In-depth exploration of the current state, potential limitations, and future prospects of the DeFi low-collateral lending ecosystem

Author: Jack, 0xlol

Original Title: 《The Crypto Loan Economy》

Compiled by: Captain Hiro, DeFi 之道

Introduction

Over-collateralized loans are one of the most popular lending methods in the crypto economy. For individuals who need to invest in new assets without selling their current positions, this is a great option. However, a question arises: in a world of income inequality, can over-collateralization truly meet the purpose of lending? Over-collateralization not only presents issues of capital efficiency, but it also prevents people from accessing cheap capital to achieve their financial goals. Yet in a trustless financial system, under-collateralization will only make your liquidity providers poorer. So, what is the solution?

Since ancient times, humans have engaged in lending activities. The earliest examples date back to 2000 BC in Mesopotamia, where farmers borrowed seeds and animals to increase agricultural output. This social activity brought complexities of trust and risk, but also led to significant leaps in economic growth. Those with wealth found new mechanisms to derive returns from their assets, while those lacking wealth gained the ability to borrow to create wealth.

Leverage is the cornerstone of entrepreneurship and creation, helping individuals take risks to develop new products and services. Currently, based on interest income from all private and public institutions, the modern global lending market is valued between $7-8 trillion (excluding the value of outstanding loans).

Global lending market forecast, source: Research and Markets

The value of the global lending market in 2021, along with estimates for 2025 based on a 6% compound annual growth rate. 2021: $693.229 billion, 2025: $880.95 billion.

More importantly, DeFi aims to pair peer-to-peer lending with institutional lending, thereby stimulating exponential growth, and DeFi is currently poised for rapid expansion. While institutions currently have a firm grip on the lending market, DeFi can provide individual lenders with more opportunities and help offer higher yields on personal assets by eliminating the need for intermediaries. In fact, the peer-to-peer lending market is expected to reach a value of $1 trillion by 2025.

Source: Statista

The bar chart shows the forecasted value of global peer-to-peer lending from 2021 to 2025. 2012: $1.2 billion, 2013: $3.5 billion, 2014: $9 billion, 2015: $64 billion, 2025: $1 trillion.

This rapidly changing environment is attracting the attention of many DeFi protocols. As our financial system begins to shift towards transparent ledgers (blockchain), they hope to open their doors in a highly lucrative market. While DeFi offers an option for a decentralized and trustless financial future, low-collateral (UC) lending is currently a bottleneck for the accelerated adoption of DeFi.

TradFi or DeFi?

While traditional finance (TradFi) systems have rich credit scoring, KYC (Know Your Customer) documentation, and legal protections to recover funds, they come with various issues. For instance, even with laws preventing racial discrimination in the lending process, a report studying 2 million traditional mortgage loan applications in the U.S. in 2019 found that Black applicants were 80% more likely to be denied loans compared to economically comparable white applicants, primarily due to biased algorithms (Forbes). DeFi offers an opportunity in some cases by relying entirely on smart contracts and on-chain identities to reduce identity-based discrimination in the lending market.

The issues in the lending market do not stop there. Putting your money in a traditional bank means you have no custody over your assets. There have already been multiple instances of governments and banks using these assets to bail out 'friendly fraudsters.' After the mortgage lending crisis in 2008, several major banks were bailed out by the U.S. government, ultimately passing the burden onto taxpayers.

The DeFi lending market proposes a new era of possibilities, where only those who choose to deposit liquidity into protocols can be held accountable for mismanagement of funds, rather than large governments pushing the tax burden onto those who have never engaged in bad business practices, such as those selling subprime loans. DeFi offers liquidity providers (LPs) and token holders of protocols a new opportunity to participate in governance. In TradFi, if you put money in a bank, you earn single-digit annual interest and have no say in who the bank lends to. But DeFi allows lenders to have a say in who they lend their money to, representing a true form of financial sovereignty.

Most importantly, 31% of adults worldwide do not have bank accounts, while 91% of the global population can use smartphones. DeFi only requires an internet connection to provide unbanked and underbanked communities with financial services they have never had access to before. If executed properly, low-collateral lending can provide these individuals with financial leverage, fundamentally changing their lives.

The Beginning of DeFi Lending

Decentralized lending began with the simplest form: over-collateralized lending. Many popular platforms, such as AAVE, Compound, and Maker, play this role, where users deposit their cryptocurrency assets as collateral to borrow amounts less than the value of their deposits. For example, on AAVE, the maximum loan-to-value (LTV) ratio for DAI is 77%, meaning that for every $100 deposited in DAI, users can borrow up to $77 worth of assets.

On the other hand, low-collateral lending can be partially collateralized or completely uncollateralized. In a partially collateralized lending system, an individual deposits $25 and borrows $100 worth of assets, while in an uncollateralized lending system, an individual can borrow $100 without any collateral. Low-collateral lending is also referred to as unsecured lending, while over-collateralization is typically referred to as secured lending.

At the time of writing, the total value locked (TVL) in the entire DeFi market is approximately $174 billion, with the lending currency market encompassing $36 billion in TVL. However, of this $50 billion, less than 5% comes from unified lending protocols. We analyzed the fundamentals and technical aspects of seven such protocols. Risk assessment modules and lending mechanisms are among the few distinguishing factors. Additionally, it should be noted that in this article, we do not consider flash loans (also known as "single-block liquidity") as low-collateral lending.

Before delving into the relevant details, let’s first try to understand the different purposes of these protocols:

• Maple Finance, TrueFi, and Clearpool: Low-collateral lending protocols aimed at institutional borrowers.
• DebtDAO: Provides revolving credit lines for growth-stage DAOs and protocols (against future income). It offers a new way for DAOs and protocols to raise funds without diluting or releasing native tokens.
• Goldfinch Finance: Provides low-collateral lending for real-world businesses in exchange for fiat currency.
• Gearbox: Offers 0% funding rates for whitelisted protocols, with leverage up to 4x.
• Teller Finance: A low-collateral lending market currently supporting USDC business, such as cryptocurrency-backed mortgages in Texas and personal loans in Singapore, up to $6,000.

Risk Assessment Methods

It is clear that when dealing with low-collateral lending in the anonymous and trustless financial realm, risk is our primary concern. No one wants to lend money to a random Twitter anonymous user and hope to get their funds back instead of seeing them spend a large sum of cash. So, what actually prevents some people from stealing all the borrowed funds in real life rather than on-chain? Well, banks and other lending institutions are likely to know your whereabouts, they influence your credit score, and may cause a lot of social and financial trouble for defaulters.

In fact, each on-chain protocol uses its own customized risk assessment methods to address this issue. The different risk assessment methods mainly arise from the different purposes of each protocol, such as different target customer groups and levels of control over borrowed funds.

Analysis of different risk assessment methods

A comparative analysis of the different risk assessment methods used in secured lending protocols. These methods include: pool delegation, community voting, third-party risk assessment, and smart contracts.

Maple Finance, TrueFi, and Clearpool all have similar processes before the loan approval stage. First, each institution obtains a whitelist for individual protocols through an application process, fills out KYC/KYB (Know Your Customer/Business) forms, provides financial documents, and undergoes a credit score assessment executed by the protocol. The credit assessments of each protocol vary slightly, but they fundamentally analyze the business risks and reliability of the applicants. While the initial whitelist stage is quite similar in the lending application process, the differences between institutional lending protocols occur in the specific loan approval process, which we will discuss below alongside the risk assessment methods used by non-institutional lending protocols.

Pool Delegation: Maple Finance, DebtDAO (not yet released)

Maple Finance and DebtDAO rely on professional credit analysts who stake their native governance tokens ($MPL and $DEBT) to evaluate and approve loans. However, Maple Finance allows pool delegates to negotiate loan terms with borrowers, such as interest rates, maturity dates, and collateral ratios. DebtDAO risk analysts simply approve or reject their loan requests, and this process is non-negotiable. The work of these pool delegates is rewarded, as they receive a portion of the interest from borrowers.

Maple Finance's engine

The credit approval system of Maple Finance, which relies on pool delegates staking MPL to approve borrowers' loan requests.

Community Voting: TrueFi

Similarly, TrueFi requires staking its native token $TRU to approve loan requests, but it does not rely on a single pool delegate to fulfill this role; instead, it distributes risk assessment across all stakers. Each loan request requires at least an 80% approval rate to open the lending pool for deposits. Like DebtDAO, the community simply votes for or against the loan proposal. The terms are proposed by the borrower, and there is no negotiation phase in the entire process. 10% of the interest generated by the protocol will be paid to stakers of $stkTRU as a reward for approving loan responsibilities.

TrueFi's engine

TrueFi's credit approval system relies on community governance voting from $TRU stakers to approve borrowers' requests.

Third-Party External Organizations: Clearpool, Goldfinch

Clearpool Finance requires whitelisted institutions to stake at least $500,000 of CPOOL to qualify for credit assessment. This distinguishes the protocol as it requires institutions to stake governance tokens rather than using credit risk analysts. It considers the staked amount as a sign of good faith from the institution, indicating that outsourcing credit analysis to the third-party partner X-Margin is worthwhile. Once the staking is approved, the institution's lending pool can accept funds.

Goldfinch also requires businesses to stake their native token $GFI, but not a strict nominal amount; instead, they must stake $GFI equivalent to twice the audit fees. Goldfinch then outsources the credit check of financial statements and the legitimacy of identities to its third-party partners: Persona or Parallel Markets. If the entire process goes smoothly, each business receives a unique identity (UID) non-transferable NFT linked to its designated wallet to avoid witch attacks. After whitelist approval, borrowers must reach a 20% threshold in the first-loss capital pool. Then, 6 out of 9 randomly selected auditors (individuals holding GFI dollars for voting rights) need to approve the loan for the borrower to receive funds.

Predetermined Smart Contracts: Gearbox

Gearbox is a completely permissionless protocol relying on its credit accounts. These are predetermined smart contracts set for each type of pool, binding lenders and borrowers together, orchestrated by what is known as credit managers. Its credit accounts hold both the collateral deposited by users and the borrowed funds. The borrowed funds can only be used for whitelisted protocols and/or exchanged for specially permitted tokens to avoid extreme volatility. Since the funds deposited by users can be liquidated by smart contracts at determined risk levels, this protocol does not require credit assessments for users.

Gearbox protocol's engine

The credit account of the Gearbox protocol, as a predetermined smart contract, allows borrowers to engage in permissionless lending.

While Gearbox may seem like an interesting protocol, it is not a true form of low-collateral lending; it can facilitate everyday lending activities like auto loans and mortgage lending that TradFi currently stifles.

Off-Chain Integration: Teller

Teller has its own category because it differs from other lending protocols. Teller creates independent products for specific cases, and once the proposed borrower passes Teller's proprietary credit risk algorithm, it becomes a market platform connecting borrowers and lenders. In fact, USDC.homes on Polygon relies on Teller's credit assessment before funding loans from TrueFi. Teller's algorithmic credit risk protocol securely calculates credit and banking data off-chain, generating personal lending terms for users and routing them on-chain. Its inputs include provided collateral, risk premiums, lending to bank balances, lending to monthly net income, and lending to monthly income.

First-Loss Capital

While risk assessment models provide an initial protective wall for borrowers in low-collateral lending protocols, the "first line of defense" system offers additional protection against loan defaults for liquidity providers by using the liquidity of other stakeholders as "first-loss capital."

Stakers as First Loss: Maple Finance, TrueFi, Goldfinch

Maple Finance requires pool delegates to stake 8-12% of the pool's lending balance in MPL and USDC on a one-to-one basis. In this way, the protocol aligns the incentives of pool delegates with those of liquidity providers. Additionally, in the event of a default, these funds can easily be liquidated in the form of USDC. Besides pool delegates, anyone can join the reserve pool, further strengthening the protocol's first line of defense, earning 1/10 of the interest fees from the pool and a share of $MPL rewards. Similarly, Goldfinch relies on the supporters of the borrower as the first line of defense. Supporters deposit funds into the "primary pool" to earn higher nominal interest rates, as they can earn 20% interest from ordinary liquidity providers in the "senior pool" for taking on additional risk.

Goldfinch lending, source: Goldfinch

TrueFi relies on participants staking $stkTRU to gain voting rights, $TRU issuance, and protocol fees as the first line of defense. In the event of a default, up to 10% of $stkTRU may be liquidated to repay loans. Importantly, staking $stkTRU has a cooling-off period, meaning it cannot be withdrawn immediately, and users must wait until the cooling-off period ends.

Goldfinch has senior and junior pools.

Insurance Pool as First Loss: Clearpool

In each block, if a borrower defaults, a certain proportion of each pool's equity will be transferred to the insurance pool as a claims pool for liquidity providers. The current governance vote sets this at 5% of the pool's interest.

Spigot Function: DebtDAO

DebtDAO has developed a "Spigot Function" that controls the revenue contracts of borrowers and holds a fixed proportion of revenue while transferring the remainder to the borrower's treasury. In the event of a default, the protocol can claim the held assets to help repay interest and principal, acting as a debt collector. It can be argued that as the only true form of debt collection in DeFi, the Spigot can be applied to other protocols. The only limitation of this function is that its best application is for protocols and DAOs, given that if individuals know the Spigot will claim any funds sent to their wallets, they can create a new wallet to avoid penalties.

Analysis of various first-loss capital methods

A comparative chart showing the advantages and disadvantages of the three main first-loss capital methods implemented by secured lending protocols; staking, insurance pool, and Spigot Function. The evaluation criteria include whether the method aligns with liquidity providers' incentive mechanisms, allows for immediate repayment, provides the possibility of recovering 100% of the investment forever, and whether the method is automatically executed by smart contracts.

Types of Tokens Available for Lending and Pool Risks

In addition to risk assessment methods, the types of tokens provided for lending are also an important factor distinguishing low-collateral lending protocols. While individuals can use UniSwap to swap lending assets for preferred assets not provided by lending protocols, users may wish to avoid facing slippage, transaction fees, and increased smart contract risks. Given that USDC is considered the safest stablecoin in the DeFi ecosystem, each of the aforementioned protocols offers popular and reliable USDC. TrueFi offers USDC along with USDT, BUSD, and TUSD, while Maple provides USDC and WETH. Gearbox has the most variety, offering DAI, USDC, WETH, and WBTC for lending, as they have the ability to restrict funds for use only in whitelisted protocols.

Analysis of different assets available for lending on different protocols

A comparative chart showing the lending assets available under secured lending protocols. Available assets include: USDC, USDT, TUSD, WBTC, and WETH.

Another additional feature of low-collateral lending protocols is the provision of multiple lending pools for the same asset. The significant advantage of having multiple pools for the same asset is the ability to offer liquidity providers pools with different risk levels. Maple Finance, Clearpool, and Goldfinch provide multiple pools for each asset type based on the specified borrower, while TrueFi, DebtDAO, and Gearbox offer only one pool for each asset type. For example, on Maple Finance, users can deposit USDC into a pool with Maven 11 Capital at an annual interest rate of 11.3% or a pool with Alameda Research at an annual interest rate of 10.5%, with the difference in interest rates due to the estimated risk differences determined for each institution.

Potential Limitations of Low-Collateral Lending in DeFi

While all these protocols are exciting and driving the entire DeFi space forward, it is necessary to discuss some potential drawbacks of the current state of low-collateral lending. Most importantly, scalability is a core principle of growth potential for such lending platforms. While the concept of pool managers works well for protocols like Maple Finance, the success of these protocols is only because they are limited to institutional borrowers. If these protocols plan to scale and allow the general public to participate, the bandwidth required from pool delegates will far exceed their available time to review and assess borrowers. TrueFi also faces similar scalability issues, as the protocol requires each loan request to be approved by votes from $TRUE stakers, and if it were limited to institutional investors, it would not have to worry about this issue. While both protocols lack the potential for large-scale expansion, Maple's approach using pool delegates is superior to TrueFi's community voting mechanism simply because it allows for more educated decision-making. In reality, community voters with distributed responsibilities will not pay as much attention to details as an individual responsible for a pool.

On-chain 'algorithmic decisions' for credit scoring would be easier to scale to the masses if they could break free from institutional borrowers, but due to the limited availability of on-chain data, its scalability is relatively weak at present. Additionally, on-chain credit scoring may be vulnerable to witch attacks if it cannot be linked to a single real-world identity. In the context of on-chain low-collateral lending, a random individual could slowly build good credit through one account to eventually obtain large loans without a repayment plan. They could rinse and repeat this process through multiple wallet addresses. One way to avoid this situation is to implement KYC requirements, linking one wallet to one person. However, for security and personal reasons, individuals often have multiple wallets, with their assets distributed across these wallets, making the concept of "KYC linking a single wallet" somewhat limited and inefficient.

Moreover, for privacy reasons, individuals may not want their KYC details stored on-chain (even if the information is encrypted). In this light, off-chain credit scoring based on traditional finance not only contradicts the spirit of DeFi but also introduces the risk of storing KYC documents and personal information in a risky off-chain system.

An Optimistic Future

Even with various associated risks, one can see multiple low-collateral lending protocols thriving in this space. Maple Finance and TrueFi seem to be establishing themselves as current leaders, having a first-mover advantage in the institutional lending space, as evidenced by their current dominance in total loan issuance.

Total loans from various low-collateral lending protocols

The bar chart shows the share of each protocol in the overall DeFi secured lending market, based on total loan issuance as of April 5, 2022. TrueFi: $1.4 billion, Maple Finance: $1.26 billion, Goldfinch: $115 million, Clearpool: $41.7 million, Gearbox: $6 million.

Clearpool faces a tough battle against the two protocols mentioned above, but it may see some growth due to offering variable interest rates instead of the fixed rates of Maple and TrueFi.

Statistically, among all protocols, Maple appears to be the fastest-growing, being the only token with significant growth among the seven analyzed protocols. While one might blame TrueFi's generous incentive distribution mechanism for harming the token's price appreciation rather than the protocol itself, on the other hand, Maple deserves praise for its strong growth and highest utilization rate among all the aforementioned protocols.

Price performance of various on-chain secured lending protocols, source: Delphi Digital

As DeFi continues to spread into the mainstream, the integration of on-chain credit scoring with KYC requirements or Teller's off-chain credit scoring for retail investors will drive personal lending from a manipulated off-chain ecosystem to fair on-chain protocols. Off-chain credit integration systems are relatively scalable, and existing institutions seem open to it, as Teller has already established strong partnerships thus far. We expect to see individuals accepting on-chain loans in 1-2 years, even without realizing that Teller is the backend infrastructure.

In addition to credit assessment, legal protections will also be key to establishing trust between lenders and borrowers. TrueFi currently has legal agreements with their institutional borrowers, granting them the right to execute loans and recover any penalties in case of default if necessary. If this framework were applied to personal lending and executed on-chain through agreements like EthSign, it could further boost confidence in retail low-collateral lending.

While on-chain low-collateral lending is still in its infancy, its rapid rise and current success have laid a solid foundation in this field. It opens the door for similar protocols looking to enter DeFi. One can expect this concept to further expand into untapped consumer bases. Every innovation in this space brings us one step closer to DeFi disrupting TradFi. Perhaps one day, the rebels will storm the Bastille in Paris (a symbol of the oppressive rule of the French monarchy over its subjects)!