth is more of a survivor's bias; those who do not pursue quick riches in the crypto circle eventually become wealthy over time.

Chain Catcher: Finally, from the perspective of industry trends, the value of DeFi products has been recognized by the industry. What do you think is the current overall status of value discovery in this market? This market experienced a significant adjustment last September. Do you think a similar pullback could happen within the next six months? How can one grasp certainty amid such market uncertainty?

Dashuo: Market trends are difficult to predict; the only certain thing is the growth of data indicators, which is the biggest difference in this bull market. The recent news of Coinbase's valuation reaching $100 billion has also given the market great confidence. Currently, DeFi is competing with CEX for users, and recently the market share of DEX has reached about 10%, which is astonishing.

Uniswap rose after many big players believed that DEXs could not be asserted, so as an emerging industry, we must "Stay hungry, Stay foolish," and not make hasty conclusions. Keeping an open mind is very important.

After this rapid rise, everyone can feel some bubbles, but I often hear experienced investors around me say, "There is no top in a bull market." As long as TVL and user numbers continue to grow rapidly, we can still remain optimistic.

It won’t take six months; as long as this crazy growth continues, there will definitely be a significant adjustment. The best scenario is that every time DeFi reaches a new level, mainstream prices also rise to boost TVL.

In an uncertain environment, seeking projects with good data indicators that are undervalued is a strategy to increase certainty, but this premise is after allocating DeFi blue chips.