This article was edited and compiled by Deep Tide TechFlow volunteer 0xz from a podcast of the Delphi Labs team during their vacation in Portugal. The guests were Jose Macedo, the head of Delphi Labs and the managing partner of Delphi venture, and Yan Liberman, the co-founder.They mainly introduced the birth of Delphi Labs, how to help Axie design the economic model, the story with Terra and the projects being incubated, etc...

The growth path of Delphi Labs may be worth studying and researching by many practitioners. Without a high starting point and a glamorous background, relying on the core competitiveness of economic model design to become the existence of T1 in the current industry, casting Axie's brilliance, being stuck in the Terra ecology, station In the upper reaches of VC, outside the first level of involutional war,After all, who would refuse the help of a master economic modeller?

1. Founded Delphi in 2018. After a long period of time, I didn't get any external capital, and I can only make a living through consulting business.

2. Greater Alpha added value comes from working with meaningful projects, not with other funds.

3. Most users of the project only think for themselves, and Token is a coordination mechanism that can balance the interests of most users and make the project develop more efficiently.

4. Axie's experience gave us a lot of confidence. This is the first time we found a project through research, then through consulting work and project cooperation, and finally grew together with the project.

5. It is inevitable for centralized institutions like CBC to regulate stablecoins, but it is very beneficial for stablecoin projects like LUNA.

6. The current total supply of USDT has reached 78 billion US dollars. Once a new reliable stable currency is born in the future, its ceiling will be very high. In other words, the stablecoin market is currently the largest market in the crypto space.

7. If the assets on the chain want to obtain sufficient credit endorsement, they need to continuously increase collateral. Not all collateral is good, and as the collateral reaches a certain level, this model will definitely appear in the center All the problems with stablecoins.

8. Another interesting point of Calculus is that it tries to create as many incentive mechanisms as possible to try to keep the value anchoring truly maintained. In essence, this is similar to a game for self-entertainment.

9. A good issuance strategy is a moat for the project. The whole significance of the Token issuance mechanism is to improve the ability to discover value.

10. The real innovation of encryption is to create an incentive system in a decentralized way, so that decentralized projects can provide the same level of service as centralized services, that is, to use decentralized methods to rebuild centralized world.

11. DAO is everyone votes on everything? I don't think so, I think governance needs to be specialized, i.e. need specialized committees or groups that are empowered to make decisions in a certain area and then report to the rest of the DAO.

12. We believe that everyone's on-chain identity will be based on personally owned NFTs, but to do this, they need to be combined through specific operations. Additionally, I think NFTs can also be an integral part of project loyalty metrics and retention strategies.

13. In many cases, it is very difficult for people to hold NFT for a long time. Although there are many solutions to solve this difficulty, such as empowering NFT and giving related discounts to people who hold NFT, there is a problem in these solutions. They all It is zero-sum. In this model, there will always be someone who sacrifices some of his interests.

secondary title

Birth of Delphi Labs

Jose：Contrary to what the outside world thinks, we're not that good, we're only brought together in the first place to survive.

Yan：Indeed, we started out as a consultant,I founded Delphi in 2018. After that, I did not get any external capital for a long time, and I can only make a living through consulting business.

The original intention of creating Delphi at that time was to use its own skills to set up a fund team to truly study the factors that drive the violent fluctuations in the value of Token. Later, we gradually understood that the greater added value of Alpha actually comes from cooperating with meaningful projects, rather than cooperating with other funds. Therefore, we began to provide consulting services for some projects that have been launched or have not yet launched, and provided them with a suitable Token economic model as much as possible to grow together with them.

Later, with the deepening of the company's business, we established Delphi Labs,The main work of the laboratory is to design a complete token economy system, or consider modifying some meaningful sub-modules in the token model.

In the beginning, the lab was always open to clients, because we made a profit from it, but as we scaled up, we could pick and choose clients and serve those projects that were really worthwhile. We choose to grow with those projects, for example, in the remuneration we choose,secondary title

Collaboration with Axie Infinity

Jose：Collaborating with those projects is exciting for us because we have to keep learning to progress with them. For some time in the past, we cooperated with Axie to design their Token model. Axie has achieved great success now, and we have learned a lot of valuable experience that can be used in the next design.

Yan：The cooperation with Axie is the first complete cooperation between us and the project, from excavation, to investigation, to participation, and finally growing together. We started paying attention to the game track at the end of 2018. At that time, there were many games, but most of them were fake marketing, which was bad, but Axie was very different.

For us, discovering Axie was really rewarding, but there wasn’t much we could do at the time, and there wasn’t any way to get involved in their projects, we had to stay in close contact with the team during the due diligence process, and we contacted every team members and talked to them to really understand what they wanted to do. So when they were trying to start thinking about launching a token in 2019, we jumped right in and helped them design it, and we've maintained that passion throughout 2020.

Yan：secondary title

What is Token Economics?

Jose：After the cooperation with Axie was completed, many people came to ask me how to design a good Token economic system. This problem needs to be analyzed in detail for specific projects. It is necessary to design the entire model from top to bottom according to the corresponding project, design the corresponding Token incentive mechanism, and of course design the corresponding security module. For some projects involving leverage, we also need to design better insurance mechanisms. And all of these designs require a thorough understanding of the project, at least knowing who is the supply side, who is the demand side, and when the project needs to reward whom.

Yan：Indeed, a thorough understanding of the project is very helpful to think about the positioning of the project Token. Looking back in 2017, Token was just a way to quickly raise funds without too many rules (ICO), which spawned a lot of bad bubbles, and also It directly triggered the bear market in 2018. In the bear market, people's perception of Token gradually tends to be different. Most people think that Token must depend on the specific value, and it can no longer be a pure speculative air, so there is the so-called fundamental valuation.

Jose：That's right, I think of designing the economic system for Token in 2017. At that time, cooperating with the project was like begging, and the Token economic system they required was always full of problems.

Yan：Don't do that again, it's exploiting users and very wrong.

Jose：Of course, this is not the case now, but at that time, project parties really only regarded Token as a financing tool, and they had no real willingness to do things after financing, which was very unhealthy.

Yan：Yes. But shortly after that, Token’s economic system began to gradually return to normal. At least many project parties have realized the importance of a good economic system. For example, if two projects of the same type, the project with a better Token design has a higher probability of winning , which is why the Token economic system is very valuable as a whole.

Of course, a deeper point of view is that most users of the project only think about themselves,Token is a coordination mechanism that can balance the interests of most users and make the project develop more efficiently.

Jose：If you want to learn more about the design process of the Token economic system, please simulate an environment in your mind: how everyone will act in the mechanism we designed. Is the behavior pattern behind group actions stable? Then try to find that unstable point, try to fix it, and then repeat the previous process many times until it is stable.

Yan：This kind of simulation is very helpful for the design. It is not limited to the simulation that has not happened, but more importantly, it is a review. You can review some meaningful projects and think about why they are successful? And why did it fail? Then pick the components of different projects and put them together, think about it together What will happen? Quite interesting.

Jose：More importantly, when we start designing the Token economic system, we must have a super-in-depth understanding of this project, we must understand everything about it, and understand how the project mechanism works? Who is the supplier? Who is the demand side? What is the growth plan? Use these to find out how we can help.

secondary title

From Consulting Business to Building

Jose：In addition, I also helped them design a new security balancer module, and helped them complete two hybrid studies, to be precise, design a new product. Working with these teams gives us the added bonus of a good name.

In fact, in the process of helping the project design, we gradually figured out what we wanted to do. We thought the only way to achieve these things was to start building the project ourselves early on, and the lab naturally progressed to the next stage.

Yan：secondary title

Stories with Terra Luna

Jose：Many of our current incubations are focused on Luna and Solana. The reason is that a year ago, when we started the lab, we identified Luna and Solana as two Layer 1s worthy of attention. Of course, one of our key areas of focus is stable coins. The development of stable coins can be traced back to a long time ago, including ESD and liquidity.

Yan：Yes, we're very interested in that, and we think it's a huge market.

Our idea when we published the report in late 2018-early 2019 was to provide something actionable. For example, what kind of Token is more valuable when the market accepts it? If you look back at our earlier reports, the content formats were very similar, partly because we had a clear value-add model. At that time, we realized the functions and huge market space of stablecoins, so we gradually began to study the economic models of stablecoin-related tokens and made relevant progress.

Jose：Remember that stablecoins were one of the topics we talked about when we first met offline in 2019,We all love its value capture. It is one of the symbols of the encryption economy, and it is also one of the few Tokens that can represent the fundamentals of encryption.But the problem is that the demand for credit is completely unrelated to the demand for stablecoins. It is very difficult to increase the supply of tokens for stablecoins and requires a growing demand.

Yan：To be precise,Stablecoins can be understood as raw materials. Assuming that the output product is credit, the raw materials required for the product to work are stablecoins.

Jose：It can also be understood as a branch of credit.

Yan：Yes. Some existing stablecoins do a good job of anchoring their value. But there are still some problems, such as USDC, in fact USDC can be frozen, it is not truly decentralized.

Jose：The worse problem is that as the scale of assets on the chain expands, many decentralized stablecoins will have all the problems of centralized collateral and stablecoins. At present, many assets that are not on the chain need to go through a centralized review to obtain credit endorsement.If the assets on the chain want to obtain enough credit endorsement, they need to continuously increase the collateral. Not all collaterals are good, and as the collateral reaches a certain level, this model will definitely appear centralized and stable All issues with coins.

Yan：You are right. In fact, the thunderstorms of some projects in the past two years have verified your judgment. After 2019, algorithmic stablecoins have become popular. I think the core idea of ​​Suanwen is to design a truly decentralized stablecoin that does not require any collateral.Another interesting point of calculating stability is that it tries to create as many incentive mechanisms as possible to try to keep the value anchor truly maintained. In essence, this is similar to a game for self-entertainment.

But this does not mean that he is wrong or meaningless. Personally, I have learned a lot from the calculation of stability, and gradually began to analyze the decisive factors in the stable currency economic model. For example, in the market maker model, supply is the product of leveraged demand, and further specific to ESD (Empty Set Dollar), we can find that it is difficult to maintain a balance between the two, that is to say, it is difficult to maintain stability.

Yan：The more common method is to expand supply and collateral through incentives. The more extreme measures are similar to the current unlimited QE of the US dollar. The price of this method is to lose control over supply and collateral, because you need to Provide a portion of the collateral in advance. Although this can indeed solve many problems to some extent, it requires close cooperation with decentralized community organizations, and I guess you have noticed that many times this cooperation is often relatively inefficient. This is another problem.

Jose：You explained it well, essentially this model is growth through the Ponzi model, even though they call it demand supply and liquidity instead. Stablecoins like ESD use a Ponzi model called liquidity mining to guide and maintain a stable Token supply and liquidity.

But this will cause a problem. In this model, the market maker is the liability of Token, and financial incentives are needed to guide the supply. But at a certain point, once the Ponzi incentive stops, there will be a large number of outstanding stablecoins. Demand, which can actually be understood as the liability of Token. From another perspective, the reason for this problem is that there is currently no non-speculative demand.

Jose：Once a problem occurs, the only possible next scenario is a death spiral. So we spent a lot of time designing the incentive model to maintain the liquidity of stablecoins. During the design process, one of the main conclusions we came to was that requirements matter.

Although we can design anything in the incentive mechanism, in the end there will always be some people who want to use Token as a stable currency, because we need the needs of these people. No matter what the next cycle is, no matter what the incentives are, stable demand is important for stablecoins.

After we have experienced the design process of LUNA and many other things, we are more convinced of the correctness of this conclusion. Although the design of the incentive mechanism of LUNA itself is very meaningful, what is more important is that LUNA tries to create a It succeeded to a certain extent because of the stable demand of the stablecoin UST, which has nothing to do with its incentive model.

Yan：Ideally, the pursuit of a high-quality economic model needs to be considered separately from the token, but if you just want the economic model to be stable, you only need to build an ecology that truly depends on this stable currency Token.

Jose：LUNA is actually such a model, so in our design this year, we imitated LUNA in many places. Of course, this matter must have been communicated with the project party and the foundation in advance.

In fact, there is one very important point about LUNA, that is, I think LUNA is not an experimental project. Because many people, including us, build on it, and we spend a lot of money building it and investing in it. But as far as the overall stablecoin sector is concerned, this entire sector is indeed an experiment, but because of the uncertainty, the upside is huge.

In fact, whether viewed from the inside or outside the circle, the stablecoin sector has a lot of heavyweight benefits. This has been discussed in our previous report, so I won’t talk too much today. But you only need to understand one thing, that is, the supervision of stablecoins by centralized institutions like CBC is inevitable, and such supervision is very beneficial to stablecoins like LUNA. In addition, although ALGO's stable currency is running stably, careful analysis shows that its risk of run and insolvency is very high.

Jose：Generally speaking, if this kind of decentralized stablecoin system can be truly established, there is another huge advantage, that is, stablecoin tokens with efficient capital utilization can resist risks to a great extent, which is the so-called counter-economic cycle sex.

To put it simply, when the cryptocurrency bear market first begins, the price of Luna Token will definitely drop, but during the bear market, Luna will be very stable and even behave like a bull market. Therefore, more and more teams now hold UST. As the demand for UST increases, the liquidity of Luna will increase accordingly. Therefore, LUNA will have obvious counter-cyclical characteristics, which is also the most attractive part of LUNA.

In fact, the development of the encrypted world is unlimited under the premise of having super-efficient capital, a completely decentralized review system and stable coins at the same time. These three conditions will greatly promote encryption. Taking stablecoins as an example, the current total supply of USDT has reached 78 billion U.S. dollars (updated),Once a new and reliable stablecoin is born in the future, its ceiling will be very high. In other words, the stablecoin market is currently the largest market in the crypto space.

Jose：At present, UST does not have enough influence, which makes us have to consider the potential downside risks of UST. So we started looking at the related ecosystem, and we saw some cool stuff.

Observing the ecosystem really helped us understand its potential more deeply, but we realized that some core early elements seemed to be missing, for example, many projects lacked a good AMM, and it was as bad as Terra swap, missing many similar Things based on traditional financial markets, such as anchor, anchor is a special financial tool that is optimized for specific transactions. And like related credit agreements, we thought it would be better to invest in these things than wait for it to happen, so we started the Mars and Astro ports.

Yan：But with the development of LUNA, this downside risk is getting smaller and smaller, because more and more platforms and products will be built on it, so that it will generate more demand. But what I'm more interested in seeing is whether LUNA will survive and thrive after the downside and the sell-off, which is important.After all, in a bull market, all projects will live well, and only those projects that have survived the sell-off are good projects.

secondary title

A Generalized Credit Agreement: Mars

Jose: Yes, we have a lot of ideas on how to improve financial markets and AMMs, and we chose to start with Mars.

Mars has two core innovations,The first is a new smart contract lending we designed, which is new mainly because it enables trustless lending, where no one else needs to know anything about the lender or ask them what they are going to do to repay the loan other than what is written in the smart contract.The second is tokenized collateral. As more and more collaterals are tokenized, more and more collaterals will be put into smart contracts with liquidation conditions, which allows the project to execute More and more different types of trustless loans.

Unsecured loans are one of the more important content in DeFi, but this is not a matter of relatively high capital utilization. So people are trying to make some modifications on top of that, like DAO credit history, identity scoring for Carmen and arc X, and other cool stuff.

From our point of view, there is another option, which is smart contract lending and tokenization of everything. As more things are tokenized, when people make smart contract loans, it is easy to write down what they need to do in the future. If there are tokenized collateral and liquidation conditions, we can basically be sure that the relevant loan is not required. trust.

If we think about how this market works from the perspective of the traditional financial market, we can find a big difference. For example, in the traditional financial market, our customers can be one-way borrowers, but here, our customers Only when you become a depositor can you become our borrower, which leads to the limitation of the customer base and demand side who can borrow money. Mars has made great improvements to this problem, that is, in the mine Chili introduced the Alpha strategy, which can be regarded as a major innovation in the history of DeFi development to a certain extent.

A simple summary is that the supplier directly enters the market to deposit funds. After the mining pool is opened, miners will come in to find leverage to expand their income, and they will borrow UST from the financial market and use the mined coins as collateral. For example, imagine a mirror-UST mining pool. People can enter the mining pool by depositing a mirror, for example, by depositing a mirror worth 100 US dollars to borrow UST worth 100 US dollars, so they are equivalent to 2 times leverage, that is Turn $100 into a nominal temporary $200. Of course, due to leverage, the risk will also increase, so the mining pool has a liquidation condition. If the mirror drops to a certain amount, the principal of the miner will be liquidated.

Of course, this also means that the liquidity pool is diluted, and the borrowed UST will first be used to repay the debt, and then the remaining debt will be repaid by selling mirrors. In other words, throughout the process, miners obtain corresponding delta income through leverage, and through high leverage, the overall capital circulation efficiency has also been further improved. One of the functions of Mars is to allow many assets to obtain higher leverage based on it.

The above is just a future use of Mars, and Mars can be expanded to aggregate more things, such as designing anchor mining pool strategies to change leverage, such as depositing a certain amount of LUNA into the anchor mining pool, and then using B Borrowing more LUNA in UST as collateral, the whole strategy is essentially a leverage.

secondary title

$MARSEconomics

Jose：One of the real keys to Mars is the Token economic system. After all, designing the Token economic system has always been our advantage.

Another real reason we like him is its design thinking on the security module. Maker is actually an overview of it, and I think once Token holders are making risky decisions that will ultimately affect third parties, then they will be affected accordingly. To put it simply, Token holders should understand the decisions they make, because Token holders have gained benefits, so they should also bear the adverse effects of their wrong decisions, so after Maker, they should also Add Mint-related content.

This also means that Token will be used to mitigate the impact of bad debts in the system, which is the security module.

Yan：Yes, bad decisions are common, most of the time it's Maker that is heavily influenced, plus, allowing some sort of very volatile collateral to be collateralized and not being able to judge the level of over-collateralization can lead to very bad decisions too. So we should enable a certain type of collateral based on the specific volatility, because once the Token is sold too fast and cannot recover, it will be in trouble.

Jose：It's the balance between capital efficiency and risk taking. It's like you want to be as capital efficient as possible and give people as much leverage as possible, but as leverage increases, even if your loan terms are very good, with huge fluctuations caused by wrong decisions, you will get It's a pile of bad debts.

So, what we want to do is to allow people to benefit from the system while also directly taking risks, which will make everyone responsible for their actions, thereby improving overall security, because not everyone is in the security module have voting rights and can obtain corresponding benefits. This allows those who have not pledged Tokens to still gain an influence similar to voting, and make their risks in the system smaller than those who use security modules and enter the Token cooling period, and they can be sold in advance To reduce losses, this is also an important mechanism we designed for Mars.

In Mars, you can pledge Mars to the pool, which is the X-Mars liquidity pool. You can think of it as the same thing as the X-sushi pool. Only a maximum of 30% of the Mars in the pool will be cut in the event of systemic bad debts, and only X-Mars can be voted on.

And a very important point is that when you only hold Mars and have no pledge, you will not have much risk in the system. Similarly, you will not get much profit, so you do not need to pay any fees. Also, people without stake will not get direct voting rights, as we want only those who are deeply tied to the system to be able to vote. This is very important in an economy like Mars where there is a line of credit, there is a lot of risk after all, so, for that matter, I recommend that you carefully review the smart contract that is giving you the line of credit.

Also, we need to make sure the collateral is of high quality. More precisely, it is to ensure that the risk parameters encoded into the collateral are of high quality.

That's why our lab spends a lot of time helping Mars set up the risk team and the risk framework for both, you know, we've never done anything like this before.

A pioneer of something must take huge risks. And this idea of ​​encoding risk parameters is super cool, it's a huge inspiration for us in financial markets, by looking at different assets, rating them according to their trading risk and economic risk, and then calculating the relevant risk parameters, so We can save double calculations in the mining pool, and only need to add the corresponding leverage.

In the mining pool, we need to figure out several questions, one is how much leverage can be extended to this asset? Second, how volatile is the asset? The third is how much liquidity does it have? Fourth, is there any counterparty risk? Fifth, how much utilization do we want it to have? Then we will aggregate these issues together as a reference for managers to make decisions.

Yan：You are right. In addition, I think that if you want to implement some functions on Mars, you really need a very powerful AMM. At present, there is no such thing in Terra, which directly leads to the incubation of Astro port.

In fact, we have always been interested in AMMs. First of all, the whole idea of ​​AMMs is very interesting. In addition, AMMS is very important for any kind of transaction, or DEXs are very important for all on-chain transactions.

Jose：I don't quite agree with this, for Mars, the first element is to increase leverage, especially for mining pools, because it needs enough liquidity to liquidate.

Yan：Yes, of course you are right. The increase in leverage is certainly important, but AMMs are also something we have spent a long time researching, and I think Dex is a very important part of our existing design. What our existing design lacks is to really connect with users Consistent Token ownership and governance design. There are many good and bad ideas about our current design, and we thought about it for a long time before we decided to let our Delphi Labs hatch the Astro port.

Jose：Before we officially start discussing the Astro port, I think there is one more thing worth discussing, and that is Mars' dynamic interest rate model, which is another major innovation of Mars.

Generally speaking, Mars has three major innovations, one is to provide a wider range of credit limits; the second is the Token economic system of Mars; the third is the dynamic interest rate model of Mars, which is relatively common in traditional financial markets Things, applying it in Mars, helps to ensure that the pool is developing in a positive direction, but at the same time it can avoid excessive utilization, thereby preventing the liquidity of the pool from being drained.

Because when the pool utilization rate is 100%, the pool is unable to flow. With the support of the dynamic interest rate model, it is possible to ensure that the entire system has sufficient solvency to avoid or reduce large-scale bad debts because it has sufficient assets, because most liquid assets cannot be withdrawn here. The dynamic interest rate will dynamically adjust the account according to the risk of the asset, and then through visualization, you will see a curve. When the asset utilization rate is relatively high, the interest rate will rise sharply, thereby encouraging people to repay their debts. At that time, the system The income will be reduced, and it will be difficult for new depositors to come in and borrow money from lenders.

At present, the model we designed works well, but there are some problems in the operation process, that is, it is not very sensitive to changes in external market conditions. For example, the pool utilization rate suddenly changes from 100% to very low. A quick way to calculate the right interest rate.

For example, when the alpha was released, the goal of the STC pool and other things was to achieve super high returns. It was often encountered that the utilization rate of Cream reached 100%, resulting in the real-time income of the alpha pool being almost 500%, which was very high, and people were even willing to bear up to 100% interest rate to borrow enough Cream, which leads to the asset utilization rate of the pool is almost close to 100%.

And what we're trying to do is to minimize that with a dynamic rate model where we use control theory and a PID controller, and we've written an additional report on it, which isn't here To be more specific, if you are interested, you can visit our website. We have made the model open source. The basic idea is to maintain an optimal utilization rate range. Because the interest rate is constantly changing in each time period, so in each area Blocks are targeted for optimal utilization.

secondary title

Astroport Overview and Economic Model Design

Yan：I think Astroport is a prerequisite for Mars to be successful. Of course, this is not the only reason that prompted us to incubate the Astroport project, there are many other factors.

We had the idea to incubate Astroport when Kyber swap first appeared in the early reports. In terms of the design and practicality of Kyber swap, it is very meaningful, and it is a very important foundation for any decentralized financial project. So we have been paying close attention to this field, and researched and designed many different Token models.

Of course, there have been many similar projects before, so we ideally want to create some products that have never appeared before and make the best products of its kind, such as Token economic design on top of the best DEX.

One of the biggest gaps among different DEXs currently lies in the consistency relationship between arbitrage users and corresponding Tokens. You said before that the main way for many DEXs to solve this problem is to appreciate Token in an indirect way, so as to maintain a good consistency relationship. But I think it's too exhausting.

Jose：No, I am referring to the kind of operation that is well controlled. It will not exhaust the water, but will make the ecology healthier.

Yan：I think Astro is a good opportunity for us to realize this idea. We often see a disconnect between the price and value of Token, high or low, which is cruel, but also very realistic.

Generally speaking, most platforms prefer to obtain sufficient liquidity by releasing a large number of Tokens. Although this will make the price of Tokens cruel, it does play a role in allowing Tokens to continue to serve as an incentive mechanism, and will eventually stabilize into a stable This mechanism is designed to allow a part of the fees to flow into Token stably, so as to maintain the relative stability of Token and form a relatively virtuous circle.

The platform issues tokens to stimulate additional liquidity. In theory, this will attract more trading volume, which creates the so-called flywheel working principle. The two elements of the flywheel are token value and token usage/incentives.But we think there is a better way for platform users to interact with the platform through Token. It should be noted that the better way we think is definitely not by disconnecting the connection between Token holders and liquidity pools. achieved.

Similar to the previous Sushi model, assuming that the pool charges a handling fee of 30bips, then 5bips will be deposited in the pool for repurchase, which is equivalent to only requiring a handling fee of 25bips. Our Astro model is basically consistent with this idea. A part of the transaction fee is deposited in the pool to obtain X-Astro liquidity. As time increases, this part of X-Astro is equivalent to a 50% increase in value, that is to say, an X-Astro Astro is equal to 1.5 Astro.

And your X-Astro will be worth more Astro, because there will be a steady stream of transaction fees for repurchase. This is a typical standard model, but ideally, you can choose another way of interaction, which is to lock up your part of X-Astro. Based on the number of locked positions, you will get a certain amount of VX-Astro, Just like the mechanism of Curve.

In addition, voting rights are allocated based on the shares of locked X-Astro, and each voting right is related to the number and time of locked X-Astro. This allows those who have more X-Astro and lock up for a longer period of time to have more voting rights, because they take the same risk. Of course, in the end our design established a transaction fee of 30 basis points, of which 5 basis points were used to repurchase Astro, 20 basis points went directly to the liquidity pool, and the other 5 basis points were used to motivate Astro’s voting rights.

To put it bluntly, the more transaction fees you pay, the more voting power you have, and vice versa.

What can voting rights be used for? First of all, as time goes by, there are more and more projects on Terra. If you are a Token holder of a certain project and you want to bring liquidity to this project, you will have many motivations to obtain this voting right.

In addition, similar to Curve, you can also obtain an additional reward due to inflation through enough voting rights, instead of just interacting with the pool. So if you are providing liquidity for Astro, you will have an incentive to hold Astro to get enough voting rights, because you want to get the rewards you should get as a liquidity provider.

In general, the tokens of all projects on Terra, X-Astro, and VX-Astro have enough motivation to think about getting enough voting weight.

Voting rights are really important, whether you want to vote or you want to get extra rewards, you need enough voting rights.

Jose：secondary title

Astroport's release strategy

Jose：The design of Token actually has many problems that need to be solved. As you said, the Token issuance strategy is indeed very important. It’s not because the project needs money, but a good issuance strategy is needed, and the shares of public offerings and private placements will be better allocated, which will make the distribution of project tokens better.Therefore, a good distribution strategy is a moat for the project.

This is another reason why the Curve model is so good, because of Curve's issuance strategy, LPs holders will eventually become Curve Token holders. Any good project side hopes to distribute Token to the right people. In addition, another issue related to the issuance strategy is,itemThe project actually needs enough float to withstand the selling pressure of chips and stimulate demand, because if the demand falls below a threshold, the project is likely to die.If the float of the project is insufficient, as time goes by, the liquidity pool and the market will slowly fade as the tokens are unlocked, which is really detrimental to community building.

Yan：That's right, if your Token design requires a lot of interaction with users, you will need enough floating and liquidity pools to promote the development of the community. Then we will enter the next stage. In this part, we need to figure out who is the ideal holder of Token.

Jose：I think the latter question you asked is the key. In my opinion, the ideal should be the provider of the liquidity pool holding LPs, but at the same time, I also want other users who do not provide liquidity to also become Token holders, but this will also have some problems. Although we know that there is a mechanism to alleviate this problem, which is to hand over part of the rights of Token distribution to liquidity providers or suppliers, and then wait for the emergence of market makers, but we have not done so, because if we do that If not, the fluctuation of the entire project will be very small, so the distribution of the project's Token will be very slow, so the development speed of the project will be greatly affected.

Jose：Another problem is that if we do not personally participate in the initialization of the AMM, the resulting liquidity will be very little. First of all, if there is no AMM, the price cannot be determined, so someone must start the AMM with a certain price, but generally speaking, the AMM started by others will definitely cause problems, which is a huge problem for the entire system. Disruption, and during that time, liquidity will be weak, so we avoid having someone else start an AMM. So sometimes, we can try to start it as an auction, and in order to avoid regulatory issues, we can put the proceeds of the auction into the DAO, making this look like a public offering.

Yan：But the risk of doing this is that we still fix the initial price under the assumption that a lot of value has been discovered, which will have some potential risks.

Jose：So in general, there are five things we want to do.The first thing is that we want to make the distribution of Token chips better. The second thing is that we want to allow the project to discover value, and we want to have some better ways to maintain the balance between supply and demand and set an initial price for AMM. The third thing is that we want to make the project liquid enough to absorb all the demand. The fourth thing is that we want the project to have liquidity from the beginning, so that there will be no turbulent periods of low liquidity. The last thing is we want to minimize regulatory issues.

Yan：You're right in generalizing that some of the issues you mentioned have generally been resolved in our second phase of the project's release. In addition, the first half of your question, in fact, in this project, the distribution method has been fixed in the form of a protocol, so I think it has been solved.

Jose：This is indeed the case. We are indeed very satisfied with the release mechanism of this project. In the release mechanism of this project, we have added a lot of things we imagined to solve the problems we encountered before.

Yan：Then, the second half of your question, in fact, I can understand how we can formulate an issuance strategy so that Token can be sent to LPs providers, that is, how to let more users migrate from Terra ecology to Astro above. The idea is that providers of LPs can deposit their funds in Astro for a period of time, and in return they will receive $Astro accordingly.

Jose：That's what we've been suggesting all along,Lock the funds, and then you will get rich rewards in the future.

Yan：And, in fact, when you lock up to obtain liquidity, it does not mean a short-sighted or mercenary decision. As time goes on, the liquidity you lock up and the value of the Token you get will increase. This is a very good ecology.

Jose：This period of time window can be as long as 1 week. During this period of time, anyone can deposit and lock positions and enjoy the benefits, but there may be some restrictions later.

Yan：Overall, everyone gets a prorated share of the rewards, depending on how long and how much they deposit. During this week, the next stage of project issuance will also be implemented, which is the distribution of Token chips. Our idea is that at this stage, about 2.5% of the tokens will be airdropped to depositors of $LUNA, and 75% of the tokens will be distributed to those who provide liquidity. The actual situation (low liquidity must be avoided), and 10% of Tokens are distributed in order to avoid supervision, and the remaining Tokens are the same as other normal uses of other projects.

Yan：In fact, this is very similar to LBP, but there are still many differences. The main difference lies in the initial chip distribution.

Jose：This is the case, or it can be said that we have added some hard coding on the basis of LBP.

Yan：Another difference is that in this project, we also need to invest a certain amount of capital, so we need to use it reasonably and avoid related legal issues.

Jose：So we try to find a way to decentralize the distribution as much as possible, so that it becomes a project initiated by the community.

Yan：Regarding the ideas related to community issuance, we believe that those who receive airdrops have the right to choose to sell, or continue to hold, or choose to lock up and expand their shares. Each choice is free. It is conceivable that for those who want to participate in the agreement, if he has an airdrop, he will definitely continue to deposit into Astro. If he does not have an airdrop, he will also find a way to obtain Token to participate in the governance of the agreement. For those who do not want to continue to participate in the agreement, even if he sells all of them, he will get a lot of income. So, generally speaking, getting an airdrop is a good choice.

Jose：Not really. While liquidity providers are definitely the core users of any project, the problem is that the community is made up of many different people. There are a lot of enthusiastic people in the community, people who discuss high-spiritedly, and people who write explanation videos. They want to join the agreement while protecting the development of the project. Therefore, let them participate in the project at an early stage like LPs providers The construction is very meaningful.

Yan：When the project is issued, the voting weight assets will also be considered, so the asset part with this voting weight is very valuable, because you can use it directly to get rewards directly. This kind of design appeared relatively early on the previous Mango. For those who provide liquidity, deposits in this kind of mining pool generally encounter the problem of unavoidable impermanent losses. That is to say, on the one hand, you You need to deposit Astro tokens. On the other hand, you also need to deposit UST equivalent to Astro. If there are 100 Astro and 200 UST, it means that each Astro is worth $2. If there are only 100 Astro and 100 UST A UST is only worth $1, so for those who provide liquidity, the risk is really high.

Yan：Therefore, in fact, the so-called price discovery depends on how much UST is deposited in the pool. This matter has also been verified on the previous mango. We can choose to deposit a large amount of liquidity to effectively improve value discovery. speed.

Yan：This is like an auction. The more UST deposited, in fact, the more profits for liquidity providers, but the more risks, so it is necessary to solve this problem through the second pool. Going back to the example of Mango, if you deposit a large amount of USDC, the more you deposit, the higher the price of Mango, which will increase the cost of other people participating in the agreement, resulting in only a few people playing in the end, because no one If you are willing to pay too high a premium, it will lead to fewer and fewer new deposits, which will eventually cause a liquidity crisis. Therefore, the buffer of the second pool is very necessary for the stable development of the entire liquidity.

Jose：Yes, we may even choose to actively drain liquidity if that happens.

Yan：In the beginning, people will be very willing to actively deposit USDC equivalent to Astro, but this will lead to an increase in the entire entry barrier, because in that case, the price of Astro will be higher than they want, and they It is hoped that Token will be sold. To some extent, this will cause panic, but the problem is that people who have already deposited will continue to leverage. They deposited X before, and now it may already be 5X, so in fact, the flow of the entire system Sex is very sufficient, and the panic of those who have missed the space is well eliminated, and they will enter one after another. After the project develops to a certain stage, we will encourage those people to withdraw a certain amount of liquidity, so that Token can maintain a better stable price, all of which can be understood as part of the design of the issuance mechanism.

Yan：This is how we think about the distribution mechanism. Many of our designs are generally thinking about how we should modify this mechanism if there is a problem, and then design a better mechanism until we cannot find more problems, or the problem cannot be solved. Taking the deposit mechanism as an example, the most basic problem faced is the problem of no one depositing and withdrawing deposits. This is actually a problem caused by panic in a game situation, because everyone will think that if he withdraws the deposit , What about my deposit, so we have to withdraw the deposit. Our solution to this problem is very simple. It is to make a binding at the beginning of the deposit and withdrawal behavior, and release the binding at the last block, so as to effectively alleviate this panicked liquidity extraction.

Jose：That is, you can deposit money in other blocks and withdraw money in the last block, thus getting a better price.

Yan：The whole meaning of this entire issuance mechanism is to improve the ability to discover value. However, generally speaking, this is difficult to do, because it is impossible to find an effective way to perceive the price changes in the market. In addition, the market price is actually very easy to be manipulated by humans. The idea with this is that people can control the amount of withdrawals over time. In the early days, the flexibility of deposits and withdrawals is very high, and then as time goes by, the amount of withdrawals will be limited, the amount of withdrawals will gradually decrease over time, and finally will gradually stabilize at a stable withdrawal amount, thus Reduce human manipulation of prices and improve market perception of prices. This will create a more effective value discovery model, and the entire lock-up period is about 3 months.

Jose：This means that all participants, including those who just want to obtain UST by staking Astro, need to go through a 3-month lock-up period. During this period, their locked-up rights will be reflected through LPs, so the project Gained great mobility.

Yan：Yes, if there are 100 Astros and 100 USTs in the pool, then the fair pricing of Astros is 1 UST, so if you deposit 10 Astros, your asset utilization is probably 5 Astros and 5 USTs, of course You can achieve a certain degree of income through this deposit method, but at the same time there will be some problems. For example, someone may transfer all the liquidity immediately after the project is fully issued, which will greatly affect the stability of the project. . Our solution to this problem is that we provide additional centers for those who want to participate in the liquidity pool, deposit Astro or UST, 1% of which will be allocated to these people, so, in this way, we It is equivalent to incentivizing people to lock and hold, thereby alleviating this problem.

Jose：So, generally speaking, my view on the second stage of the release is that it is the chip distribution stage following the first stage. The first stage is to distribute the chips to the users who actively participate in the agreement. The second stage is to let the chips carry out value discovery on the basis of the first stage, thereby increasing the value of the chips and increasing the bargaining power of the initial users to sell the chips , even in the process, the project itself does not benefit much. This is significantly different from public offerings, because the main purpose of this stage is essentially to increase the benefits of users who actively participate in the agreement. At this stage, the fluidity of the entire project has once again been greatly improved, which can solve all the problems we raised at the beginning.

Yan：Yes, in the second phase, the plan for depositing Astro and UST is a design that we think is very super practical. And we actually applied this model to Mars, because for Mars, the core point is the utilization rate of assets. In the first stage, the rewards people get will be calculated based on the Astro deposited in the pool and the time. Mars also has a lock-up mechanism similar to this one. Then for Mars, in the second stage, people will be encouraged to deposit UST as much as possible to get more rewards.

Yan：But it is undeniable that what is done on paper is always superficial, and we know that this matter must be practiced. No matter how we simulate, there will always be various unexpected omissions in our design, so we also need to constantly learn from the market experience.

Jose：secondary title

Other incubation projects

Jose：So far, we think our design is relatively good, but if you find any loopholes in our design, please contact us as soon as possible, we will give a very generous bounty, and we will provide it to you Corresponding Offer.

There are three other topics about the lab,One is other incubation projects of the lab, the other is the design of Governance 2.0, and the last is the long-term plan of the lab.

Jose：Currently, our lab is incubating some cool stuff on Terra.For example, Levana, which is actually similar to a leveraged Token agreement, is based on Mars, and its ultimate goal is to become a perpetual contract platform on Terra.It's really cool. And he is currently working closely with Astro and Astro V2. The role of Astro is to provide him with a centralized liquidity pool.

Jose：We're really excited about Levana, and its founder, Jonathan, is a really smart guy. I had a great chat with him at Levana's community meetup. In addition, he and I also participated in the hackathon on Terra and participated in the establishment of Ryan, which is similar to a cross-chain anchor. We are currently trying to make an angel agreement Saberra, if this can be successful, maybe we will soon be able to announce that future purchases can be settled with UST. It's really cool that our Delphi organization was the first to try something like this, and even if it fails this time, we'll keep trying things like this.

Yan：Yes, in fact, our ultimate goal is to enable Terra to be used and recognized as much as USD. We will continue to incubate good projects and promote the improvement of Terra's products, so that Terra's UST, no matter from investment From a perspective, or from a long-term perspective, a great consensus can be obtained, and this is our goal.

However, the risk of doing this is that we will inevitably try to produce some relatively failed products, which will affect the ecology of Terra, but we have indeed done a lot of things to avoid this risk, so the continuous product Growth and utilization improvements are really important.

Jose：Another project we are currently incubating is Everland on Solana. This is a unique project that can only appear on Solana.

secondary title

Delphi Labs Vision

Jose：Next, let's talk about our thoughts on the laboratory.We want the lab to be the best place for the brightest minds to come together and make the coolest stuff.Also, this lab of ours is really not interested in the rapid incubation model, such as incubating 200 projects next year or something like that.We prefer to work closely with super smart teams on projects that interest us.Our lab has a very good team of about 25 people, including development engineers, lawyers, and others. The goal of the lab is to make the projects we are interested in as successful as possible and to help them focus on what they want to do.

For us, the real innovation in crypto is that we create incentive systems in a decentralized way, so that decentralized projects can provide the same level of service as centralized services, that is,Use decentralized methods to rebuild the centralized world.

Jose：This means that incentives are very important, but it also means a new governance structure. We spent a lot of time researching the new governance structure. Inspired by Yearn Governance 2.0, I think all of us are a little skeptical of the idea of, is the DAO where everyone votes on everything? I do not think so,I think governance needs to be professionalized, i.e. need specialized committees or groups that are empowered to make decisions in a certain area and then report to the rest of the DAO. Based on this idea, we have designed many DAO governance systems.

Yan：You are right about one thing. Looking back, a lot of things come down to incentives. The research on incentives is actually a business that the laboratory is doing, and we also welcome people with ideas to join us to build together. Incidentally, retrospective incentives are very important wherever they are. Modeling such incentives has been elegantly designed in our lab and validated in real projects. Generally speaking, we are a small entity in the entire encryption circle, but we will try to help interested project parties with ideas to complete their ideas. We have some areas that we are very good at, and we hope that Collaborate with the smartest project builders to realize the greatest ideas possible. As I mentioned before, we will not choose to do 200 projects, we only need to seriously do a few interesting and interesting projects, which is a real win-win for the project party.

Jose：But this does not mean that we will abandon the projects that we have cooperated with or have not cooperated with, but the projects have been produced for a long time. Most of their main needs are the design of the Token economic system. This is in our laboratory It can be fully satisfied, and the design of the Token economic system in our laboratory can also be used to help many other project parties. In addition, we have a risk and quantitative team, they do the modeling work based on the project body. We also have a legal team that advises on issues such as governance being hacked and how to better comply with regulatory requirements for decentralization. Because every project looks different, we have a very professional development team working on this, plus, all of our code will be open source. We will also publicly provide the liquidity pools we can provide on Terra. We will also be open to our ideas about chip distribution design. Of course, we also prefer to cooperate with those project parties who have their own complete teams. We will support them from a very professional perspective and play a practical role in these fields to help them.

Jose：As far as openness is concerned, I think what we do is relatively good. For example, when you join the laboratory, you can not only access all laboratory resources, but you can even access all Delphi resources. As a project side, you can benefit from our research analysts, we have all expert level analysts in different areas of the transaction process, such as venture capital analysts. I think it's a very attractive value prop for builders. For us, the lab is like a studio and incubator, to be precise, like a venture capital studio and incubator. We don't intend to expand too quickly. We enjoy building projects, building projects that interest us, and we think we will definitely have a place in the crypto space.

Jose：That's what our lab wants to do now, for us,Our limitation is mainly whether we can meet great project founders or entrepreneurs.Summarize

Summarize

Jose：We are currently very interested in Layer2 projects, for example, we are very interested in Starkware. If you have an L2 related idea, please contact us.

Yan：We're really interested in Starkware. We are happy for you to go to build anywhere, Layer2 related things are some areas where we have spent more time recently.

Jose：At this time, I think the encryption field is at an inflection point, just like NFTs are at an inflection point of market changes. In addition, we can now see that the hot mainstream solutions in 2017 have become less competitive in today's market, so it is really an important turning point for the encryption field.

I think everyone at Delphi is very excited about NFTs. I'm as excited about NFTs as I was about cryptocurrencies in 2017, but I'm skeptical about most things in NFTs. I think we can use them in new ways in our projects, in an interesting composable way. IWe believe that each person's on-chain identity will be based on personally owned NFTs, but to do this, they need to be combined through specific operations. Additionally, I think NFTs can also be an integral part of project loyalty metrics and retention strategies.

Yan：In addition, NFT actually has a lot of sticky communities, so we can use NFT to achieve specific influence. In many cases, it is very difficult for people to hold NFT for a long time. Although there are many solutions to solve this difficulty, such as empowering NFT and giving related discounts to people who hold NFT, there is a problem in these solutions, they are all zero And yes, in this model, there will always be someone who sacrifices some of his interests. But by associating NFT with the personal identity of the project, it can be empowered in a positive and positive way, thereby maintaining a good ecology. In general, this will be an identity on the chain, and it does not require your name or your credit score.

Jose：In fact, this is an auxiliary tool of another Token economic system similar to the incentive mechanism. I think NFTs are a really important tool, and we're now thinking about how we can use it in really cool ways to incentivize the right behavior that we want in these projects.