Partnered With:

Celer's new expansion plan: to minimize the threshold of DeFi

Read this article in 40 Minutes

Layer2.finance is not a new DeFi protocol, but a "General Entrance to the DeFi World" with low barriers to entry and no need to trust a centralized third party.

Original source: Celer Network

Today we officially released the latest solution of Celer Network: Layer2.finance. Layer2.finance is committed to lowering the threshold for using DeFi, so that the general public no longer has to worry about high transaction fees, and can freely and conveniently use existing DeFi protocols to obtain high returns. Layer2.finance is not a new DeFi protocol, but a "General Entrance to the DeFi World" with low barriers to entry and no need to trust a centralized third party.

As we all know, the existing Layer-2 expansion solutions require the overall migration of the DeFi protocol and ecology , which will not only affect the composability of DeFi, but also cause the fragmentation of ecology, users and liquidity. Layer2.finance realizes "in-place expansion" of the existing DeFi ecosystem. Since there is no need to migrate DeFi to Layer-2, while ensuring that the liquidity of the DeFi protocol is not fragmented, it can also ensure its composability. Layer2.finance allows multiple users to aggregate the completed fund allocation transactions through multiple accounts and multiple accounts into a single transaction through Layer-2 rollup. This aggregation process is implemented through an innovative generalized function call from Layer-2 to Layer-1, which can ensure the same security as Layer-1 without requiring additional trust in any centralized third party . Below we will introduce in detail the reasons for releasing layer2.finance, the specific solutions, and the launch plan.

The landing speed of DeFi is greatly decreasing, or it will become an exclusive game for giant whales

In the past two years, we have witnessed a double-hundred-fold increase in DeFi users and locked funds. Such an international financial system that is open and free and does not rely on third parties has the potential to make people excited: We all hope that DeFi will quickly start a rampant growth mode, and enter the homes of ordinary people from a cypherpunk utopia as soon as possible.

But when we carefully study the data on the chain, we find that the growth of DeFi has slowed down significantly recently. DeFi, an open utopia that should be non-discriminatory and borderless, allowing ordinary people to stand on the same starting line as large institutions and share the effective financial market for all mankind, is disintegrating into an aristocratic besieged city that ordinary people cannot enter.

< /p>

Figure 1: DeFi user growth slows down

As shown in Figure 1, after the rapid growth last summer, the overall growth rate of the DeFi ecosystem has slowed down, and the number of new daily users has dropped to around 2,000. And if we look at some "blue chip" and "entry-level" DeFi For projects, such as Maker, yEarn, and Curve, the slowdown in growth rate is more significant.

Another important indicator is the number of DeFi daily active addresses (DAU). To some extent, daily activity may be a more accurate indicator of DeFi activity, because it can better rule out the phenomenon of multiple accounts for one person, and can truly reflect how many people are actively using DeFi products every day. We can see from the figure that after the 10X growth last summer, DeFi DAU basically stopped growing significantly and stabilized at a level of 40,000.

So what is the concept of 40,000 DAU? Very little.

If we take "single", "on the Apple mobile phone system", "on the US market", "comprehensive If you look at the "ranking around 1500" games (note the addition of four restrictive attributives). The DAU of this game is more than twice the daily activity of DeFi in the world. If this trend is followed, in another year, the overall DeFi activity may not exceed that of a very ordinary mobile game.

No, wait a minute. DeFi users are not growing, which makes no sense at all! Even some of the most stable and reliable DeFi applications can achieve a risk-free annualized return of about 10%, and regardless of whether the interest rate on current deposits in the United States is 250 times higher than that of US demand deposits, it is not a problem to hang many private equity funds. Let's not talk about the older generation of users, but the young and middle-aged white-collar workers who were born in the Internet age, have some understanding of finance, and have higher incomes. Not to mention rushing up, there are tens of millions of users, and each person puts 10,000 yuan. Eight thousand to test the waters? It is true that there are objective problems such as blockchain understanding thresholds and user usage habits, but as the old saying goes, money can make xxx, and strong capital potential can of course level all thresholds.

So we can't help but ask: why? Why don't a large number of users each carry thousands of funds and pour into DeFi?

The high blockchain transaction fee is the culprit. Blockchain transactions treat transactions of different values equally as long as they have the same calculation and storage costs. But ironically, it is this kind of "non-discriminatory treatment" that has created a DeFi ecosystem that is extremely discriminatory against small-scale users. This leads to a 10% annualized return for "Xiaoyu" level users, which is a pipe dream.

Let's look at two examples:

Example 1:

- ; Xiaohong deposited 5,000 yuan to the exchange and wanted to play DeFi

- RMB 4,000 in ETH, and 4,000 RMB in USDC

- Xiaohong took both ETH and USDC into her wallet and spent 60 Block handling fee

- Xiaohong wants to deposit 4,000 USDC in Compound to earn 10% annualized income

- Xiaohong was surprised to find that the handling fee needs to pay 420 yuan

- Xiaohong is confused: In order to earn 400 yuan a year, I have to pay 420 yuan first. What kind of operation is this?

- Xiaohong gave up and continued to explore the DeFi world

Example 2:

- ; Xiaoli saved 50 million to the exchange and wants to play DeFi

- ETH was exchanged for 49.999 million USDC

- The ETH was taken into the wallet

- Xiaoli put 30 million in Compound, and the handling fee was exactly the same as Xiaohong’s: 420 Yuan. In less than two hours, Xiaoli recovered her 420 yuan transaction cost of handling fees

- the remaining 19.999 million , Xiaoli played high completely. After spending 60,000 yuan to try 50 different DeFi protocols, she found a strategy with low risk and a comprehensive return of 50%.

- Within a month, Xiaoli earned a net profit of 450,000, Xiaoli is very happy

We can clearly see that the "early majority" users we want to see are nakedly blocked by transaction fees in DeFi outside the door.

Thinking more carefully, although a user like Xiaohong cannot enter, a capital like Xiaoli Big crocodiles see high profits and low proportion of handling fees, and more and more will pour into the DeFi ecosystem, which further exacerbates the marginalization of small users like Xiaohong, and eventually forms a big A closed ecology monopolized by capital.

Culturally speaking, this is not cypherpunk, but realistically speaking, blockchain technology and applications cannot attract Users outside the circle are, after all, a major obstacle to the sustainable development of the entire ecology.

This problem needs to be solved urgently.

Hurry up and use Layer2 rollup to expand? calm!

Friends who have been paying attention to DeFi for a long time should have heard of Layer2 expansion. For POS side chain technologies like SKALE and Matic, Since they do not have Layer-1 security and are no different from building an Ethereum bridge with an independent chain, we will not include them in the scope of Layer-2 expansion.

Layer-2 rollup is a technical direction that has been popularized recently and widely recognized by the community (whether it is Optimistic Rollup or ZK Rollup). For the introduction of different rollup technologies, please refer to Vitalik’s article (https://vitalik.ca/general/2021/01/05/rollup.html). We will also publish a series of rollup Chinese articles in the future, so I won’t go into details here . In addition to using rollup technology as a separate DeFi application, the current popular opinion in the community is that there will be an EVM-compatible rollup expansion chain in the future, and the existing DeFi ecology can be redeployed and migrated to one (note that there are not multiple, Because multiple will destroy composability) on the rollup chain, then everyone can happily and quickly use DeFi applications. If so, what are you waiting for, hurry up!

It's a pity that the ideal is full and the reality is skinny. This technical direction is facing some fundamental challenges.

From a technical point of view first.

For a single-chain rollup architecture, computing and storage resources are still limited. In order to ensure the liveness of rollup (that is, to be able to generate blocks stably), it is necessary to keep synchronization among multiple block producing nodes. Once there is such a requirement for multi-node synchronization, no matter what block producer node selection rules and algorithms are used, to be honest, it is a very good result that a single rollup chain can increase the capacity by 50-100 times compared with the main chain. The current popular node selection point synchronization scheme is also similar to a DPoS solution. This roughly gives us a scaling cap of 2-4 million daily active addresses. Obviously, this is far from the large-scale landing of DeFi entering thousands of households as we hope.

Although the long-term is not enough, can we use this method to solve the urgent need first? Unfortunately, the current technical architecture of rollup requires the support of a lot of surrounding ecological infrastructure, and will have a profound impact on the security assumptions in the use process. If you want to migrate the DeFi ecology to a separate rollup chain, you must first solve a series of problems, as follows:

- How should users interact with the application when the chain is running normally and when the chain is paused and rolled back? What kind of status shall prevail?

- Are tokens of the same type treated equally?

- In the absence of support from surrounding ecological facilities, how to allow users to quickly enter and exit the Layer-2 rollup chain?

- Blockchain status query services (similar to Infura and Alchemy), now need to expand capacity by 100 times, and simultaneously access Layer-2 and Layer-1 perform complex verification and transaction forwarding. Is such a service ready?

- When the block node tries to stop the progress of the block, how does the client effectively notify the client and ensure that the user Switch to Layer-1 interactive mode for security?

- For those DeFi protocols that rely on oracles, are these oracles ready to resist possible Layer-2 bans? verbal attack?

- If there is no state synchronization delay assumption of the same magnitude as Layer-1, do most DeFi protocols need to re-evaluate their security sex?

There are many more questions like this. All these problems require a lot of time and research to find reasonable solutions and trade-offs before we throw 40 billion into rollup. But 10,000 years is too long to seize the day and night.

Even assuming all technical issues, there are good compromises, except for the fact that there is an upper limit for capacity expansion If the problem cannot be solved, is it all right?

Unfortunately this is not the case, because from the economic game model From a financial point of view, going down this path will inevitably lead to liquidity fragmentation and greatly reduce the market efficiency of DeFi.

As mentioned above, Layer-2 rollup expansion is not a free lunch. This expansion is exchanged for various "troubles". In system design, it is called "tradeoff". These trade-offs include more complex user interactions, delays in main chain-level security confirmation, more complex surrounding infrastructure, risk of block rollback and reconstruction, risk of banning attacks, and being front-running by transactions, etc. For small-scale users, these trade-offs are tolerable, because if you don’t endure these trade-offs, you can’t get the benefits of using DeFi, which is a comparison of 0 and 1.

But for giant whale users, compared to the huge amount of funds they hold, Layer-1 transaction procedures The fee itself is a very small part. The reduction of the fee based on this small percentage has limited benefits for giant whale users, but the potential troubles and security uncertainties brought about by the above trade-offs far outweigh the reduction. Benefits of handling fees. They will obviously not migrate to such a Layer-2 system, but will continue to stay in the Layer-1 DeFi ecosystem.

In summary, if we go down this road, even if various technical problems are solved, it is more reasonable However, we are still trapped in a DeFi ecosystem with an upper limit of expansion and fragmented liquidity. If we want to break through the upper limit of expansion of a single rollup, we must open multiple rollup chains with independent DeFi ecology, which will further lead to liquidity Fragmentation and overall market efficiency decline. It can be described as a dilemma.

We need a new solution.

Layer2.finance: Don’t migrate DeFi to Layer-2, bring Layer-2 to DeFi

In order to solve the above challenges and realize the wide application of the DeFi ecosystem, we use Layer-2 rollup in different ways and introduce Layer2 .finance. Similar to any rollup, users can deposit funds into the rollup chain of Layer2.finance by making a deposit transaction to the rollup contract on Layer-1.

However, unlike most popular approaches, Layer2.finance will not migrate the DeFi protocol itself to Layer2. Simply put, Layer2.finance builds a model similar to a decentralized intermediary. Users store funds on the Layer-2 chain, and through a special Layer-2 transaction (transaction), indicate which DeFi protocol their funds want to be placed in. On Layer2.finance's second-layer rollup chain, "transaction" is no longer "A sends money to B", or a direct smart contract function call, but a "fund allocation instruction", such as "I want to put ¥500 to the Compound protocol.” The rollup architecture of Layer2.finance implements a trustless "capital allocation instruction summary" function. For example, within 2 hours, there are 10 people who successively issued transactions of "I want to deposit X yuan to Compound", and Layer2.finance rollup will aggregate all these transactions and realize "two-layer to one-layer" After summing up the total fund allocation, a single call is made to the Compound contract to directly allocate the sum of all funds. Through such a mode of gathering sand into towers, the originally required 10 transactions are simplified into one transaction. When more people gather, the transaction fee shared by everyone will be less. In computer terms, the original O(N) operation is simplified to O(1) operation, which greatly improves efficiency and reduces cost.

Figure 2: Architecture of Layer2.finance

Next, we use an example to illustrate the architecture of Layer2.finance.

Generate Fund Allocation Scheme

Alice, Bob and Charlie (referred to as A, B and C) first put funds into the rollup fund pool contract, just like treating It is the same as various common rollups. They then start spending their funds on Layer2. Account and fund states on Layer2.finance rollup are not just address-to-value mappings, for each user any part of their funds can be in three different states: 1. Idle; 2. Committed; 3. Allocated .

"Committed" means that a user has issued a layer 2 transaction to express an intention to allocate funds to some DeFi strategies (which we’ll explain later) or withdraw funds from certain DeFi strategies. We need this "committed" state because there is a delay between when a newly aggregated funding allocation proposal is published on-chain and when it is actually executed. In the next section, we discuss the causes of latency and the tradeoffs of choosing between ZK and Optimistic rollup.

The user's "allocated" state funds appear as a certain amount of "allocated" received after performing fund allocation on the chain Liquidity Certificate". It is similar to the cToken or yToken concept used in Compound and yEarn. These liquidity certificates with a yield can later be used to redeem underlying liquidity and return the yield to idle funds. In Layer2.finance, we refer to these interest-bearing yield liquidity certificates as stToken (st: strategy strategy).

When users are deploying and withdrawing liquidity, both stToken and idle liquidity can be in the "committed" state.

Now, A, B and C start sending transactions to invest their idle liquidity in different strategies for yield ,As shown in Figure 3. They can also withdraw their liquidity from the protocol they have allocated. Block producers aggregate these transactions on the rollup blockchain. Whenever these small transactions are aggregated to a large enough amount, the transaction fee is relatively low, or after waiting for a certain period of time, the block producing node in the rollup will generate a block in the rollup chain. The block producing node not only publishes the state root of each user account state to layer 1, but also publishes the aggregated "fund allocation plan" of the entire pool to rebalance between different connected DeFi revenue protocols. In this particular example, the allocation of funds would be "Allocate $300 to Strategy 0 and $500 to Strategy 1".

The block node is a logical concept that will eventually become a decentralized Celer State Guardian Network (SGN), To ensure resistance to censorship and increase the security of the rollup chain.

From the generation of fund allocation plan to the real fund movement Latency: Choice between ZK and Optimistic Rollup

After generating blocks and hashing the state root After the distribution plan and funds are published on the chain, there will be a delay until the actual flow of funds occurs. It is worth noting that Layer2.finance has greatly expanded the simple information interface between the second layer and the first layer, that is, the simple state of depositing and withdrawing funds such as "who has how much money" has been extended to a more general " "Instruction interface" to express the fund allocation plan, that is, "how much funds are allocated to which agreements"". This is essentially a function call from layer 2 to layer 1.

When using Layer2.finance's Optimistic Rollup, the delay is to ensure that the total funds allocation proposal is finalized before funds are transferred challenge time window. Unlike withdrawals, there is no third party taking funds from the system, and due to the long delays during rollup and the simplicity of the challenge proof, the challenge period here can be shortened from weeks to hours.

When using ZK Rollup, the validity of the state root does not require a challenge period, so the delay in moving funds depends on the block duration .

In version 1.0 we chose to use Optimistic Rollup to simplify and speed up iteration. In the future we plan to migrate to ZK Rollup to reduce execution latency as the protocol matures.

Implementation of Fund Allocation Plan

When the delay is over, the distribution of funds will be performed on the chain. Instead of directly interacting with each different DeFi protocol, we define a common abstraction layer called "Strategy Adapter" (pluggable strategy strategy). They have a similar concept to yEarn's vaults. This layer serves three purposes:

- Mobility Certificate (stToken) abstraction. The strategy adapter tracks the basic value of the total deposit liquidity, and generates a certain amount of income certificate tokens (stToken), which are received by Layer2.finance users;

- ; Easy to extend. With the public interface defined for the Layer2.finance rollup contract, it is easier to extend the right side of the protocol to any protocol with yield;

- Flexible policy configuration. Strategy abstraction provides the flexibility to use different strategies (such as collateralization ratio) to the same yield protocol or even combine multiple DeFi protocols into one strategy.

We intend to cooperate with the developer community in the DeFi field such as yEarn, Curve, Aave, Compound, etc., to build adapters, and use More users access these products.

When the execution of fund allocation actually happens on the chain, the Layer-2 block producer layer will follow the Layer-1 transaction A Layer-2 is generated and a state update is made, adding each user's new stToken to the Layer-2 state (A, B, C in this case), and clearing those funds in the "committed" state. In this step, we use generalized reverse communication from layer 1 to layer 2 to achieve this.

The above is a complete cycle of the system, layer2. Move funds between corresponding different strategies.

The advantages of Layer2.finance as a DeFi expansion and speed-up solution

We summarize the main benefits of using Layer2.finance for a broad audience:

- Low cost. Since Layer2.finance pools small amounts of liquidity together, it is much less expensive to distribute funds and generate yield per user. Although there is a certain cost of data availability, the more people use it, the more it will actually reduce the cost of interacting with DeFi protocols for individual users. We estimate that it is entirely possible to reduce the cost by a factor of 1000;

- maintain composability, no liquid fragments, and No need to migrate DeFi protocols. Layer2.finance can expand the DeFi protocol in place, so it will not cause any liquidity dispersion, and will further improve the composability of the DeFi protocol;

- Stackable scalability. If a single Layer2.finance aggregation chain becomes too congested or the state becomes too large to maintain, a new chain can be added to connect to the same liquidity system in Layer-1 DeFi. Although it is limited by the capacity of the chain in the short term, when using ETH2.0 in the future, the upper limit of scalability will be increased almost infinitely.

- Flexible DeFi sharding. Besides scalability, there are many reasons to start multiple Layer2.finance chains, such as different batch configurations, different policy risk characteristics (whether high or low risk), etc. This effectively enables functional sharding in the DeFi ecosystem.

Limitations and trade-offs

Although we believe that Layer2.finance will bring DeFi into the world A solution at scale, but we also need to explain the limitations and trade-offs of this solution. At the same time, we can better discuss whether such a trade-off is "worth it".

- Delay due to aggregation. Whether we choose the ZK rollup method or the Optimistic rollup method. There is an inherent delay between when a user expresses an intent to make a fund distribution and when the fund distribution actually takes place. This is due to the Layer 2 aggregation that Layer2.finance relies on. However, we think this is a reasonable trade-off, otherwise these small liquidity users would not be able to benefit from DeFi at all. We can adjust this delay dynamically to find the right compromise between cost and reactivity.

- Limitations of advanced operations. Different from the original interface of the DeFi ecosystem, Layer2.finance provides users with abstracted strategies. Users cannot perform advanced customized combinations of multiple DeFi protocols. However, we don’t think this is a problem because advanced DeFi users are not the target audience for Layer2.finance.

As for other restrictions, the most obvious ones are transaction queue jumping and transaction ban. However, they are not unique to Layer2.finance, other rollup constructs have such problems as well. In fact, considering Layer2.finance's shardable overlay scalability, the severity of such problems is actually much smaller than that of traditional rollup.

Startup Plan

When will we launch Layer2.finance?

Our goal is to launch Layer2.finance 1.0 based on Optimistic Rollup in the first quarter of 2021 , of course it depends on the number of protocols we support and security audits. We plan to upgrade to ZK Rollup after the initial phase of iteration. We will support some of the most popular protocols such as Curve, yEarn, Compound and Aave in version 1.0.

Contract Economics

We will not issue new tokens. Layer2.finance is an ecosystem project that uses Celer's open source layer 2 technology stack. After the initial iteration, critical infrastructure for block producers will run on Celer's Stateful Guardian Network (SGN). You can become a verification node of SGN by staking CELR tokens. CELR will also be a governance token used to vote on the implementation, integration and configuration of new policies. In order to quickly increase the liquidity on the platform, we are likely to consider liquidity mining. Detailed economics and fee models are outside the scope of this article and will be described in more detail in a future article.

Hope that Layer2.finance can work with the community to help DeFi enter thousands of households as soon as possible .