DeFi’s Breakthrough Moment: Inside the Financial Transformation of EtherFi, Aave, Maker, & Lido
A financial deep dive into the revenues, expenditures, and tokenomics shaping the future of DeFi
Executive Summary
The following report is designed to explore some of the most influential DeFi protocols from a financial lens, included is a brief technical overview of each as well as a deep dive into their revenues, expenditures and tokenomics. Given that we lack access to scheduled, audited financial statements, we used on-chain data, open-source reports, governance forums and conversations with project teams to estimate Full Year financial reports for Aave, Maker (Sky), Lido & ether.fi. The table below showcases some of our most important takeaways throughout the research process, providing readers a high level look at the current standing of each protocol. While earning’s multiples are a common way to conclude what is overvalued & what is undervalued, there are key considerations around dilution, new product lines & future earnings potential that tell a more complete story.
*DAI Savings Rate included in Cost of Revenue, but Aave Safety Module is not
**Does not include ether.fi token incentives, as they were in the form of an airdrop
***Is a very rough estimate, using growth rates, interest rates, ETH price appreciation & margin estimates from new products (GHO, Cash, etc) - not to be construed as investment advice
****Aave is currently looking to revamp tokenomics to include an AAVE buyback & distribute
Through our analysis, we came to the conclusion that we are in the process of witnessing several protocols make the flip to sustainable profitability after years of bootstrapping liquidity and building moats. Aave seems to have reached an inflection point, achieving its first months of profitability and rapidly growing a new, higher margin lending product in GHO. ether.fi is still nascent, but has racked over $6B in TVL securing it a spot as a top 5 protocol by size. The liquid restaking leader has also learned from some of the shortcomings of Lido, launching a number of other higher margin, ancillary products to best utilize their billions in deposits. Read the full report to better understand our calculations, estimates & what each community is doing to drive value to their protocol.
Problem Statement & Definitions
On-chain data and analytics have steadily improved since the rise of DeFi in 2020, with the likes of Dune, Nansen, DefiLlama, TokenTerminal and Steakhouse Financial playing a pivotal role in creating real time dashboards on the state of crypto protocols. At Kairos Research, we believe that an important piece of fostering credibility within the industry is to push for standardization across protocols and DAOs, to showcase financial performance, health and sustainability. Profitability is often overlooked within the crypto, but value generation is the only way to sustainably align each of participants within a protocol (users, developers, governance & community).
Below are a number of definitions that we will be using throughout our research & financial statements to attempt to standardize similar costs across each protocol.
Gross Revenue / Fees: This encompasses all of the revenue generated by the protocol, belonging to both the users of the protocol and the protocol itself.
Take Rate: The percentage of fees that users are charged by the protocol.
Net Revenue: The revenue left for the protocol after paying out users of the protocol and accounting for the cost of revenue.
Operating Expenses: A wide range of protocol expenditures, including payroll, contractors, legal & accounting, audits, gas costs, grants & could encompass token incentives and more.
Net Operating Income: The bottom line dollar number net of all costs that accrues to the protocol and token holders, inclusive of token incentives that are related to the operations of the protocol.
Adjusted Earnings: Adding back one time expenses to earnings to more accurately predict future earnings, subtracting known future costs that are not currently being expressed through earnings.
Introduction to the Protocols
We will provide a high level breakdown of the core products offered by each of the highlighted protocols within this report, which were selected to include some of the most established protocols across a range of crypto subsectors.
Aave
Succinctly, Aave is a “decentralized, non-custodial liquidity protocol where users can participate as suppliers, borrowers or liquidators.” Suppliers deposit crypto assets, in order to earn lending yield and earn borrowing power themselves, such that they can either leverage or hedge their deposited position. Borrowers are either the over-collateralized user looking for leverage & hedging, OR arbitrage traders utilizing atomic flash-loans. In order to borrow against their crypto collateral, users of Aave must pay either a fixed or variable interest rate on the specific asset they are borrowing. Aave’s protocol fees are the total amount of interest paid on open (unrealized), closed or liquidated positions, which is then split between the lender/supplier (90%) and the Aave DAO Treasury (10%). Additionally, when a position breaks above its stated Loan-to-Value maximum, Aave will allow a “liquidator” to close the position by taking on the risk of the collateral and paying off the remaining debt balance. Each asset has its own liquidation penalty, which is then split between the liquidator (90%) and the Aave DAO Treasury (10%). A newer product offered by Aave is an over-collateralized, crypto-backed stablecoin, dubbed GHO. The introduction of GHO allows Aave to offer loans without needing a stablecoin supplier on the other end of the transaction, giving them more flexibility on interest rates and allowing them to cut out the middleman and earn 100% of the borrow interest from any outstanding GHO loans.
Aave transparently displays all of the DAO’s income, expenses and current run rate data through a Tokenlogic dashboard. We were able to pull the “treasury revenue” data from August 1 - September 12th and apply a multiplier to annualize the numbers to arrive at $89.4M in net revenue. To arrive at the top line, gross revenue number, we relied on TokenTerminal’s income statement data to estimate margins - as Aave’s revenue is now less simple than 90%/10% due to the split between traditional lending markets, liquidations and GHO borrowing. Our 2025 predictions are heavily based on assumptions, including a relative uptrend in crypto asset prices that will lead to increased borrowing power. Further, Aave’s net margins increased in our model due to a likely shift towards GHO borrowing over third party stables & a revamp in the protocol’s Safety Module, which will be further explained later.
Crypto’s leading lending market is on pace to have its first profitable year during 2024, as supplier incentives have dried up and active loans have continued to trend upwards, sitting north of $6B actively borrowed. Aave is clearly a large beneficiary of both the liquid staking and restaking markets, as users deposit LSTs/LRTs, borrow ETH, swap the ETH for the liquid staking tokens and follow the same process again. This is called looping and it allows Aave users to generate the net interest margin (APY associated with the LST/LRT deposit - Aave borrow interest) without taking on tremendous price risk*. As of September 12th, 2024, ETH is Aave’s largest outstanding borrowed asset with over $2.7B in active loans across all chains. We believe that this trend, enabled by the concept of Proof of Stake + Restaking, has changed the landscape for on-chain lending markets - greatly increasing the utilization of a protocol like Aave in a pretty sustainable manner. Before looping grew in popularity, these lending markets were dominated by users looking for leverage against their crypto, such that they would tend to only borrow stablecoins given they were heavily skewed long.
The introduction of GHO created a new, higher margin lending product for Aave. It is a synthetic stablecoin from which none of the borrow fees need be paid out to suppliers. It also allows the DAO to offer slightly below market interest rates, driving the demand for borrowing in otherwise difficult environments. GHO is undoubtedly one of the most important parts of Aave to monitor going forward from a financial perspective, given that the product has:
High upfront costs (Technical, Risk & Liquidity)
Audits, development work & liquidity incentives will continue to slowly dissipate in the coming years
Large relative upside
The outstanding GHO supply is $141M, which is only 2.35% of Aave’s total outstanding loan balance and 2.7% of the outstanding DAI supply
Nearly $3B in non-GHO stablecoins (USDC, USDT, DAI) are currently borrowed on Aave
Higher margin than Aave’s lending markets
While there are other costs to consider while issuing a stablecoin, it should be cheaper than needing to pay out suppliers
Maker DAO’s net revenue margin is 57%, compared to Aave’s 16.31% margin
The protocol’s native token, AAVE, is trading at a $2.7B fully diluted valuation (FDV), giving it roughly an ~103x multiple on its estimated $26.4M in annual earnings, however we believe that this could be effectively compressed over the coming months. As mentioned above, favorable market conditions should increase borrowing power, spur a new demand for leverage and likely be accompanied by liquidation revenue - which has been quieter throughout 2024. Finally, even if GHO’s market share growth is just the result of the stablecoin eating into Aave’s traditional lending market, it should have an immediate positive impact on margins.
MakerDAO
Maker Protocol is a decentralized organization which issues a stablecoin (DAI) against a wide range of crypto & real world collaterals, such that users can both lever their assets and that the crypto economy can have access to a “decentralized,” stable store of value. Maker’s protocol fees consist of “Stability Fees” which are generated by interest paid from borrowers and the yield generated by the protocol’s allocation to yield generating assets. These protocol fees are split between the MakerDAO and depositors of DAI into the DAI Savings Rate (DSR) contract at a rate dependent on the DAO. Like Aave, MakerDAO also charges a liquidation fee. When a user's position falls below the necessary collateral value, the loan is closed through an auction process and the protocol takes a portion of the remaining position value to discourage liquidations and protocol stress.
MakerDAO has flourished over the last several years, benefiting from liquidations during the speculative ups and downs of 2021 but ALSO creating a more sustainable and less risky line of business as global interest rates increased. Introducing new collateral assets like United States’ Treasury Products and USDC allowed Maker to make their assets productive and generate returns superior to the standard DAI borrow rate. When exploring the DAO’s expenditure, a couple things are clear to us:
DAI being deeply ingrained throughout the crypto ecosystem (CEXs, DeFi) has allowed Maker to avoid dishing out millions of dollars in liquidity incentives.
The DAO has done an impressive job in prioritizing sustainability
Throughout 2024, Maker is on pace to generate a net protocol income of roughly $88.4M. Given that MKR has a $1.6B valuation, it is trading at a 18x multiple to net protocol income. In 2023, the DAO voted on revamping the protocol’s tokenomics to return some of these earnings to MKR holders. As DAI accrues to the protocol through continuous borrow interest rates (stability fees), Maker builds up a system surplus which they target to keep around $50M. Maker’s Smart Burn Engine utilizes this surplus to buy back and burn MKR through their Surplus Auction. According to Maker Burn and visualized by Steakhouse below, 11% of the MKR supply has been bought back and used for burns, protocol owned liquidity or treasury building.
Lido
Lido is the largest liquid staking provider on Ethereum, a platform which connects ETH holders with a decentralized network of validators that are willing to stake their assets. When the user stakes their ETH through Lido, they receive a “Liquid Staking Token” or a fungible representation of their underlying staked balance, such that they can avoid both unstaking queues & the opportunity cost of not being able to utilize staked ETH within DeFi. Lido’s protocol fees are the ETH yield paid out for validating the network, which is split between the stakers (90%), the node operators (5%) and the Lido DAO Treasury (5%).
Lido is an interesting case study for DeFi protocols. As of September 10th, 2024 - they have 9.67M ETH staked through their protocol, roughly 8% of the entire ETH supply and more than 19% of the staking market share. At $22B in total value locked, there is arguably no other protocol that has taken control of the market as well as Lido. However, you can see above that they are still shy of profitability. What changes could be made to ensure that Lido is able to cash flow in the near term?
The DAO has made massive strides in cutting costs over the past two years alone. Liquidity incentives were hugely important in bootstrapping stETH as power users naturally gravitate towards the LST that is the most liquid and ingrained throughout the ecosystem. We believe that the DAO will be able to further decrease this line item as it becomes clear that stETH has an impressive moat. If Lido were to completely cut liquidity incentives - like Maker - we estimate that they would already be in positive territory for 2024. Even with the cost cutting, $7M in the black might not be enough to justify LDO’s $1B+ fully diluted valuation.
In the coming years, Lido must look to either expansion or cost cutting to grow into their valuation. We see a couple potential paths to growth for Lido, either the ETH network wide staking rate increases from 28.3% or Lido pushes hard to expand outside of the Ethereum ecosystem. We believe that the former is fairly likely on a long enough time horizon. For comparison, Solana’s staking rate is 65.5%, Sui’s is 79.5%, Avalanche’s is 49.2% and the Cosmos Hub’s is ~61%. By doubling the amount of ETH staked & maintaining its market share, Lido would be able to generate another $50M+ in net revenues to offset their costs. This assumption is also severely over-simplified and does not account for ETH issuance getting compressed as the stake rate increases. While increasing current market share is also possible, we saw Ethereum’s social consensus grow very skeptical of Lido’s dominance during 2023, which marked the top of its blitzkrieg growth.
ether.fi
Like Lido, ether.fi is a decentralized, non-custodial staking and restaking platform that issues liquid receipt tokens for their user’s deposits. ether.fi’s protocol fees consist of ETH staking yield AND Actively Validated Service revenue that is paid out for providing economic security through the Eigenlayer ecosystem. These fees are broken up by ETH staking yield, which is split between the stakers (90%), the node operators (5%) and the ether.fi DAO (5%) and then Eigenlayer / Restaking rewards, which are split between stakers (80%), node operators (10%) and the ether.fi DAO (10%). ether.fi has a number of other ancillary products that generate notable revenue, including “Liquid” which is simply a number of restaking & DeFi strategy vaults that attempt to maximize yield for depositors. Liquid charges a 1-2% management fee on all deposits that is accrued back to the ether.fi protocol. Additionally, ether.fi recently launched Cash, a debit/credit card product that allows users to either spend their restaked ETH or borrow against it for real life purchases. Cash grants users the ability to earn cashback and avoid crypto offramps & gas fees, in exchange for an annual fee paid in ETH.
ether.fi is the undisputed market leader in liquid restaking as of September 2024, with $6.5B in TVL across their restaking and yield products. We have attempted to model out the potential protocol revenue from each of their products within the above financial statement, using the following assumptions:
ether.fi Stake will have an average 2024 TVL of ~$4B, assuming the current stake remains constant for the rest of the year
The average ETH staking yield will fall around 3.75% for the year
EIGEN pre-market FDV is roughly $5.5B & the restaker rewards emission schedule of 1.66% during 2024 & 2.34% during 2025 means that ether.fi should be directed top line EIGEN revenue of ~$38.6M in 2024 and ~$54.4M in 2025
By looking through EigenDA, Omni & other AVS reward schemes, we estimate a total of ~$35-45M will be paid in rewards to Eigenlayer restakers, an annual yield of 0.4%
Cash is the most difficult revenue line to model out, given its recent launch and the lack of transparent precedence across the space. We worked with the ether.fi team to gauge the demand for pre-orders, alongside the cost of revenue seen by large credit card providers to come to a best estimate for 2025 - something that we will watch closely over the coming year.
While we understand that ETHFI token incentives are a cost to the protocol, we have decided to leave them at the bottom of the financial statement for a handful of reasons, including: these expenses are heavily frontloaded due to its airdrop and bootstrapping, these expenses are not a necessary cost of business moving forward AND we believe that EIGEN + AVS rewards will be sufficient to offset the diminishing ETHFI emissions. Given that withdrawals have been enabled for some time & ether.fi has already seen the heaviest of its net outflows, we think the protocol is closer to settling in on a longer-term sustainable TVL target.
Token Value Accrual and Scoring System
Beyond simply evaluating these protocol’s ability to monetize their product, it is worth exploring where each protocol's earnings end up - extremely relevant to the crypto industry. Regulatory uncertainty has been a driving factor in creating a vast landscape of earnings distribution mechanisms. Dividends to token stakers, buybacks, token burns, accumulation within treasuries and many more unique approaches have been taken to attempt to allow token holders to participate in the upside of protocols and have a reason to participate in governance. In an industry where token holder rights /= shareholder rights, market participants must thoroughly understand the role their tokens play in the protocols which they govern. We are not lawyers & are taking no stance on the legality of any distribution method, simply exploring how the market could react to each.
Stablecoin / ETH Dividends:
Benefits: Measurable benefit, Higher quality payout
Drawbacks: Taxable event, gas claims, etc
Token Buybacks:
Benefits: Tax free, constant buying pressure, growing treasury
Drawbacks: Subject to slippage & frontrunning, no guaranteed return to holders, concentrating treasury in native token
Buyback & Burn:
Benefits: Same as above, increases earnings per token
Drawbacks: Same as above + no treasury growth
Treasury Accumulation
Benefits: Increases protocol runway, diversifies the treasury, still under control of DAO participants
Drawbacks: No immediate benefit to token holders
Tokenomics are clearly more of an art than a science and it is difficult to know when distributing earnings to token holders is a more effective use of capital than reinvesting it in growth. For the sake of simplicity, in a hypothetical world where a protocol has already maximized its growth, owning a token that is redistributing earnings will increase a holder's internal rate of return and take risk off of the table every time they receive some form of payout. We will explore the design and the potential value accrual of ETHFI and AAVE below, both of which are currently being revamped.
Forward Looking
Aave
As of today the GHO supply is 142M. Exploring the current dynamics, the weighted average borrow rate for GHO is 4.62% and the weighted average stkGHO incentive payout is 4.52%, with 77.38% of the total GHO supply staked in the safety module. Therefore, Aave is earning 10bps on $110M worth of GHO and 4.62% on the unstaked $32M. There is of course room for GHO borrow rates to drop below 4.62%, given global interest rate trends and stkAAVE discounts, so we have added projections for GHO’s impact at 4% and 3.5% as well. There should be many opportunities for Aave to foster the growth of GHO during the coming years, the chart below breaks down how the path to $1B in outstanding GHO loans will impact the protocols earnings.
While Aave is clearly positioned for growth, Marc Zeller also put forth a temperature check within Aave’s governance forum to revamp the protocol's expenditure as well as Aave’s native token, AAVE. The premise of the revamp is that Aave is quickly becoming a profitable protocol, but that they are currently overpaying for an imperfect safety module. As of July 25th, Aave had $424M in the safety module, primarily consisting of stkAAVE & stkGHO, both of which would be imperfect assets to cover any bad debt due to slippage and depeg risks. Further, through token emissions the protocol is incentivizing AAVE secondary liquidity, such that slippage could be minimized if stkAAVE must be used to cover bad debt.
This concept could be radically revamped if the DAO votes to instead utilize aTokens like awETH and aUSDC for the safety module, while isolating stkGHO to only cover GHO debt. stkGHO would never need to be sold to cover bad debt, simply seized and burned. aTokens like the ones mentioned above are extremely liquid & make up a majority of the protocol's debt. In the event of under collateralization, these staked aTokens could be seized and burned to cover any bad debt. The goal of the proposal is to lower expenditure on both safety module & liquidity incentives, while garnering more effective backstopping. Current safety module spend is ~44.3M annually, making a 10-20% efficiency improvement is absolutely massive to the protocol and to AAVE as well. Zeller further explains the role of stkAAVE under the new plan in the image below.
If the proposal is voted forward, it should have beneficial implications for the AAVE token as it will now have more constant demand, but also will allow holders to be rewarded without having the downside risk of getting their stkAAVE seized to cover bad debt. We are uncertain of the tax implications of the staking contract wrapper, but it heavily favors long term holders of AAVE through the constant buy pressure & redistribution of tokens to stakers.
ether.fi
Establishing an earnings multiple for ether.fi is tempting, given the protocol’s success in quickly creating a sustainable business model. For example, the protocol’s development team and DAO move extremely quickly and have attempted to showcase their financial positioning through a successful proposal that will use between 25*-50% of revenue generated from the Restaking & Liquid products to buy back ETHFI for both liquidity provisioning and the treasury. However, using 2024 earnings numbers to come up with a fair valuation may be futile and complex given the lack of AVS rewards, heavily frontloaded startup costs and the fact that the majority of its product suite is brand new.
The ETHFI token has a fully diluted valuation of $1.34B and is on track to be slightly profitable (barring liquidity incentives) on the year, giving it a very similar profile to Lido’s LDO. Of course ether.fi has to survive the test of time, but it appears that the protocol has a potential route to profitability much faster than Lido and a higher earnings ceiling given continued success with the broader product base. Below is a conservative breakdown of how AVS rewards will contribute to the protocol’s earnings beyond simply ETH staking and EIGEN incentives, which should add a much needed boost to earnings without a drastically different cost base than Lido. The AVS rewards yield is the rewards that restakers will accrue simply from AVS payouts.
As seen with Lido, liquid staking/restaking is a highly competitive sector that keeps margins relatively thin. ether.fi has shown a strong understanding of this limitation and has focused on simultaneously dominating market share while building out ancillary products that fit well into their broader restaking & yield generation thesis. Below are reasons that we believe these other products fit well under the ether.fi umbrella:
Liquid: We firmly believe that LRT power users are familiar with DeFi lego blocks and are looking to maximize yield, drawing them to products that could automate their DeFi strategies. Once AVS rewards are actually “live,” there will be dozens of risk/reward strategies and a new native form of yield within the crypto economy.
Cash: Much like LSTs, LRTs are a superior form of collateral to vanilla ETH, given that they are sufficiently liquid. Users can either use liquid restaking as a yield bearing checking account or borrow against their asset for daily spending at nearly zero cost (restaking rate - stablecoin borrow rate).
Conclusion
Many of the numbers used throughout this report are subjective, given they are projections based upon partial year data, trends, market conditions and discussions with developer teams. We plan to follow up on this report at the end of 2024, to both paint a more accurate full year picture & dissect how the market reacts to the growing earnings of these top DeFi protocols.