Due to the lower liquidity of RWA collateral, the approach to handling defaults differs from traditional lending protocols.
RWA protocols introduce a mechanism called "Tranche" to focus on how asset losses should be allocated among Lenders.
This article will explore how various RWA protocols respond to defaults through various cases.
In this piece, we will delve into the crucial topic of handling risk management for Real-World Asset (RWA) tokens. We will analyze how various protocols are tackling these associated risks and discuss any troubles they may have experienced so far. Following this, we will explore the ways in which DeFi protocols are penetrating the lending market collateralized by RWAs. To conclude, we will evaluate a range of perspectives about RWAs in the web3 domain, while also considering the unique benefits and constraints these assets may introduce to the Web3 industry.
In the following section, we will be discussing the notable default risk associated with using RWA as collateral and the strategies that have been put in place to mitigate this risk. Before diving into the details, it’s crucial to comprehend the risks tied to the use of token collateral in traditional DeFi and the ways existing DeFi protocols tackle default risk. We will commence with some fundamental context, then shift our focus to how defaults in loans backed by RWA tokens differ, and examine how various RWA protocols approach these challenges.
In the previous DeFi model, when a user deposits stablecoins into a protocol, they only need to take care of the mechanisms of the protocol whether they are able to protect the value of the deposits. In other words, it is crucial to evaluate whether the protocol would be able to perform processes such as liquidation when the price of the collateral token of a borrower drops. For example, if you were a depositor of USDC on AAVE, all you had to worry about was that when the price of the ETH collateralized by the lenders dropped, the protocol would liquidate the lenders’ positions and protect your USDC. The same is true for MakerDAO. All it had to do to keep the value of DAI at $1 is to liquidate the borrowers’ collateral positions before the value of the collateral held by MakerDAO fell below the value of the DAI issued as loans.
So from the depositor’s point of view, they only need to check if the protocol is using proper mechanisms from major protocols like AAVE or MakerDAO, and if there were any cases of defaults. From the protocol’s point of view, when defaults occurred, it checks what is wrong with the mechanism, and only modifies the mechanism appropriately to deal with the default.
1.1.1 Examples
In March 2020, MakerDAO experienced a large number of defaults due to a sharp decline in token prices. The problem at the time was that MakerDAO’s liquidation mechanism was poorly designed, preventing liquidators from participating in the liquidation process as required by the protocol. MakerDAO made improvements to the protocol during the recovery phase, including a significant revision to the collateral auction process of their liquidation mechanism.
In addition to liquidation mechanisms, existing DeFi protocols have risk parameters associated with liquidation to prevent defaults. With a help from a token economy modeling firm called Gauntlet, DeFi protocols such as MakerDAO and AAVE have set loan-to-value (LTV) thresholds for liquidations to occur, and adjust these parameters whenever they believe the market is in a bear market to ensure a protection of their collateral assets. For example, in December 2022, the AAVE community increased the collateral required for stablecoin loans in response to a slowdown in the cryptocurrency market.
1.1.2 Summary
In previous DeFi, if a protocol gets the right price for its token assets and disposes of its collateral at that price, it will not default. In other words, in the traditional DeFi model, if a protocol defaults, it is a problem with the protocol design, thus it means that the protocol is poorly designed or has not properly considered the risks that may arise in the protocol.
1.2.1 Awareness of Borrower Defaults
Protocols dealing with RWA think about defaults a bit differently: they recognize that each RWA token collateral may default, and they don’t consider it the protocol’s obligation to prevent it from happening.
In conventional DeFi, when the price of the collateralized tokens falls below a certain level, they can be sold to the on-chain market. However, RWA tokens do not have a market where they can be sold instantly at a set price. Therefore, even if the price of RWA falls, it is difficult for the protocol to properly sell the collateral to protect the value of the protocol assets. Furthermore, no matter how tokenized RWA is, it is still unclear whether token ownership on the blockchain reflects ownership of the actual RWA. These two factors make it actually impossible to quickly liquidate collaterals on the blockchain and redeem the borrower’s loan position when the RWA token collateral needs to be liquidated.
1.2.2 Tranche: A Mechanism for Responding to Defaults
For this reason, the RWA protocol assumes that a default can occur at any time, and the focus is not on preventing the protocol from defaulting through liquidation, but rather on how to allocate the loss of assets to Lenders in the event of a default. To be more specific, some Lenders may be willing to make deposits because they want a high-interest return, even if there might be some loss of capital, while others may prefer not to incur any losses, even if the return is low. The RWA protocol categorizes Lenders and their deposits based on their risk preferences, even for the same vault.
This structure is called a tranche in traditional finance, and most RWA protocols adopt the tranche structure when accepting deposits from lenders. Tranches are usually divided into senior tranches and junior tranches. A senior tranche is a form of funding that is protected against losses but acquire limited returns, while a junior tranche is a pool that can receive a higher return in exchange for being the first capital to be exposed to losses.
Vaults that accept lender deposits in a tranche structure are also designed in the same way as tranches used in traditional finance. When a borrower repays the amount they borrowed plus interest, the repayment is first allocated to the senior tranche and then distributed to the Lenders who deposited in the junior tranche. Any remaining amount will be distributed to the junior tranche. In the event of a Borrower bankruptcy, any losses on the outstanding amount will be applied first to the junior tranche. When the junior tranche alone is unable to recover all of the outstanding amounts, the losses will be applied to the senior tranche as well.
1.2.3 GoldFinch Finance’s senior tranche
TrueFi’s Structured Credit Vault and Centrifuge Tinlake’s Tin & DROP both have senior and junior tranches per vault. However, GoldFinch Finance has a slightly different tranche structure. Here, lenders who deposit assets directly into a specific vault are called backers, and all of the backer’s deposits are recognized as junior tranches. Conversely, lenders who deposit money into the GoldFinch protocol rather than into a specific vault are called liquidity providers, and their deposits are recognized as the senior tranche. The distribution of how many backers deposit capital per vault determines how much of the capital pooled in the senior tranche is allocated to each vault. The result of this is that each VAULT is able to borrow more liquidity than the capital provided by the BACKERS, and in that sense, the model is called a leveraged model.
Let’s speculate on why GoldFinch is using a different senior tranche structure than other RWA protocols. First, GoldFinch’s target borrowers are all businesses in the developing world, even if they have different characteristics, so it seems natural to bundle all of the vaults that GoldFinch has lent to and create a deposit product for their risk and return, and then package that senior tranche as a diversified investment in financial products in the developing world and offer it to users. Other RWA protocols, on the other hand, are so different in terms of who they are lending to, that I don’t think it would have made sense to bundle all of the loans in the protocol into a single deposit product.
1.3.1 Funding the Junior tranche is important
As described above, in the RWA token collateralized lending protocol, if the borrower is unable to repay the loan, the protocol is exposed to non-repayment damage, and the lenders of these vaults may not be able to get their tokens back. Considering that lenders are mainly token holders who deposit stablecoins and want to earn a return with relatively little risk of losing their assets, there needs to be a way to limit the size of the lender’s loss in the event of a borrower’s bankruptcy.
In the current RWA token-backed loan protocol, the junior tranche fulfills this role. The fact that the junior tranche acts as a buffer against losses creates a situation where ordinary investors can invest in RWA protocols with relative ease. In other words, in RWA token-backed loans that are not “risk-free,” the junior tranche is acting as insurance for each loan position. RWA protocols need to be concerned with ensuring the size of the junior tranche deposit, monitoring that the junior tranche that exists in each vault is providing a sufficient return to the junior tranche depositor.
1.3.2 The ratio of junior and senior tranche sizes should be set well
If the size of the junior tranche is too small relative to the size of the loans originated in the vault, the junior tranche alone will not be able to absorb enough of the losses; in that respect, the junior tranche should not be too small. Conversely, if the size of the junior tranche is too large, there’s a chance that you won’t receive enough interest income for putting money into the junior tranche because of the structure of the junior tranche, which is distributed to the senior tranche and receives any leftover interest. This means that you need to set the size ratio of the junior tranche to the senior tranche appropriately. This ratio can be set by considering the following factors
Borrower’s probability of non-repayment and expected losses The higher the probability of non-repayment in your vault, the larger the junior tranche should be. To determine a borrower’s probability of non-repayment, you can use the following information: The Borrower’s past lending and repayment history, the price stability and ease of disposition of the RWA pledged by the Borrower, and overall market conditions.
Competitiveness of the junior tranche in terms of yield The interest yield that can be expected from the junior tranche should be naturally higher than the senior tranche, and at least moderately higher than the interest yield that can be earned on stablecoins in other DeFi, so that token holders have an incentive to deposit their assets in the senior tranche.
The RWA protocol should constantly monitor the above factors to track whether the junior tranche can maintain its size. If the return on junior tranche deposits is low, it is advisable to reward junior tranche depositors with tokens from the protocol.
1.4.1 Default Cases
In 2022, news of defaults in the vaults of the RWA protocols have been coming out in rapid succession: first, a $3.4M loan from TrueFi to Blockwater Capital defaulted. Invictus Capital, a crypto fund that borrowed money from TrueFi, also failed to repay a $1M loan on time. Maple Finance also failed to get paid back the 2,400 WETH it lent to Auros Global, a crypto fund. In addition, a series of defaults in vaults managed by Celsius, Alameda, Orthogonal Trading, M11, and others led to Maple Finance defaulting on a $54M loan in December 2022. In addition to Maple Finance, $4.4M in defaults occurred at TrueFi and $2.5M at Centrifuge.
One of the commonalities of the defaults that occurred after October 2022 was that the borrowers were crypto funds, who often put their borrowed assets on exchanges like FTX. They planned to use the loans from the RWA lending protocol to execute their strategies, such as market-making and arbitrage, and make profits to repay the loans and interest. However, when FTX collapsed, their loans were frozen on the exchange, and they were unable to repay the loans from the RWA protocol. In addition to FTX, there is also a case of Invictus Capital that suffered large losses when the Terra stablecoin collapsed and was unable to repay its loan.
In the RWA protocol, when a vault defaults, the losses are shared among the Lenders who funded the vault. When Babel Finance defaulted on a $10M loan from Maple Finance in July 2022, the unpaid amount of the default was charged to the vault’s Lenders, resulting in a loss of about 3.2%.
1.4.2 Protocol Mechanism Improvements
The direction in which Maple Finance changed its mechanisms in the wake of last year’s defaults provides insight into additional default risk management methods that the RWA protocol could consider in addition to the mechanisms described above.
Borrower Diversification — One of the reasons why Maple Finance experienced such a large loan default last year (over $50M) was due to the high concentration of crypto funds in its borrowers. Since the borrowers are all players affected by the crypto market, when an event like the FTX collapse happened, many vaults experienced simultaneous non-repayments. From the protocol’s point of view, such a large number of non-repayments at once creates a burden on the protocol to recover the non-repayment through the treasury created by the protocol. Maple Finance has also recognized this issue and has decided to diversify its borrowers from 2023.
Empowering Vault Administrators — Previously, vault administrators could not declare a default in the event of a borrower’s non-repayment until a certain amount of time had passed. On its face, this may not seem like a problem, as the default is triggered by a mechanism defined by the protocol. However, since there is no way to liquidate the borrower’s collateral to force them to repay the loan, the RWA protocol requires the lender to freeze the vault assets as soon as possible to preserve the lender’s assets as much as possible when it is determined that the borrower has lost the ability to repay. With this in mind, Maple Finance has modified the protocol to allow the vault administrator to declare an immediate default if, in the vault administrator’s opinion, the Borrower has lost the ability to repay.
Aligning Vault Administrator Incentives — An issue that arose during the 2022 default was that some vault administrators did not share accurate information about the status of the vaults they were managing and the repayment status of their borrowers. Given the nature of the RWA protocol, which relies not only on the logic of the protocol, but also on the trust relationships between Lenders, Borrowers, and administrators using the protocol, this behavior by vault administrators can lead to greater losses due to defaults. In this respect, rules are also needed to force vault administrators to deposit a large amount of assets in the junior tranche of a vault when they create it.
Since the second half of 2022, one of the most talked about issues in the RWA ecosystem has been the emergence of several projects among existing DeFi protocols that have announced their intentions to incorporate RWA. MakerDAO and AAVE are the most prominent examples, while Frax Finance has also been actively discussing RWA-related mechanisms. In this part, we’ll discuss the context in which MakerDAO and AAVE have tried to introduce RWA, and what form it’s taking.
2.1.1 Purpose of Introducing RWA
Stable growth of the protocol — One of the main agendas of the MakerDAO community has always been to create a means to stably increase DAI issuance. One of the problems MakerDAO has faced since running the stablecoin since 2018 is that it has been unable to continue to grow DAI with only ERC20 token collateral. MakerDAO recognized that accepting only token collateral would make DAI issuance too dependent on long demand for the cryptocurrency. By incorporating RWA collateral, MakerDAO hopes that DAI will provide utility as a lending instrument for entities looking to borrow on the basis of RWA or credit. Under this strategy, MakerDAO will no longer rely solely on cryptocurrency market conditions, but will also be able to create demand for DAI based on the real economy. As a result, DAI’s issuance will be able to grow more stably.
Monetization — MakerDAO also initially created a demand for DAI through the DAI deposit fees offered by the protocol, known as DSR, and based on this demand, the interest paid by borrowers became the primary revenue stream for the protocol. However, the MakerDAO community soon realized that the yield from tokens was not high or sustainable, and they started to focus on the yield from RWA, which is quite high and sustainable, and they started to issue DAI directly to entities that want to borrow against RWA and make the yield from RWA the main source of revenue for the protocol.
2.1.2 Adoption Criteria
Whenever MakerDAO creates a vault for a particular RWA or makes a loan secured by an RWA to a particular borrower, it relies on input from its community members to make the decision. In making these decisions, the MakerDAO community has decided to use the following evaluation criteria
Scalability: First of all, MakerDAO’s position is that the RWA must have a high enough issuance volume and a large enough market size to increase DAI issuance.
Stability: As RWAs are used as collateral for DAI issuance, it is important that the collateral remains stable in value. To that end, the MakerDAO community determines if the price of the RWAs is sufficiently stable and if there is sufficient liquidity in the market to dispose of them. Furthermore, the protocol assess whether there is any risk that this could jeopardize the stability of MakerDAO.
Demand: Even if MakerDAO accepts the RWA as collateral and issues DAI, it doesn’t make much sense to accept the RWA as collateral on MakerDAO if it’s better to borrow from a traditional financial institution with the RWA in terms of interest and other financial costs. In this regard, the MakerDAO community evaluates the utility of borrowing on MakerDAO from the perspective of the RWA owner, i.e., if MakerDAO and the RWA owner are mutually beneficial.
Regulatory Risk: In addition to dealing with RWA, the MakerDAO team also considers whether there are additional regulatory or legal entities that need to be involved in the process of incorporating the asset into the protocol, and whether there are additional technical, economic, or regulatory risks.
2.1.3 Steps required for implementation
In addition, MakerDAO has summarized the steps required to adopt RWA as follows.
How can RWA assets be converted into tradable assets on MakerDAO? In order to hold RWA as collateral on the protocol, users need to be able to tokenize it and have a legal mechanism in place that allows them to dispose of it in the event of a liquidation.
Is it possible to prevent a DAI borrower from pledging the same asset again in the traditional financial world? Since the legal rights of tokenized RWA are separated from the rights in the real economy, there is a risk that a borrower may borrow from MakerDAO with their RWA and then take out another loan in real finance. If this happens, it is unlikely that the borrower will be able to repay the loan they received from the RWA protocol. In this regard, MakerDAO tries to prevent borrowers from using their RWA as collateral to obtain another loan.
Can price information, etc. be reliably retrieved? How will the oracle be designed? Even if there are differences in the way collateral is disposed of in the liquidation process, the value of RWA still has a significant impact on the borrower’s repayment potential because RWA acts as collateral. In that regard, MakerDAO is concerned about how to reliably fetch the value of those RWA assets.
2.1.4 Adoption Status
In July 2022, MakerDAO created the Real World Finance Core unit, which aims to increase MakerDAO’s impact on real-world finance. Through this unit, the protocol team has been making RWA-backed loans to actors in real-world finance and continues to create vaults for them. These vaults are working by issuing DAI directly to borrowers needing loans. You can check out the active RWA vaults here. The main vaults are as follows (as of March 2023)
(Source: MakerDAO)
The amount of DAI issued as RWA collateral on MakerDAO has now exceeded $600M, and the pace of DAI issuance has been increasing rapidly, especially since the end of 2022. Vaults like RWA009-A above are particularly significant because they provide funds directly from MakerDAO that would otherwise be lent by a government-regulated bank. In terms of protocol revenue, it’s also helping to improve the protocol’s bottom line, with DAI issued as RWA collateral worth around $650M generating around 75% of the protocol’s total fee revenue as of January 2023.
Against this backdrop, MakerDAO continuously reviews vaults that lend DAI as collateral for RWA, such as short-term bonds, institutional credit, and energy production facilities.
(Source: @steakhouse)
The MakerDAO community has expressed that it is inefficient and slow for each RWA to have to go through governance to decide whether or not to handle the asset. As a result, MakerDAO is looking to partner with RWA tokenization and RWA collateralized lending protocols such as Centrifuge (+ Tinlake) to allow the RWA protocols to take care of the necessary steps in handling RWA assets. MakerDAO is also in talks with Maple Finance about a partnership. Here again, Maple Finance will be utilized as the infrastructure to facilitate RWA-based collateralized loans, and MakerDAO will likely issue DAI to Maple Finance and adjust parameters such as DAI issuance limits and default interest rates for each vault based on the risk of the vault. In a partnership with RWA Protocol, the roles that MakerDAO expects the other protocol to play are as follows: 1) onboarding and managing borrowers, 2) conducting legal proceedings or organizing legal structures, and 3) handling defaults.
2.2.1 Background
The rise of GHO — In July 2022, the Aave Governance Forum, in an agenda item called “Introducing GHO”, proposes a plan to issue a stablecoin called GHO based on the Aave protocol. Like other stablecoins, GHO would be backed by multiple asset classes. Still, instead of being crypto-specific, it was notable that it would be a mix of assets that were less correlated to each other, such as RWA and credit score.
In the GHO mechanism, instead of a single entity managing all collateral, an entity called a facilitator is authorized to issue GHO tokens. Facilitators propose a method for issuing GHO to the community, and if the proposal passes, GHO can be issued in the manner proposed by the facilitator, within certain limits. Interestingly, Facilitators do not have to insist on issuing crypto-collateralized stablecoins; in fact, GHO can be issued/burned in any way that each Facilitator defined, including collateralization against multiple asset classes, algorithmically, credit-based or delta-neutral position-based, and the more diversified the Facilitators, the more stable the price of GHO will be.
(Source: Aave)
The Importance of RWAs in DeFi — In November and December 2022, the Aave community began to discuss the importance of RWAs. In “Helping Aave Think Through RWAs”, Punia from Warbler Labs noted that RWAs could help the Aave protocol in two ways: by providing stability through the diversification of stablecoin collateral outside of crypto and by paving the way for GHO to scale as a global stablecoin by leveraging a larger pool of assets than can be leveraged with crypto.
In “Why RWAs are Important to DeFi”, the author argue that while DeFi can achieve a certain amount of growth (10x~) by focusing on crypto natives, there is a clear limit to the size of the crypto market and that RWAs are necessary to achieve even greater growth (100x~) beyond that. DeFi needs to get some market share from TradFi to enter the mainstream eventually.
DeFi and TradFi are analogous to the reversal of the ratio between e-commerce and in-store transactions, where people prefer e-commerce to in-store transactions now that the internet is available. The same is true for DeFi and TradFi: as blockchain technology develops and becomes more active, people will favor DeFi over TradFi.
Finance is a field where user needs are always present, and people always want flexibility in their financial applications and funds, which is a universal phenomenon. DeFi can serve them better in all these ways.
The visibility and interoperability of DeFi make it safer and less volatile than centralized protocols and better serve the economic system as a whole.
Success on MakerDAO — The success of MakerDAO further strengthens the case for RWA on Aave Protocol. Securing a steady stream of revenue from a DeFi protocol is crucial to its sustainability, and according to the article “Outlining an RWA Strategy For Aave”, as of November 2022, MakerDAO received over 50% of the protocol’s revenue from RWA.
(Source: @steakhouse)
The dashboard above shows the evolution of MakerDAO’s revenue by asset, with RWA accounting for well over 50% of revenue as of March 2023. With RWA accounting for a larger and larger share of total revenue over time, it’s clear that RWA is more important than we might think.
2.2.2 Effects and expected results
According to the article “Outlining an RWA Strategy For Aave”, the risks of Aave including only on-chain assets as collateral for GHO are as follows:
During market downturns, there is a risk of accumulating liabilities in the protocol if there is a lack of market liquidity in the relevant assets, i.e., there is a risk of protocol default if only crypto assets are treated as collateral.
The lack of liquidity in a protocol eventually makes it vulnerable to economic attacks or liquidation.
Cryptocurrency assets are highly sensitive to market conditions, and during market downturns, the cash flow of lending protocols can deteriorate due to lower demand for leverage.
Ultimately, the Aave community expects to see the following effects when RWA assets are offered as collateral for GHO. This overall statement is broadly consistent with the effects the MakerDAO community expected when they decided to accept RWA.
Because RWA is more stable and less volatile than crypto assets, its risk is more dependent on macroeconomic flows, and it is more robust to the volatility of the narrower crypto market itself.
RWA borrowers’ financing is often for productive businesses, not speculative trading. This is a much larger scale than crypto speculation and is consistent regardless of bull and bear markets.
2.2.3 Aave’s RWA adoption strategy
RWA is a completely different asset class compared to traditional cryptocurrencies. There’s a lot of discussion in the Aave community about how RWA should be implemented. In the article “Outlining an RWA Strategy For Aave,” the author has broken down the different elements of DeFi to see which ones are being advocated.
Asset types: The author argues for introducing less risky, lower-yielding RWA first to see how it works in the crypto market to build long-standing trust in RWA facilitators, and then progressively introduce riskier, higher-yielding assets. The asset classes should not be directly correlated to the crypto market, for example, by excluding loans to makers, miners, DAOs, etc.
Debt Ceiling: In the real world, adopting GHO as a fiat currency will take time, so we must assume that all borrowers will initially exchange GHO for an existing fiat (USDC, USDT, USD, …). This is why the assigned debt ceiling should be proportional to the liquidity available to sell or redeem for existing stablecoins, as, ultimately, it is important for GHO to remain soft-pegged to USD to maintain its value. To summarize, the debt ceiling should be such that it will not depreciate if a significant amount of GHO is sold into the market.
Interest Rates: The author argues that Aave’s current interest rate model should not be applied to RWA and that interest rates should be based on underwriting and TradFi-like deals. This is because Aave’s facilitators can set interest rates for RWA differently than other asset classes.
2.2.4 Adoption Status
On December 28, 2021, Aave partnered with Centrifuge to launch the RWA market for the tokenization of RWA (https://rwamarket.io/). Offering Centrifuge Tinlake’s TIN & DROP model within the Aave service, RWA market allowed depositors to earn yield on stable, crypto-correlated RWA collateral, and Centrifuge issuers to use the crypto market as an alternative source of liquidity by lending through Aave.
(Source: https://rwamarket.io/)
Since the RWA market is a closed market, only users who have been verified through a KYC process to become a depositor can participate as an investor. Issuers (Centrifuge Issuers) can create pools for their assets, and LP tokens correspond to the DROP of Centrifuge Tinlake. Depositors (investors) can stake their deposits with USDC to purchase DROP from that pool.
Immediately after launch, the size of the RWA market grew rapidly, exceeding $20M in April 2022, but later, during the market downturn, the market size shrank, and the total market size was around $7.5M as of March 2023. Fortunately, no capital defaults occurred, and the protocol itself works as intended.
Through the operation of the RWA market, Aave has learned that several key challenges need to be addressed to bring RWA to DeFi in earnest.
It needs various domain experts related to Risk Assessment (RA) or Underwriting.
Lenders should consider their strategy carefully, considering the duration and liquidity of RWA collateral, as it is an inherently different class of asset than highly liquid types like crypto.
RWA’s funding depends on Aave’s stablecoin liquidity, which can often be affected by factors outside of Aave.
RWA facilitators. The example of the RWA market has shown us the potential for bringing RWA to DeFi. We’ve seen how the discussion of an RWA facilitator as one of the facilitators for the issuance of GHO is progressing. As of December 2022, according to the Aave community proposal “Centrifuge’s Perspective of an RWA Facilitator for GHO”, Centrifuge is continuing to attempt to become a full-service RWA facilitator for GHO.
In the previous article, we identified a number of key challenges that need to be addressed to make RWA scalable. In the Aave protocol, we expect the RWA facilitator to have the following roles:
Compatibility with DeFi: This refers to ensuring that RWA collateral is well integrated into the on-chain ecosystem.
Trusted infrastructure: Conducting intermediation involving RWA collateral requires a high level of transparency within a sound legal framework. A trusted, professional brokerage framework is required to onboard RWA into the on-chain ecosystem successfully.
Centrifuge is currently developing an index token based on the assets held by the Centrifuge pool. As an index to the entire pool, purchasers of the index token will purchase a portfolio of centrifuge assets, which is the result of risk diversification thought out by the collective intelligence of the credit group within the ecosystem. The RWA pool can also be included as a portfolio composition of the index, and this may be the best case of RWA experience as an Aave GHO facilitator.
So far, we’ve described the mechanisms used by the protocols that handle RWA tokens and why they are designed the way they are. In this part, we’ll take a quick look at the current state of the protocols that handle RWA tokens. We’ll start with the overall trend in lending volume and then cover how each protocol is doing.
3.1.1 The growth and contraction of lending to crypto funds
Below is a graph showing the active loan volume of protocols that are lending against RWA tokens. Before 2022, TrueFi accounted for the majority of RWA collateralized loans, which is partly due to TrueFi being the first RWA collateralized lending protocol. After that, Maple Finance grew rapidly until the second half of 2022, when Maple Finance and TrueFi bisected the RWA collateralized lending market.
(Source: rwa.xyz)
The composition of borrowers that Maple Finance and TrueFi lent to during this time was dominated by crypto funds that engaged in crypto-based market making and trading, such as MGNR, Auros capital, Folkvang trading, and Alameda research. Until the second half of 2022, it seems that Maple Finance and TrueFi experienced rapid growth as trading companies needed more funding due to the crypto bull run.
However, as we move through the second half of 2022, events like the collapse of Terra Protocol and the FTX bankrun have taken a toll on major crypto funds, and the volume of lending to crypto funds has dwindled accordingly. As a result, the weight of Maple Finance and TrueFi in the graph above virtually disappears from 2023 onwards.
3.1.2 Stable growth in lending to non-crypto funds
While Centrifuge and GoldFinch haven’t seen the same explosive growth as Maple Finance, they have grown steadily since 2022 and appear to have maintained their overall loan volume through the big events of the second half of last year without much damage. Here, Centrifuge is lending across various asset classes, including loans secured by real estate and loans secured by freight invoices. GoldFinch is lending to FinTech companies and real estate firms in developing countries. The figure below shows Centrifuge’s lending by sector, and they are actually lending to various sectors, including real estate, developing countries, and institutions.
(Source: @j1002)
Centrifuge and GoldFinch seem to have been spared much of the damage last year because their lending is not concentrated in crypto companies but rather in non-crypto companies, and in that regard, the demand for borrowing from them remains strong and has been growing steadily regardless of the state of the crypto market.
3.1.3 Increasing RWA Collateralization in DeFi Lending Protocols
In addition, as mentioned above, stablecoins and lending protocols like MakerDAO, AAVE, and FRAX are actively entering the RWA collateralized loan space and are slowly increasing the percentage of RWA assets within their protocols. FRAX has been discussing the allocation of FRAX to RWA collateralized loan pools through governance forum discussions since June of last year and has voted on partnerships with companies like TrueFi and PropFi. While we’re still very new to the RWA collateralized lending mechanism in DeFi, it will be interesting to see how fast this space grows over the course of 2023.
3.2.1 Attracting real economy players to the DeFi ecosystem
It’s still true that most of the users of products like DeFi and NFTs are concentrated in a small number of people who have been owning and trading cryptocurrencies for a long time, so to grow the Web3 ecosystem, we need to onboard more users into the Web3 ecosystem. Various initiatives are underway, such as the Web2 X Web3 partnership, and RWA token-backed loans could be one of them.
If RWA collateralized loans can provide value to investment firms and institutions that traditional finance can’t and bring their liquidity into the DeFi ecosystem, it could act as an impetus to grow the overall size of the Web3 ecosystem once again. By slowly bringing more institutions into the Web3 space through the RWA protocols, blockchain can be positioned as an infrastructure for new financial technologies. Bringing RWA tokens into the DeFi ecosystem is a big step towards broadening the user base.
3.2.2 DeFi can have other means of monetization besides tokens
As previously discussed, RWA collateral allows blockchain projects to generate revenue from sources other than cryptocurrency. The problem with traditional crypto-based incentives is that they’re too dependent on market conditions, and they don’t provide meaningful or sustainable returns.
(Source: rwa.xyz)
However, by utilizing interest income from RWAs as a new revenue stream, DeFi protocols can have a revenue stream other than their newly issued tokens, which can be attractive from a DeFi perspective, especially if the risks associated with RWAs are well-understood and managed. In fact, if you look at the rwa.xyz webpage, you’ll see that the average interest yield from RWA handled by DeFi protocols is 10%. Since most of MakerDAO’s revenue comes from its RWA vault, protocols can broaden their revenue streams if they get RWA right.
3.2.3 Secure demand that is not affected by the DeFi cryptocurrency market
One of the biggest problems with blockchain projects in any domain, whether it’s DeFi or NFTs, is that they’re all too heavily influenced by the cryptocurrency market, represented by Bitcoin. This problem is exacerbated for DeFi lending protocols because the people who deposit fiat tokens (ETH, UNI, etc.) on DeFi lending protocols and try to lend are usually people who are trying to leverage their long positions by borrowing. Conversely, people who deposit stablecoins on DeFi lending protocols and want to borrow fiat tokens represent short positions.
So if you look at DeFi protocols like MakerDAO or AAVE, you can see that the size of the protocols increases and decreases dramatically depending on market conditions. But if it includes the entities that own RWA in the user base, these protocols have a demand that’s relatively unaffected by the crypto market and that allows them to operate with some degree of predictability in terms of the size of the protocol.
3.3.1 Trust issues arise
In addition to the advantages in terms of cost and transaction speed, the ability to achieve expected results through blockchain and smart contracts without having to trust a centralized party has been one of the most important values of DeFi. However, most DeFi activities that use RWA tokens as collateral are based on trust. Lenders have to trust a third party in each vault when they deposit their money, and then they have to trust that the borrowers will repay the loan honestly.
Suppose you look at what’s happened in the RWA protocol over the past year. In that case, we’ve had vault defaults and borrowers not redeeming because the third party in the vault was not acting honestly. We’ve had cases like that, and we’re going to continue to have these trust issues until there’s a more solid legal and institutional foundation around the RWA token, and even then, we’re likely to be using this protocol with some level of trust issues compared to fully crypto-native DeFi.
3.3.2 Only certain users are eligible for loans
In the RWA Lending Protocol, we mentioned that a certain amount of trust is required between lenders, borrowers, and third parties, and it is true that RWA token-backed loans are used for under-collateralized lending. However, because of the need for trust between actors, the group of users who can borrow through the RWA Lending Protocol is still concentrated in certain institutions. Look at products such as Centrifuge, GoldFinch, Maple Finance, etc. You can see that the entities that are borrowing or acting as administrators through this protocol are concentrated in institutions with a certain amount of reputation and expertise. In this respect, there is no market where ordinary people can easily borrow against RWA as collateral.
3.3.3 Does it really reduce costs?
Smart contracts can indeed reduce the cost of many arbitrations that occur during financial transactions. However, in the case of RWA protocols, the process of tokenizing the RWA, identifying the borrower, and the risk of the RWA incur additional costs because it requires experts with knowledge of the RWA and entities to handle legal and administrative tasks. In fact, some protocols require establishing a special-purpose vehicle for each vault. In this respect, there are still multiple costs associated with RWA. This limitation will continue to exist until the infrastructure for each RWA is in place and the risk assessment criteria for each asset are somewhat established.
RWA has a lot of value in that it can solve some of the limitations of the blockchain ecosystem and some of the inefficiencies of traditional finance simultaneously. Once we get some administrative guidelines around RWA in 2023, RWA will get a lot more attention from the blockchain market, and that’s why RWA tokens are getting a lot of attention, and this market will grow in the future.
3.4.1 The need to recognize a different identity from traditional financial or Web3 products
But for a protocol that deals with RWA, it’s based on a very different philosophy than what many DeFi or blockchain projects are based on. This is because everything is not handled through smart contracts like the existing Web3 products, but to some extent, you have to trust someone and accept the problems that come with it. This is a very different characteristic from the existing Web3 products that people who have been in Web3 have seen, and the protocols that deal with RWA are located somewhere in between the existing financial system and Web3 products.
In this regard, when protocols deal with RWA, it is necessary for them to clearly understand not only the value they can bring to the existing financial sector and the blockchain ecosystem through their services but also why they are trying to deal with RWA and the various problematic situations and constraints that may arise in the process, both for those in the existing financial sector and the blockchain ecosystem. And protocols should be aware of the problems that may arise due to the transparency and centralization of information as they operate and improve them so that the RWA market can develop into a more mature industry.
3.4.2 The importance of risk management
As I keep mentioning in this article, the protocols that deal with RWA are still centered around lending protocols, and the lending that happens here is in the form of credit. Since the lending itself is often done for leverage purposes, the lending system itself can be destabilized at any time during the process of over-expansion and investment. This risk can be even more devastating to the protocol, especially since the lending is not regulated and monitored like traditional finance but can be done at the discretion of a third party that exists in each vault.
Most RWA protocols are not yet fully equipped to deal with these risks, so protocols are currently forced to rely on opaque information disclosure (compared to traditional Web3 products), the behavior of certain actors, and the unregulated nature of their mechanisms. Against this backdrop, each protocol must consistently take care of the way to manage these risks.
Thanks to Kate for designing the graphics for this article.
We produce in-depth blockchain research articles
The approval of the Ethereum ETF could boost yields for Ethena's synthetic dollar, sUSDe, much like the yields observed following the Bitcoin ETF approval.
Bluefin rapidly grew by transitioning through Arbitrum and finally settling on Sui. Bluefin now has $26.3 billion in trading volume and holds 70% of the market share, showcasing strong user base and liquidity Bluefin plans to become a comprehensive decentralized trading platform by introducing features like a DEX aggregator, easy on-ramps, and cross-chain functionality. Similar to Jupiter on Solana, Bluefin aims to leverage Sui's growth to achieve expansion and generate synergies
SynFutures is a decentralized exchange for perpetual contracts that combines AMM and order book features to provide efficient liquidity, improved price stability, and significant trading volume growth following its migration to Blast L2
Let's take an in-depth look at the structure of Ethena's USDe, which claims to be an on-chain dollar asset, and the problems that USDe aims to solve.