The credit market in traditional finance is for the most part inaccessible to many - a luxury beyond the grasp of 1.7 billion people in the world. For starters, you’d need to possess a credit score, and/or sufficient assets (read: collateral) or source of income to sniff a look from the banks’ credit assessors. This is also not to mention the excessive red-tape and lengthy processes involved for loan approvals by traditional old-fashioned banks:-
Bank vs. DeFi Loans by Mark Cuban
Decentralized lending attempts to remove as much of these barriers as possible, allowing anyone anywhere to secure loans on platforms such as Aave (AAVE), Compound (COMP) or Maker (MKR). With the first generation of decentralized lending protocols though, there is no need for bank accounts nor tedious credit checks as users needed to overcollateralize their lending positions. But therein also lies an issue in this form of trustless, overcollateralized loans - collaterals. Users who are underbanked / unbanked likely cannot afford the necessary crypto collateral (or even possess much of any valuable assets) to begin with. For these segments, the crypto market needs to provide an alternative avenue - enter uncollateralized crypto loans.
DeFi Lending started with Overcollateralization
Decentralized money markets such as Aave are operated trustlessly, with loans being executed within minutes via smart contracts, sans third parties, and KYC processes. In order to maintain this trustless nature, DeFi loans are predominantly overcollateralized to provide protection against defaults. This means that users will have to deposit an amount that exceeds the value of any loans. For example, a collateral factor of 70% would mean that $100 worth of collateralized deposits only allows users to borrow up to $70 worth of assets. The collateral is locked to manage counterparty risks, funds that could have otherwise been deployed elsewhere to be more productive. Overcollateralized loans are simply not an efficient allocation of capital.
Then there’s also the usage of overcollateralized DeFi loans and its target audience. As noble as it sounds to serve the unbanked, the truth of the matter is that overcollateralized borrowing does not reach the borrowers in need. Rather, it’s more commonly used as an instrument for various financial strategies, such as entering a leveraged long position through recursive lending and borrowing. This has also allowed many crypto maxis to access immediate consumption without selling off their positions. For instance, if you were bullish on ETH but also needed cash for iPhone 20, instead of selling off your ETH, you could deposit your ETH on Aave and take out a loan on USDC / USDT to buy the phone.
That sounds great and all but the question remains - how can I get a loan if I didn’t already have enough money to begin with? After all, that is precisely why I needed the loan in the first place! Imagine asking for a $200k loan to buy a house, only to be told to first make a deposit of $300k.
The World of Undercollateralized Lending
This is not to say that over collateralized loans are not useful, but rather, they serve an entirely different purpose from what the average borrower needs. The requirement of overcollateralization with crypto prevents a vast majority of borrowers in the world from participating. By lowering or even removing the need for collaterals, crypto loans can become more accessible to a wider pool of audience. The idea sounds like an unattainable holy grail but do not dismiss it just yet - there are actually already DeFi protocols offering such service to their users.
An undercollateralized loan refers to a loan that is not fully / at all collateralized. Should the loan default, the collateral (if any) will not be able to fully cover the principal. By and large, the market of undercollateralized lending (including non-collateralized lending) can be segmented into 8 different types:-
© Clear Chain Capital
According to Defillama, lending and borrowing is the second-largest segment in DeFi right behind Decentralized Exchanges (DEX) in terms of total-value-locked (TVL), boasting just over $44billion or ~28% of the entire $191billion crypto market TVL (excluding borrows) at the time of writing.
© DefiLlama
While that is certainly impressive, it still pales in comparison with the $11 trillion credit market residing within traditional finance. Figuring out undercollateralized (or uncollateralized loans) will be key to the next stage of evolution for DeFi lending, allowing users to access without (too much of) the usual red tape and bureaucracy gatekeeping loan access in traditional finance.
Currently, undercollateralized lending is still far from being the go-to solution for lenders and borrowers, even within the crypto community. As of 19 February 2022, the TVL of undercollateralized lending protocols is only sitting at around $1.2 billion, not even 3% of the total lending TVL. It is clear however that the community favors third-party assessment protocols at this point, with Maple Finance, TrueFi and Goldfinch leading the pack, and Wing Finance rounding up the top 4 spot as a credit native credit score platform. Bear in mind however that this does not take into account flash loans, which is also a type of no-collateral loan.
For the purpose of this article, we will be focusing on flash loans, third-party risk assessment, and crypto native credit score - the three most popular methods of undercollateralized lending.
Flash Loans
© CoinGecko
Flash loans are the most widely used under collateralized loans in the crypto industry. In 2021 alone, flash loan transaction volume on Aave exceeded $3 billion. As its name suggests, a flash loan involves borrowing and repaying the loan in a “flash”, with everything happening within seconds in the space of one transaction. Both borrowing and repayment will have to happen simultaneously in order for the transaction to go through, which means that there will be minimal risk for both parties since any deviation will just mean that the transaction is reversed.
Flash loans have racked up quite a bad reputation due to their involvement in multiple DeFi protocol exploits, but the bulk of flash loans (~80%) are actually employed for arbitrage purposes. The design of flash loans makes it ideal for arbitrageurs looking to capitalize quickly on the price difference between two platforms, all within the span of a single transaction. Other uses also include collateral swapping and strategic self-liquidation.
That being said, a flash loan is designed for niche purposes with no application in conventional loans or serving the average borrower.
Protocols: Aave (AAVE), dydx (DYDX)
Third-party Risk Assessment
Outside of flash loans, third-party risk assessment is by far the most popular undercollateralized lending method. Interestingly, this form of loan mirrors the traditional finance process flow closely, with the caveat of being decentralized and on-chain.
© CoinGecko: High-level loan process for bank
The debt products in banks are normally verified, evaluated, and approved by their internal credit team/credit committee, taking into account the risk profile of the applicant. Similarly, protocols such as Goldfinch and Maple Finance add a third-party credit assessor between lenders and borrowers to perform risk assessments of loan applications. Users who want to act as credit assessors have to stake a certain amount of tokens in order to qualify, and in the case of a default, the collateral that they staked will be first-in-line to be slashed. On the flip side, they get rewarded for successful loan repayments. On Goldfinch, in order to verify the legitimacy of applicants - that is to confirm whether the borrower does what they claim to do and they are indeed who they claim to be - there is even a second layer of checking by third-party auditors to verify the information provided by the borrowers. Again, this highly resembles the maker and (multiple) checker process that financial institutions adopt.
Unlike overcollateralized forms of loans, third-party risk assessment introduces human-level checks to help screen against high risk applicants and potential frauds. This opens the door for borrowers to access much-needed funds without collateralizing any assets, paving the way for DeFi towards personal loans, microfinance, and even prime brokerage. Borrowers are also set up to build up their on-chain credit scores and trustworthiness, which should facilitate future on-chain lending and borrowing even on other protocols.
The flip side is that introducing manual checkers also means running the risk of human errors and a lengthier process. Taking a loan from Aave is a matter of minutes, but having one approved on Goldfinch for instance could take up to 48 hours for the auditing alone. Other key challenges with this structure are bootstrapping a network of reliable credit assessors, including arming them with the necessary tools and data to conduct their assessment dependably, as well as attracting sufficient liquidity (or TVL) to support the uncollateralized loans.
© Maple Finance
© Goldfinch
So what are the rates like on these platforms? On Maple Finance, you can earn up to 22.5% as a lender by supplying USDC. Goldfinch on the other hand offers more options (i.e. borrower pools) that yield from 7.7% - 21.25% on your stables, even going up to as high as 39% when you factor in GFI rewards. As high as they seem, it’s also important to note that the interest rates for borrowing are proposed by the borrowers, subject to approval and acceptance by the lenders on both platforms of course.
All in all, if executed well and ignoring the irony, this could be a viable, decentralized replacement of the existing loan approval model traditionally employed by banks.
Protocols: Goldfinch (GFI), Maple Finance (MPL), TrueFi (TRU)
Crypto Native Credit Scores
“Get a credit card and build up your credit history, you’ll need it for loan applications in the future”. This is one of the most common financial wisdom dished out by elders to young adults around the dinner table, but what exactly does it mean?
Imagine applying for a loan with little to no credit history. From the perspective of the bank, it would be extremely difficult for them to assess you since there is no data to rely on. Conversely, holding a credit card and having a sprawling history of paying your accounts on time will make a good display of your credit history - a reliable borrower who is capable of handling credit responsibly. As a result, lenders will be more inclined to offer a loan at a better rate.
Similarly in crypto, a wallet address with a rich record of on-chain activities ranging from loan repayment, yield farming, governance participation etc. could be used to construct a crypto credit profile. Such a profile is unique to the wallet address and can potentially be used across multiple platforms to help facilitate loan assessments, just as how your credit history in traditional finance is stored, accessed, and assessed by financial institutions. There are however some major problems with this approach.
Turning our lens back on the average borrower, especially those that are already unbanked, there is simply insufficient on-chain data for most users to date. The pseudonymity of wallets also means that if a user were to default on a loan, he or she could just switch to another wallet come loan application time. Ironically (again), one of the solutions commonly spouted for this is to connect real-world IDs on-chain with a single wallet (i.e. off-chain integration), but this is very much against the spirit of decentralization championed by DeFi since the beginning. One way to get around this is to simply introduce zero-knowledge-proof (zk-proof) of off-chain data.
There are quite a few projects already working on this, with some such as Wing Finance and EasyFi running their own lending protocols as well. If successful, this approach will allow lending platforms to properly curate and assess a large trove of on-chain profiles, similar to The Fico Credit Score used by lenders in traditional finance.
Protocols: Wing Finance (WING), LedgerScore (LED), EasyFi (EZ)
As shown, there are also various other methods of undercollateralized loans serving other needs, below is a quick summary of what each entails:
- Off-chain Credit Integration - importing off-chain data (e.g. KYC information and credit score) to support assessment and underwriting of undercollateralized loans
- Personal Network Bootstrap - invite-only borrowing to restrict public access, allowing lenders to control the borrower pool
- Real World Asset Loans - real world assets represented on-chain via NFTs, and to act as collateral for a crypto loan
- NFTs as Collateral - NFT-backed loans which also enable improved liquidity for NFTs
Closing Thoughts
DeFi lending as a whole has a long way to go before it can go toe-to-toe against traditional credit markets in size, with the undercollateralized loan segment being further away still. The silver lining to this is that there is still plenty of space for growth as we ramp up towards mainstream adoption. In its current iteration, there are still limited use cases for DeFi lending to serve the unbanked. Sure, there is a strong argument to be made that not everyone who needs access to credit should be granted said access, but at some point partially secured / unsecured lending on DeFi needs to be introduced to capture a wider audience.
As pointed out, there lies some irony in undercollateralized lending and borrowing protocols adopting traditional finance processes, but this is a necessary step in bootstrapping the user economy. Crypto native credit scores seem promising in laying down the foundation for a fully on-chain and decentralized lending future, but that is only plausible when crypto has reached escape velocity for mass adoption. As of now, the average user simply does not have enough on-chain activities to build up a crypto credit profile.
Ultimately, the proof, as they say, is in the pudding, and for undercollateralized loans, we really need to observe these protocols (and their credit models) go through a few credit cycles to determine their reliability and soundness. During times of bull market or relative stability, credit markets may seem to operate swimmingly well, but the real stress test will be during bear markets and times of crises. However with this segment still being relatively small, the chances of cascade failures affecting other DeFi protocols seem relatively slim for now.
In conclusion, we’re still in the nascent stage of undercollateralized loans and it’s hard to say which mechanism will survive the test of time to come out on top. It is just as likely that a new innovative solution crops up to overtake the race. What we do know is that there is certainly no harm in starting early, starting now. So what are you waiting for, “get a wallet and build up your credit history, you’ll need it for loan applications in the future”!
Dillon is an Operations Associate at CoinGecko with a weakness for 3-digit APYs. As an ex-consultant, he now finds joy in laughing at consulting memes unironically.
Or check it out in the app stores
