In-depth Dialogue on the Current State and Development Trends of Decentralized Finance

Alex: For friends who haven't officially entered this industry yet, how would you introduce what Decentralized Finance is in a way they can understand?

Minda: The way I explain it to my friends now is: Bitcoin is a decentralized currency, and DeFi is an expanded version based on Bitcoin. Apart from currency, the financial systems that can be accessed in traditional finance, such as trading, payment, lending, and banking services, can actually be realized in the expanded DeFi space. You can think of it as an expanded version of Bitcoin, an application version of Bitcoin.

Alex: So they often add a question, which is to say that Bitcoin can be understood as electronic gold, a non-sovereign asset, but most of our traditional financial services seem quite convenient. What is the additional value that DeFi provides?

Minda: I think the biggest difference between what DeFi offers and traditional finance is that DeFi returns to the essence of finance, which is an information network. Traditional finance has completely fragmented this information network, with different regulations in various countries, bank regulations, and policy differences, resulting in information transmission being completely disrupted and facing significant resistance. DeFi restores this, as finance is information. Therefore, whether you are trading, issuing assets, or lending, it all returns to the transmission of information between information, with no obstacles, what is called permissionless. DeFi allows information to be transmitted as efficiently as possible, at the speed of light. In traditional finance, however, you find that information does not travel at the speed of light, or even at the speed of sound, because there are checkpoints everywhere, with border checkpoints, regulatory checkpoints, and checkpoints between banks. Therefore, in terms of efficiency, DeFi must be many times more efficient than traditional finance.

Alex: Understood, let's talk about a deeper topic. In fact, we can say that DeFi has developed more than two cycles up to today. The last cycle, which we refer to as the DeFi Genesis or DeFi Summer, was the one where a multitude of applications flourished, with many new projects emerging in 2020 and 2021. However, very few of them have survived to this day. Moreover, from the current perspective, the number of innovative new projects in this cycle is far less than in the previous one. How would you evaluate the overall state of the DeFi sector at the moment?

Minda: Actually, I think the entire DeFi space is very similar to the technology field, especially the innovation in the financial sector. At the beginning, it was a hundred flowers blooming, with all sorts of narratives emerging. Because people were not yet aware of the narratives themselves, during DeFi Summer, there were truly new financial methods coming out every day. But you find that after reaching sufficient competition, a few tracks will settle down, which we refer to as validated tracks. Looking at these two cycles, I feel that the foundation of DeFi, what we call the original primitives, has not exceeded what it was in 2019. In 2019, we just happened to be at the first wave of DeFi. You can imagine that in 2019 there were Uniswap, MakerDAO, and Compound. The entire foundation of DeFi still revolves around these three, with some variations on top, such as order book implementations, concentrated liquidity, and improvements on AMM. In lending, apart from the current pool lending, there is also this isolated pool lending, but essentially, I believe it has not deviated from these three methods.

So the current situation in the entire track is that I think there are two particularly interesting changes from two cycles. One change is that DeFi has already been heavily commoditized; every new chain and every Layer 2 that comes out has those three main components: stablecoins, lending, and AMM swaps. These three main components are present on every chain, heavily commoditized, and of course, much of it involves copying the code of existing projects in the market since it is open source, with Uniswap and Aave using this code. However, at the same time, another interesting phenomenon is that while being heavily commoditized, the concentration is also increasing. For example, the market share of Uniswap in spot trading and Aave's market share in the lending field are both on the rise. So it actually reflects that in the DeFi track, I think commoditization and increased concentration are occurring simultaneously.

In fact, over the past few years, many new DeFi applications have emerged. Of course, this is also based on everyone's changing understanding of DeFi. From the traditional so-called DeFi centered on decentralization, there are now many applications that combine De-CeFi. So I'm not saying there hasn't been innovation; in fact, the track has already become highly standardized at the fundamental primitive level, with just those three main components, and these three components have also been highly commoditized, while the concentration is increasing. However, on a more segmented level, some new DeFi applications and tracks have also appeared. I think this is a very interesting phenomenon that has emerged as the infrastructure has developed.

Alex: Yes, you just mentioned the three main components: stablecoins, lending, and AMM swaps. In addition, there have been quite a few products in derivatives since the last round. What do you think about the derivatives category? Is it suitable to be done in a DeFi way? Are you optimistic about its future development?

Minda: This point may relate to another issue, which is what the underlying logic of the entire DeFi track evolution is. I have mentioned a so-called first principle of DeFi in my previous discussions. What is the first principle? Firstly, it must be where the resistance is greatest that applications will emerge first. For example, in the past with Ethereum Layer 1, you could consider its medium's main force to be very large, even if the speed of light reaches Ethereum, due to its gas fees and low throughput, the DeFi that can occur on Ethereum's mainnet is limited, which refers to the few types we talked about earlier. For instance, why did Aave fail to implement P2P lending before, but succeeded with the pool model? It’s because P2P lending on such a high gas, low throughput main chain simply cannot operate; its efficiency is too low, and personal matching efficiency is too low. Similarly, why did the order book not establish itself on Ethereum's mainnet? At that time, dYdX had previously implemented an order book on the mainnet but later withdrew and moved to StarkNet, and now they're developing an appchain. You find that the order book also cannot run on Ethereum's mainnet, while AMM has been established.

In fact, the establishment of all DeFi applications follows a principle: it starts from low-frequency applications, such as lending and AMM, which do not have a high frequency. Stablecoins also involve large, low-frequency transactions. As we move forward, we see a shift towards higher performance layer 2s or new layer 1s emerging, where we notice the rise of medium to high-frequency applications. Regarding the perpetual aspect we just mentioned, why are centralized exchanges currently engaged in this instead of the previous Ethereum mainnet? Because centralized exchanges are places where the highest frequency applications can emerge, and perpetual trading can only exist in such environments. What's interesting now is that new high-performance layer 1s, high-performance layer 2s, and app chains are all appearing simultaneously, particularly in the space of perpetual trading with significant volumes. For example, in Base, Synthetix Futures, and Arbitrum with GMX, and now the recently popular Hyperliquid, which, like dYdX, utilizes Cosmos SDK. We find that perpetual applications are indeed high-frequency applications, and they must have high-frequency infrastructure to support them. This is why we see many perpetual applications emerging in this cycle. I believe that the perpetual applications in this cycle may not necessarily compete with centralized exchanges or any specific trading platform, due to various performance issues. However, I think with the emergence of layer 1s and app chains like Hyperliquid, which can be considered very close to the experience of centralized exchanges, there is potential for future competition with centralized exchanges in the area of perpetual trading. Of course, you cannot compare them one-to-one, nor is it necessary. After all, one exists in a DeFi or permissionless model, while the other has KYC and other elements, resembling a centralized exchange model. But in terms of performance comparison, I believe it may come very close to the experience of centralized exchanges.

Alex: You just mentioned that from 2019 to now, the three major components of DeFi or the most market-validated application types haven't changed much. So many people say that the foundational innovations of DeFi have actually been completed and believe that there might not be too many surprises in the future. This round seems to lack products that are as eye-catching as those in the previous round. However, some still believe that the potential of DeFi has not been fully realized. What is your opinion on this view? If DeFi still has significant growth potential in the future, what do you think are its driving factors? How might it evolve?

Minda: Actually, at a fundamental level, because the entire DeFi is based on blockchain architecture, it is block by block, so we see why there isn't much innovation at the fundamental paradigm level. Essentially, it is because whether it's the Ethereum mainnet, layer 2, or Solana, they all build the underlying infrastructure according to a block by block model. However, the entire DeFi, the changes in this cycle compared to the previous two cycles, lie in the fact that everyone’s understanding of DeFi has changed a lot. What was previously called decentralized finance is now not regarded as a core component by everyone. What is more important is permissionless. I see this cycle as relatively innovative, with projects like Pendle and Ethena. Pendle is a protocol for the swap between fixed and floating interest rates, a kind of fixed income protocol. Ethena is a particularly typical stablecoin for USDE. In fact, the strategy for stablecoins, I have been doing since I entered the crypto space in 2014, is to engage in what is called basis arbitrage ( basis point ). The strategies for trading have been around since Bitcoin started, and the market has been continuously involved in how to conduct basis arbitrage between spot and futures. Ethena turns this into a token, completely democratizing the so-called in-house strategies that were previously used only by some traders, allowing the market to capture the basic profits from the fluctuations of Bitcoin or other currencies. If you ask people from the previous cycles, no one would consider Ethena a DeFi project, because its entire infrastructure is built for arbitrage on centralized exchanges, and although the assets are custodial, the custody itself is not held in a contract but by custodial institutions. So from an architectural perspective, it is not identifiable as DeFi. However, from its tokenization, where anyone can mint it and swap it with existing tokens, it is DeFi. Therefore, I think we can refer to it as De-CeFI or Ce-DeFI applications. There are many such applications in this cycle, such as some projects in the Bitcoin ecosystem and some liquid-staking projects, which are very similar to this model.

So I think at this time we have actually greatly expanded the definition of DeFi, and you will find that the contribution of the entire DeFi TVL that has emerged in the past cycle includes a large number of projects like liquid staking, such as the Ethena project I just mentioned, as well as RWA on-chain projects. These actually account for a very large proportion of the overall TVL contribution to the revival of DeFi. Therefore, I believe that the evolution of DeFi is the evolution of the entire concept, from what we used to refer to as pure decentralized finance, to open finance, and now it is actually a hybrid, a mix of centralized and decentralized finance. From this perspective, I think there will be many interesting combinations arising from the opportunities in the future. Of course, if you purely look at it from the most original, fundamentalist DeFi level, there really aren't many options. Because in the past, for example, the decentralized stablecoin sector was basically not being pursued, or projects in the Ponzi stablecoin sector that were particularly degenerate are no longer being developed. Unlike during the DeFi Summer when a bunch of projects emerged, there were decentralized projects with on-chain governance.