Balancing fees and first usage report.
One month ago we released Aave Protocol and the adoption has been absolutely staggering. Indeed, time flies when exciting things happen.
Today the market size of Aave Protocol is above 15 Million USD, and the total value locked in the protocol is already bigger than many DeFi products with much longer market exposure. The Ethereum community has been praising the multiple innovations that the protocol brings to the DeFi space, including the new aTokens, which are pegged 1:1 to the underlying asset, and Flash Loans, a new DeFi primitive enabling uncollateralized loans that can be used in the context of a single transaction.
The Flash Loans have especially captured the attention of the DeFi crowd and we expect other DeFi protocols to follow our lead and implement their flavors of Flash Loans as well. Like any other building block of Ethereum composability, Flash Loans quickly allowed new creative ideas to become reality. Prime examples of this are ArbitrageDAO (a DAO with the goal to market make arbitrage opportunities by leveraging the Flash Loans) and the Maker Vault collateral swapper (which allows you to swap your collateral from ETH to BAT in one transaction).
A very detailed article has been published by Marc Zeller, an amazing read to better understand the use-cases of Flash Loans.
This article will analyze how Flash Loans have been used during this first month of infancy, and how we can properly stimulate the ecosystem by maximizing the possibilities for the “flashboys” and “flashgirls” around.
Let’s talk about volume
The total volume of Flash Loans in this first month is 9,716.6 DAI. Although Flash Loans are assets agnostic, it is interesting to notice that only DAI has been used until now.
The following chart details the composition of all the Flash Loans executed up to this point:
As you can see, the arbitrage case is the most used. There is a reason for that:
- Dexes arbitrage is usually a much more frequent event compared to liquidations
- Collateral swap has only been recently built and it’s still very experimental.
Although the sample is quite small, it’s also important to consider that arbitrage always has been the most obvious use-case while we were discussing Flash Loans over the last months before the release. As developers play around with Flash Loans, new use-cases and ideas are popping up every day. Developers are especially eager to build cross DeFi protocol tools to test the boundaries of Ethereum composability. A recent example is the Maker Vault Swap which needs Aave, UniSwap and Maker to make it whole.
What are the fees?
A 0.09% fee is collected from the Flash Loan amount, 70% of which is redirected as extra income for depositors and 30% of which is split using the same 20%/80% model of the origination fee. There are also additional fees for transactions on the Ethereum Blockchain, and these fees depend on the network status and transaction complexity.
The depositors receive their 70% the moment the Flash Loan is executed, effectively making micro APR spikes in the depositors’ yield.
What about earnings?
In the following chart, we analyze the earnings for the depositors, the protocol, and the Flash Loan executor that have been collected as a consequence of these Flash Loans:
The chart above clearly shows an uneven distribution of earnings in the arbitrage use-case. Most of the time in fact, the protocol and the depositors earned more than the actual Flash Loan user. Since we also need to subtract the transaction cost sustained by the user from the arbitrage opportunities taken with Flash Loans until now, the Flash Loan executor was only able to profit from the two that involved a bigger amount flash-borrowed. In all the others, the arbitrageur incurred a loss.
Reasons for this could be:
- Arbitrage usually involves a relatively small amount of money, and the profits are even smaller. Profits in arbitrage are rarely above 0.5%, and if a 0.35% fee is taken to reward the depositors, not much is left for the arbitrageur.
- Arbitrage usually involves gas races, where people compete in capturing the spread using progressively higher gas prices, which leads to high transaction costs.
The data suggests that a revaluation of the depositors fee for Flash Loans is needed.
Why not 0 fee on Flash Loans?
There has been a vibrant discussion in our community regarding the Flash Loan volume and how to maximize the usage and stimulate adoption. Some users even raised the fateful question, “why not 0 fees”? We have contemplated on the specifics of the Flash Loan fees, and we believe that is important that the Flash Loan executor contributes to the wealth of the protocol for multiple reasons:
- The “Fee” is not actually a fee. Only a small part of the earnings on Flash Loans are redistributed to the protocol to burn LEND. The majority goes to the liquidity providers, that make the Flash Loans possible in the first place.
- Flash Loans involve a (minimal) risk. As you may know, Flash Loans leverage the ability of the Ethereum blockchain to execute atomic transactions. This means that if the Flash Loan fails because the executor does not return enough funds, the transaction is reversed. Seems completely risk free right? Well, not completely. While very small, there is still a certain degree of risk involving smart contracts and the underlying layer (the blockchain itself). Flash Loans leverage a specific condition to work which enforces that the funds are returned at the end of the execution. There is still the remote possibility that a bug is found in the bytecode of the contract, or at a deeper level in the EVM, that might allow an attacker to circumvent this condition. Even if extremely minimal, at Aave we believe that when there is risk involved, there must be a reward.
- It enables rewards for integrators. If you are building a DeFi business, one of the most interesting features of the Aave Protocol ecosystem so far, besides the UX magic of aTokens, is the fact that everyone building on top of Aave gets a cut of the fees. Indeed the Aave protocol itself does not take fees besides burning LEND to prepare the tokenomics. However, if you integrate Aave in the background of your dapp, you will get a cut of all fees coming from your users. During our last community call David Truong, creator of the Maker Vault Collateral Swaps, asked if this feature can be applied to Flash Loans as well. As he built a Flash Loan front end, it would allow him to scale his work while earning revenue for the public good he is providing to the ecosystem. We think David is right, and we will implement a Flash Loans-specific fee sharing scheme, on top of the Deposit and Borrow scheme, with the next update of Aave Protocol.
- The Flash Loans fee is beneficial to the ecosystem. A different source of income than the normal interest coming from borrowers gives the protocol a competitive advantage compared to other similar products. This attracts more liquidity, which in turns allows bigger Flash Loans, which brings more earnings for the depositors, and so on. It’s a beneficial loop for everyone. Even arbitrageurs or liquidators could just deposit their funds previously used for liquidations in the protocol, and enjoy the earnings coming from normal and Flash borrowers.
- Flash Loans use someone else’s money. It’s important to note that what allows Flash Loans to be executed in the first place are the liquidity providers. Without liquidity available, no Flash Loans would be possible. In this regard, even if there is another “streamed income” coming from the borrowers, any usage of someone else’s money should be rewarded if it brings a benefit.
So what’s next?
As the protocol gets more decentralized, the Flash Loans Fee will be one of the parameters to vote on to adjust the overall tokenomics of the LEND Token. Since we still have control of the admin keys internally, the interim DAO needs to initiate and vote on this update.
Based on community feedback and discussion last month, it became clear that the fee needs to be rebalanced appropriately. Therefore we will issue a governance vote to reduce the Flash Loans fee to 0.09%. This amount should be small enough to stimulate new users to keep building on Flash Loans, while at the same time still guaranteeing a good source of income for depositors and the protocol.