"What We're Building is Very Big and We Should Have the Least Concentration Possible:" Aave's Stani Kulechov

This week’s episode is with Stani Kulechov the founder and CEO of Aave. When we spoke Aave was the second-largest DeFi protocol, but in the days after our conversation, it rose to overtake MakerDAO and become the DeFi platform with the most assets held in its smart contracts, at over $1B in digital assets.

It’s been an exponential rise for Aave, which launched just eight months ago and in that short time was able to climb the famous Defi Pulse Total Value Locked ranking to the top. It’s done so pushing DeFi forward by driving innovation in interest-generating tokens, flash loans and delegated credit lines, which are all lowering the barriers of entry to users from all corners of the world to access financial applications.

We also get into all the recent yield farming craziness, which he wishes was more restrained. Stani believes projects should build communities first, which can help fund audits, before launching protocols into the wild and risk losing users’ funds. There’s no trade-off between safer launches and fast growth; he believes DeFi can attract an even larger amount of users and funds by providing safer applications. If it doesn’t, it will end up losing credibility and users’ trust.

You’re a paid subscriber, which means you get the full transcript below. Subscribers also get exclusive access to The Defiant’s Discord chat for the community, here’s a new link to join.

🎙Listen to the interview in this week’s podcast episode here:

🙌 Together with Zapper, the ultimate hub for managing DeFi assets & liabilities.

Camila Russo:Before we get into all the latest developments, I'd love to get more of the background on how Aave came to be. Because before Aave, there was ETHLend and the ICO and going through all the roughness of the bear market. Can you find provide that background on the period before Aave?

Stani Kulechov: Definitely. As a protocol, we started Aave this year in January. We rebranded to Aave, but before that, we used to go by the name of ETHLend, which is short for Ethereum lending. What Ethereum lending was back then was just a small project that I founded while I was studying law in the University of Helsinki, Finland where I’m originated from. What actually happened is that we started as a community project and we wanted to create, the first lending protocol on Ethereum and test how we could create a bit more complex financial transaction than trading or transferring funds from one address to another.

One thing led to another, we had an interesting market back then and this was ages before stablecoins or liquidity pools or an idea of total locked value. I think, the transition from ETHLend to Aave was natural, as we saw liquidity starting to pull together and we started to think how we could utilize more of this locked liquidity, locked value in a more capital-efficient way. And that is how we came up with the Aave protocol.

CR: Because was ETHLend more of like a peer to peer lending structure than a pool liquidity structure?

SK: We started back then with the peer to peer model, because we heavily believed and still believe in the idea of decentralized finance, and for us, it was just financing in terms of smart contracts, because the term wasn't coined back then yet. But the idea is that you could actually democratize a bit interaction, so anyone can build permissionslessly and participate permissionslessly into the ecosystem. That is what's very fascinating.

We choose the peer to peer model also, because we try to separate risks as much as we could. Back then, the tokens that were on Ethereum, for example, didn't have that much liquidity, so we didn't want to go with a systemic model. Now, we see many models, in terms of liquidity pooling, whether it's trading or lending, that is actually a functional model if you have incentives network for liquidations, and these incentives that will ensure that your protocol will be healthy. That's why we transferred from one model to another as we saw it coming, the thing to do.

CR: Got it. And I thought, when you first launched, it was interesting to see that you were innovating right off with the A tokens, building from what Compound had done with C tokens, but modifying that structure a little bit. Can you explain, when you launched Aave, what were the main things that you were trying to innovate with at the time?

Interest-Gaining Tokens

SK: I'm a really big fan of the C tokens. What I wanted to do with our team is that with A tokens we wanted to somehow make interest-bearing tokens in a way that you could actually understand how much you're earning in terms of what you're depositing. Our goal was just to focus more on the end-user experience. For example, if I deposit 100 DAI into the Aave protocol and I get 100 A DAI as an interest-bearing token in return. We wanted to have this one-on-one ratio so that you could actually understand how much is deposited, but also that whatever you're earning in the protocol, it actually grows with the balance.

This means that we integrate we added into the A token smart contracts and an algorithm that is based on interest rate model of the protocol, which updates the balances of all of the A token holders. You will constantly see if you're holding A token Metamask or in some other basic wallet, you will see the balance all the time increasing.

Why it's important is that you actually know how much you're earning and also, it gives you access to a global permissionless savings account, which is dollar-denominated at the moment with the DAI, USDC and USDT A token versions. That's very important in places, for example, like Venezuela, other countries in Latin America. Argentina is a good example where you could just by holding A tokens, by depositing or by buying them, you hold this permissionless access to savings account this way. It's quite interesting, especially if your local currency has high inflation. I think in one way, it really democratizes and shows what you could actually do with DeFi with just this small interesting function.

“… it really democratizes and shows what you could actually do with DeFi with just this small interesting function.”

CR: I agree. For those listening or reading who are not that familiar with these types of tokens, it's a super interesting concept, which is representing a lending deposit or tokens deposited in a lending protocol with tokens in your Metamask wallet or in your Ethereum wallet. It’s a derivative of your token deposits. But what's interesting is that the token itself starts growing in value as you start gaining interest, so that's why they're called interest-gaining tokens. It's really cool that it does democratize access to a dollar savings account for anyone in the world. Aave and Compound are two main lending protocols offering these types of tokens.

From launch early in the year, value locked in Aave has completely skyrocketed. What do you think are the main milestones in this trajectory?

Image source: DeFi Pulse

Flash Loans

SK: We weren't expecting this traction at all. I mean, DeFi in general has been growing quite a lot above my expectations. I always projected DeFi to grow a lot based on things people have been building, what Aave has been building, what others have been building. What I see now is that we get more and more adoption. It's just funny because we're still in the very, very beginning, in the sense that there's many things that can be still built. We will probably see quite interesting DeFi primitives on top.

“We're still in the very, very beginning, in the sense that there's many things that can be still built. We will probably see quite interesting DeFi primitives on top.”

The biggest milestone for us was to see flash loans to get utilized, because in the beginning, we started to introduce the functionality to developers that, hey, you could build very interesting DeFi compressibility things which means, you take a couple of DeFi building blocks. Let's say you use Aave, you use Compound and Uniswap and you make a product and you use flash loans to reduce the need of actually capital to do a transaction.

We saw a few months ago that actually flash loans started to get a lot of traction. Now on a daily basis, we have quite a lot of flash loans and in some days, we have over $100 million worth of flash loans. That for me is very nice to see. Other things that have been substantially important to our team is the amount of developers that are building on top of Aave or are utilizing Aave and I think that's pretty cool stuff now. We’re trying to just innovate as much as we can, create new products and services and try to just make the DeFi experience as seamless as possible and get most out of it for the users

“Some days, we have over $100 million worth of flash loans.”

stani.eth 👻 @StaniKulechovHoly fuck, someone did 14 mm Dai flash loan, biggest ever on @AaveAave 🤯🤯🤯

sakulstra @sakulstraWith a slight delay a new version of https://t.co/cYPQpKRzA8 is available improving charts toggle functionality and re-rendering behavior ✨ @JakobSieversDK @SirSoth91 Btw. today @AaveAave broke another flash-record: https://t.co/NTBwNei7rp 14M $DAI in a single transaction ⚡️🤯 https://t.co/rY6nXkBHjF3:57 PM ∙ Aug 29, 2020165Likes24Retweets

CR: I wanted to expand a bit on the flash loans concept, which again, for those listening, it's a really interesting idea or ability for DeFi users to take out a loan on one Ethereum transaction, use that money borrowed to execute any type of trade in any other DeFi protocol outside of the platform that you took out the loan to begin with, and then pay back the loan on the same transaction, potentially with the profits that you made from whatever trade you're doing.

This needs to be automated and requires some level of skill, because you need to code these trades up. But it does allow anyone to participate in the system without having that much capital to start out with, because you're executing these trades with money that's borrowed. You get to keep those profits as long as you pay the loan back in the same Ethereum transaction. Let me know, Stani, if that was a good explanation for it.

SK: Totally. I think that was exactly, very good explanation. Flash loans are a functionality that mostly now maybe developers are using. There are ways to use it with no code functionality. For example, there's an app called Furucombo that does that you can actually just drag and drop and make functionality there it meant.

It's all about developers building new products for end-users to make them capital-efficient transactions in Ethereum. I think that's the key point where we are with flash loans. Flash loans are just a functionality, but it's part of the bigger picture of Aave protocol, utilizing capital as efficiently as we can with innovation.

“It's all about developers building new products for end-users to make them capital-efficient transactions in Ethereum.”

One of the other things amongst flash loans and A tokens is credit delegation that we announced a while back ago. That allows anyone who deposits into Aave to earn interest to delegate their credit line to someone they trust or to a smart contract that does some functionality, some strategies that are pre-programmed, there's less risk of failure.

This is the idea of average here, we calculated that every fourth user always depositing into Aave or let's say from the capital, 20% is somewhat utilized and the rest, 75% is available liquidity for flash loans or non-utilized. The credit delegation was a way of increasing that utilization rate by allowing other people that don't have collaterals but are in trust relationship or building some smart contract functionality to actually draw that credit on behalf of others. For depositors that are actually delegating credit, for them, it's pretty cool because they can earn additional yield and that's very interesting.

CR: That's very cool. I thought this was a really innovative concept and the way to get around this dilemma in DeFi, which is that up until now, you need a lot of capital to borrow, all of the loans in DeFi are over-collateralized. The reason why that happens is because for the system to be permissionless, you can't be checking somebody’s credit score or doing KYC or demanding like salary on the income statement or anything like that, it just needs to depend on the money you have. That's limited the number of users that can actually use DeFi because you're requiring to have money up front.

But this is a way to get around that; by delegating. If you have capital, you can delegate that collateral, so someone else can take on an unsecured loan. What I don't get is, how does the person who is delegating their collateral know how trustworthy the borrower is?

Credit Delegation

SK: That's a good question, because it boils down to the actual lending facility or lending model where you are entering into trust relationship. There’s an interesting scenario because there’re different kinds of trust relationships. People lend money, for example, for their family members, friends, so there's some trust. There’re others that are doing the same, but they're lending out to, let's say, consumers, businesses based on the risk borrowing.

As a function the credit delegation idea is to open the floodgates to delegate that credit and source DeFi liquidity out of protocols like Aave. But end of the day, about the underwriting, it's an opportunity for different kinds of protocol models or business models to create functionality where they can do some sort of creative risk assessment, for example.

I think, three very easy scenarios are where you're delegating credit lines to someone you can trust. You can even do that with just a verbal agreement, formal agreement with Open Law that we used, for example, with the very first delegation to diversify. The second way is you can use the very same thing. But you could, for example, a group of persons or delegators, could delegate to an entity that could be in the traditional finance, so they could delegate them to borrow stablecoins that they convert into fiat and then they lend it out, into their clients.

I could see that in the future, for example, our objective will be fulfilled, if someone in the street who bought a car or who bought a house, partially of that purchasing power is sourced from DeFi with credit delegation. It doesn't like the underwriting, it's one own form of business model. There is a project called UnionTello who are actually working on this stuff where, they're trying to bring efficiency in terms of underwriting loans in the decentralized finance and utilizing them. But I think we will see like a lot of various differences.

“Our objective will be fulfilled, if someone in the street who bought a car or who bought a house, partially of that purchasing power is sourced from DeFi with credit delegation.”

The third one, which is actually even more fascinating to me is that you could actually delegate credit to a smart contract that does some functionality. For example, we have a protocol called yEarn and one of their vaults, so they have vaults that are doing some sort of strategy, trading, yield farming, and one of the vaults is based on A lend token, which is delegated token of the A tokens to that vaults and that wallet draws credit line with the delegation from Aave. And that's a good example to have, I think, like farming activities.

CR: I'm not sure I got it, but first, let's go over the other examples. In the first case, which is kind of the simpler case is that you're delegating your borrowing power to people you know, like your friends and family. Then there's the second case, which I think is a way to scale up this operation is for entities to delegate their borrowing power to their own clients. So, for example, fintechs in traditional finance, who already do KYC on their customers and know that they are creditworthy, they can delegate their lending power and allow those customers, to borrow through Aave?

SK: Exactly.

CR: Then the third option would be for smart contracts execute that loan. You made this example with vaults in yEarn, but I didn't get that part of the example.

Smart Contract Borrowers

SK: What's interesting about this is delegating to a smart contract, you could delegate to a smart contract that can only do predefined functions. Let's say you have yEarn, you're delegating your credit to the yEarn vault and what happens there is that you give the permission of that vault to draw a credit line against your deposits into Aave.

When the vault functionality is called, it draws credit from Aave and deposits those funds into for example, Curve to yield farm and a smart contract can't do anything else, so that's the beauty of it. Because it's a predefined strategy, it's immutable and you're reutilizing the capital that is looked in Aave, in functions that can't be changed and eliminates the credit risk there. It's different from lending to people that you trust or under legal agreements. Because actually, the more you create different functionalities in smart contracts that are defined, the more DeFi grows and these opportunities will come, we might see more into this direction.

I really love when we have smart contract-based infrastructure and we don't need to go to, let's say, legal agreements and enforcement. For me, it's very fascinating what's happening on the smart contract side.

“… the more you create different functionalities in smart contracts that are defined, the more DeFi grows and these opportunities will come, we might see more scaling into this direction.”

CR: These loans would be used to execute different trading strategies by smart contracts. That's really interesting. Are there any examples of this happening yet? Or is it just something that you think would be a good way to use this functionality?

SK: Yes, smart contract-based credit delegation is already utilized by yEarn. So, they're drawing credit lines from Aave and depositing into different protocols, stablecoins to generate more yield and that's already happening. I will bet that this will be expanding in the future as well. I think I made the first proposal, I think was yesterday into the yEarn community where actually we could create more of these vaults and they expand the credit delegation in those strategies. It’s pretty interesting to see how it will develop.

CR: How much would you say right now is being borrowed through this mechanism in general, taking all these different strategies?

SK: I think the amounts aren't that big at the moment, I think we have, I’m not sure quite the number, because I haven't checked how much the utilization is there. But I would expect somewhere below one million still. It might be more, it's really, I haven’t checked how utilized it is, but it's interesting to actually follow how it's going to develop.

CR: As with so many things in DeFi, these things tend to grow exponentially and just pick up from one day to the next.

SK: We always tend to focus on the permissionless aspect of the users. But what's happening in the permissionless aspect in DeFi in terms of development, that's even more interesting because anyone can develop these products. I could imagine the future in Aave governance that someone could propose these vaults and there could be a credit delegation functionality that could be done in the user interface to choose which vault you want to delegate and what yields they give you. I could imagine the Aave governance voting on what vaults can be approved into the system, depending on their security, depending on their risk assessment and so forth.

CR: Speaking of governance, I wanted to also speak about Aave v2 and transition to more decentralized governance. Would love for you to give an overview of this upgrade of the protocol.

High Stakes

SK: Now what's interesting is the extra narrative behind, because when we launched in January and we saw the protocol growing over time, and currently, we have roughly $1.4 billion of locked value into smart contracts and the market size is even bigger, and so forth. We see the stakes are very high and we feel that as a team, we cannot anymore govern the protocol in the sense that we have at the base layer of the infrastructure.

“We see the stakes are very high and we feel that as a team, we cannot anymore govern the protocol.”

We have a base layer of infrastructure now, and we're updating it to version 2, where we are adding more optimizations in terms of gas and more functionality. But the underlying idea there is that now it's the time that we give the power of making changes to the protocol and vault on the features and changes to the other governance and this is why we're having a token migration from Lend to Aave and after that happens and after we've transferred keys to Aave governance, it's up to the governance to decide Aave what the team will be building and what other developers could build.

In the future, anyone could build features and functionality into the core Aave in the next upcoming versions or additional features on top and that's what will make it very interesting. And of course, anyone can propose updates. I think we tend to still underestimate the power of governance, because the more we give the decision making for the governance, the more we have expertise and the more we have vision, on how the projects would be developed further.

This is what got me into DeFi and Ethereum in the first place. Because it's not about few bankers deciding in a cafeteria in London on how the company should be run, but actually now anyone can participate. That also means that you need to build a product or a protocol that actually is suitable for all parts of the words and demographic. That is very fascinating for me.

“We tend to still underestimate the power of the governance, because the more we give the decision-making for the governance, the more we have expertise and the more we have vision.”

CR: I agree. In this new governance system, will it be a DAO and token-based voting? For instance, will Aave token holders propose changes to the protocol and vote on those changes through a DAO?

Aave Safety Net

SK: Exactly. The most key thing about Aave protocol is that Aave token holders are voting on decisions and most of the decisions, because of the value locked in the smart contracts are based on risks. They're already holding the risk of the locked funds there and the government activity and they're also voting on the risk parameters and what assets can be added into the protocol. I could imagine there will be in the future decentralized risk teams giving proposals on how things can be changed.

When the Aave token holders are deciding upon this risk, they're also at the same time bearing the risk, which means that in our Aavenomics, the token economics of the system, what we have is a functionality of staking where the token holders can stake their tokens into a safety module and if any sort of a shortfall event such as smart contract, hack or a bug or any failed liquidation or the governance sees that this thing led to a deficit in the protocol. Those staked tokens can be used to cover the deficit and if it’s not enough, there is a minting facility as well over a deficit.

“When the Aave token holders are deciding upon this risk, they're also at the same time bearing the risk.”

What's important here is that the Aave token holders are setting those risk parameters and at the same time, they're bearing that risk as well with providing the safety nets. The key component here is that, over the past six months where we were designing the Aavenomics, we tried to think of, ‘what is the most important thing for DeFi or for Aave?’ One of the things that ring the bell all the time was, we have to ensure the safety of the infrastructure of the whole ecosystem and our protocol. That is how we get adoption when people can trust that the code has been verified and it has been diligently deployed. But also, if there's some sort of a shortfall event and something bad happens, it could be somehow covered and life can continue. That was like our idea on how we see risk playing a very major role in governance.

CR: That's interesting. For Aavenomics, I guess the key difference with maybe other protocols is that you have this token reserve that's made by token holders staking their tokens and this reserve is there in case anything bad happens and you need to cover maybe a shortfall in collateral, like what happened with Maker in March?

SK: Exactly. What's interesting, during the past six months, we took a lot of inspiration from Maker. We took inspiration from different protocols, and the back-stop syndicate. The idea of that was to buy Maker tokens that were minted and building that system as well. Also, we're looking at other staking DAOs.

One of the biggest interesting inspiration for us was Ethereum, because in essence ether is transaction gas, but also it's a way to secure the network. Miners for confirming transactions, they're upholding the security for the Ethereum network and they're getting rewards via inflation and also the transaction costs. We saw that this element, rather than focusing completely on rewarding liquidity providers for making the market, we need to also ensure that those liquidity providers are safe and we have modules that incentivizes them to deposit and we can ensure that if there is an event as such, then their liquidations doesn't create a downside effect where people are leaving DeFi because of there's just too much risk. The risk is very high, because all of the code that is developed, it's public there. It's really important functionality we understood.

CR: Can you talk about the token distribution and how much of the tokens as a percentage will be held by Aave team and Aave early investors?

Aave Token Ownership

SK: I think in the current sense, maybe the Aave team and early investors probably have one-fifth of the token supply, which is very low, but it has quite a good perspective in the sense that the token holder base is quite decentralized already. Because we see a lot of projects when they start their DeFi project, it starts in a way where the early investors have most of the supply and gradually with different kinds of liquidity incentive schemes, the token sort of going from, let's say, the founding teams to the users, at least in terms of percentage.

What's interesting about the incentives is that we have actually a very interesting system in a way that part of the incentives are for the stakers who are providing the safety and part are for the liquidity providers who are providing the liquidity. Now, we have our governance basic and open discussion, what how to balance between the liquidity providers and stakeholders and, what's the ratio should be to incentivize the safety nets, but at the same time to bring long-term liquidity providers. This is an interesting discussion that we have been discussing.

We see that it's a bit like our community is very oriented into the safety nets, but what's cool about the Aavenomics is that if you provide liquidity and get rewards, you can compound them into the safety protocol. If you are incentivized to do that, you might be a liquidity provider that cares more about the protocols health than, for example, a liquidity provider that farms let's say, tokens to just to sell them quickly.

CR: What happens there is liquidity providers can then use those rewards to add staking to provide the safety net for Aave and so maybe that way, both incentives are aligned, liquidity providers and the safety of Aave?

Aligned Incentives

SK: Exactly. Our goal is, when a liquidity provider steps into the Aave protocol, they are incentivized enough to keep the healthiness of the protocol in mind and see the protocol as a long-term investment for them. I think that's important, because when we have different kinds of governance vaults, it will be very difficult if on one side, we have liquidity providers that they want quick returns, but on the other side, we have people who want to provide safety nets. In this way, you have incentives aligned and they're all towards providing more safety.

“Our goal is, when a liquidity provider steps into the Aave protocol, they are incentivized enough to keep the healthiness of the protocol in mind and see the protocol as a long-term investment for them.”

CR: But right now, how much each group will get of Aave tokens is still in discussion?

SK: Yes. Whatever the community decides, we will implement it. Interesting part is that those incentives can always change, the governance rewards again. I think that recently for the past six months or even the recent months, we've seen governance working pretty well in different projects and creating proposals and looking in general long-term all the protocols that, this decision-making that I have seen in DeFi in terms of governance has been really, really impressive.

CR: I agree. These tokens have certainly launched this really active community and incentivized them to really become involved in the protocols that they're using and propose different kinds of changes and participate in voting. I was really surprised to see something like the YAM community really spring into action to save this token which was born overnight. I think that was a really good example of decentralized governance.

But at the same time, we've been starting to see the temptation that, decentralized governance can suddenly become concentrated again. I'm thinking about Curve Finance and how with the incentive to lock up CRV token, the team held 70% of voting power because they were the first to lock in their tokens.

Interested in your thoughts on this, how to balance keeping a governance system decentralized with the obvious incentive for the team to also keep some level of control of the protocol, especially when stakes are so high with hundreds of millions of dollars lockes or billions of dollars in your case?

Least Concentrated as Possible

SK: That's a very good question. Because in terms of the beginning, it's very flexible if the team has power to make changes. Especially when you need to do rapid fixes or just decision making where action is needed you need flexibility. But over time, when the protocol starts to grow and there's more and more assets and funds involved that are from external parties and from the community, I think it has to start already thinking how to become more decentralized.

One of the best learnings I have learned in the past years is the more you give power to the governance, the healthier and stronger your governance is. The earlier you can do it, the better. There's no reason for the team to keep majority control of a protocol. If the protocol code is safe enough and battle tested after few months and so forth, it is fair to already start transferring the power as much as possible.

“One of the best learnings I have learned in the past years is the more you give power to the governance, the healthier and stronger your governance is. The earlier you can do it, the better. There's no reason for the team to keep majority control of a protocol.”

The way I have seen only good thing happens when well, can’t say only good things, I mean, depends on the communities. But I've seen very positive effects when you do that. Adding that as early as possible is very important. Our plan was started even earlier, but we were in a very small team and we wanted to make sure that when we are shipping code or we're doing changes, we do them very diligently and we put a lot of effort in there. Also, it's a question of development. Now, that the Aave protocol is holding a vast amount of funds, our development process is a bit slower in the sense that we are innovating as fast as we can, but we also need to be very diligent when we're deploying things.

New protocols that are launching new products and token economics, I think they should really think about it, in the very beginning, how much time they need to keep power and to what extent. Do you even need the majority power in the very beginning? How could you decentralize it a bit more? Could you actually have some sort of a timeframe of, let's say, once you deploy your main functionality and after, let's say, one month after that, you could transfer the keys to a larger audience?

In terms of protocols we have seen now, it's more about the question of fair distribution models. We have seen models where the early investors, they have majority of the token supply. You can't call it decentralized until you as a team can be challenged on every single decision you make or proposal you make. Once you reached that level, then you're decentralized enough, in my opinion and that requires that the team and the investors and even inner circle or something, doesn't hold the majority of the voting power. For me, that's a landmark of having fair distribution model and being decentralized enough.

“You can't call it decentralized until you as a team can be challenged on every single decision you make or proposal you make.”

CR: I guess what you're saying is that when you're first starting with a protocol, it makes more sense to have some control of the protocol in the hands of a reduced number of people of the team so that they're able to make sure that it's being built in the safest way possible for users and that they're able to quickly react if something goes wrong. But over time, as the protocol becomes more tested and safer, maybe when audits are done, then it's maybe the right time to start giving up that power so that the protocol can become increasingly decentralized.

You're saying that a protocol can only become decentralized once the tokens or voting power of the team are distributed among the community and not so concentrated in the team itself. I guess, it's maybe a subjective measure of how much is too much of concentration, but that's something that the community itself can discuss.

SK: In my experience, honestly, the investors and the team, they don't need as much as they think. Because what we're building is very big and we should have the least concentration possible. These protocols will be very, very big and what's different from compared to traditional networks, payment networks or any trading venues and facilities is that these might become ever bigger. You don't want to hold that responsibility. You don't want to be responsible for vast amount of funds of end-users. That's something that I have learned.

“ In my experience, honestly, the investors and the team, they don't need as much as they think. Because what we're building is very big and we should have the least concentration possible (…) You don't want to hold that responsibility. You don't want to be responsible for vast amount end-user funds.”

Innovative Distribution Models

I think, of course, you should not even deploy any code to production without audit. That's something that the community in DeFi needs to come together and just find ways to fund audits. I mean, there's ways to do that.

But beyond that particular thing, I think people should think of more innovative models, how you could come up with variations of this fair distribution model, for example. I mean, could there be a way for example, that in the beginning, when you deploy something, maybe there are 20-30 persons from the community that holds a key or like 10 persons at the very beginning? We have seen situations where you have one key into a smart contract which holds a few hundred millions worth of value and then the community starts to panic and arranging multi- signature wallets for that with multiple persons.

I think when you're deploying, it's not that interesting to copy/paste what someone else has done, but actually trying to improve a bit and thinking what could be a bit more improved iteration of what has happened before. I would really encourage people to innovate on their distribution model, how you could launch in a decentralized fashion and move towards decentralization. You don't need to be aggressively decentralised in the very beginning, but to a certain extent that you can make sure the trust and the concentration doesn't rely on a single entity.

“I would really encourage people to innovate on their distribution model, how you could launch in a decentralized fashion and move towards decentralization.”

CR: That's a fair point. There needs to be more innovation in this regard. I think we're so early in these new distribution systems that I doubt that we've arrived at the right model to do this. I wanted to touch on what you said on the need to deploy code that's been audited and how hard it's been in DeFi because of how expensive audits are.

I'm interested in this question of balancing innovation with really secure and audited code. Do you think maybe there's a middle ground there, how can a lone developer who's building a DeFi protocol go about deploying code without an audit and handling those risks? Or do you think that they simply just shouldn't do it and be always on test net before they can manage to get an audit?

Safe Launches

SK: That’s an interesting one, because the way I see there's multiple ways to handle the audits. There is a way to attract community and actually ask from the community, can we split the funds? Or is there some community auditor that could help? There's different models. You could always reach older projects, for example, Aave and ask, hey, we're building this interesting thing and it's very important for DeFi. Can you help us somehow? We have grant program available, where you can actually ask grants for whatever you're building, whatever you're researching and including audits. Those kinds of things we are helping out.

But I think one of the things we can distinguish is that there has been this a way of deploying it without anyone knowing and creating a rush into the project in a way that you want to quickly distribute the tokens and get quickly people involved. And this has had the function of not auditing some code and putting disclaimers and the disclaimers will not help if you are creating the rush there, your idea is to get people quickly to deposit, it will not help. Human psychology isn't capable of resisting that, you know. People are led by greed sometimes. It doesn't help.

That is important for a developer. Every developer that is deploying into mainnet just to understand the risk and understand that everyone needs to do their best practices and to ensure that not audited code can go into mainnet for use.

I think just by reaching out and creating communities first, there is no rush into deploying code. I mean, try to see communities grow first and then the products are grown there and the communities coming together and helping out to get those audits, asking grants. This is the healthy way of getting funding to bootstrap. If you want, you can always go to venture capital. Is not the most preferred model. There’re different ways. But definitely, I see there's more potential in actually building communities first and then building the product and then launching. Because the community will help your product build and that's a substantial help.

“I see there's more potential in actually building communities first and then building the product and then launching. Because the community will help your product build and that's a substantial help.”

CR: Would you rather see less of these protocols launching overnight and getting a bunch of liquidity with these yield farming incentives and tokens before having an audit?

SK: If your code is public, and everyone can look at it for a certain period of time and launch, that gives more reassurance. There are people from the community who can audit the code and take a look before they implement. You can arrange partnerships if you're using different protocols and discuss with their staff, their developers, they can read the code and review.

Time only is a positive thing in DeFi. We somehow created an environment recently that you need to develop quickly. Once it's ready, you need to deploy into the mainnet and then we have issues and then we're looking ‘okay, what happens now’ and then if some exploits or bug will be found. There's no rush.

There are many protocols which have been public in terms of code and have successful launches. This is a more controlled, auditable environment is what the space needs and this will help us go further and create bigger things. Even maybe there might be things that people might notice that you could improve your product. I see that more valuable than anything else.

CR: I agree and I guess to start wrapping up, I'm interested in your views of DeFi going forward. Considering this recent craziness where these protocols launching very quickly and drawing hundreds of millions of dollars have accelerated growth in DeFi, but added a lot of risk as you have highlighted.

Going forward. I mean, what do you hope to see in terms of DeFi growth? Do you think it'll be more sustained growth, not so exponential, but maybe a more sustained growth without so much risk? Do you see any risk of these protocols producing some crash, which takes people away from DeFi for a while? What do you see coming up next?

Testing Principles

SK: The risk is always there. As a DeFi protocol builder, for me, that part of life is serious. It's always there for each and every protocol. It's our job to keep these protocols as safe as secure as possible and ensure that, nothing like this should happen. That's how we support each other in the space.

The recent developments aren't helping the direction we should take because the safer we are, the more we get adoption and the more we get institutional adoption where we get large amounts of deposits and capital injections, but also we get allocations from a larger amount of mainstream users. Because if a system is seen as safe, then it's very usable to a larger extent. It takes a long time to build a reputation, but short time to lose it.

What DeFi is now having is a bit of, I would say, testing in terms of principles. If we are able to come together and improve the safety nets of the protocols, ensure that we have a way of deploying code into the mainnet that is secure, that is attractive, that’s what gets people into this space and we have a base layer for financial applications products that we have never ever seen before just because of the efficiency and security that blockchain provides. I really want to see that.

If something happens, the industry stalls a year or two and that will be very shameful, because then all that we’ve built, we’ll have to wait for the next phase where the trust is regained again. I will not like to see that happen. ‘

I would like to see that the space will grow in a healthier direction. If you take the healthier path, it doesn't mean that we are going to grow slower, it's actually might mean that we are passing more and more tests and are more battle-tested and actually capital comes quicker.

“If something happens, the industry stalls a year or two and that will be very shameful (…) I would like to see that the space will grow in a healthier direction. If you take the healthier path, it doesn't mean that we are going to grow slower.”

CR: I think that's a great point. Last question is, we started the conversation highlighting how Aave went from zero value locked at the start of the year to $1.3 billion now, the second largest DeFi protocol after Maker. Are you looking to overtake Maker, do you see that happening? [NOTE: it did overtake Maker the day after the interview]

Aave vs Maker

SK: I think, Maker as a protocol has done very substantial work in terms of issuing the Dai stablecoin. It's quieter in terms of the space, because it's a very base layer. Many of these things become a bit quieter at some point, because you're providing the very foundation layer for the ecosystem. I will see Maker and Aave will be like that in the future.

The growth of the product and the brand, it really depends on the innovation that you create. Our goal is we don't focus that much on other protocols, only to the extent that we can ensure that if a developer builds something on Aave, our composability allows them to use other things and build products that are using everything. That is what DeFi is all about.

I think these protocols that we see now in DeFi, they will be there if nothing serious will happen. I really love what we are doing now in the space in terms of those baseline protocols and I really love to see more innovation. I hope something that is built on top of Maker and Aave will become bigger, that will be awesome.

The Defiant is a daily newsletter focusing on decentralized finance, a new financial system that’s being built on top of open blockchains. The space is evolving at breakneck speed and revolutionizing tech and money. Sign up to learn more and keep up on the latest, most interesting developments. Subscribers get full access at $10/month or $100/year or 70 Dai/year, while free signups get only part of the content.

About the founder: I’m Camila Russo, author of The Infinite Machine, the first book on the history of Ethereum. I was previously at Bloomberg News in New York, Madrid and Buenos Aires covering markets. I’ve extensively covered crypto and finance, and now I’m diving into DeFi, the intersection of the two.