Unboxing BendDAO #03: Understanding the Risks - Part 1
Understanding the risk factors, their size, the likelihood of their realization, and how the company manages them are just as important as understanding revenue. Now that we understand the A to Z of BendDAO’s revenue system, we’ll go over everything about the risks of investing in BendDAO in the next few newsletters. This week, we’ll set the framework to best understand different types of risks. Next week, we’ll guide you through how to manage those risks with the underlying data.
Before diving in, let’s lay out the three main risk factors associated with BendDAO:
Default Risk: Even after BendDAO recognizes its revenue, as per the “accrual basis accounting” there isn’t a 100% certainty that that revenue would be realized (= the actual “cash” comes in).
Defaulted Asset Liquidation Risk: Risk of not being able to recover the unpaid loan by selling NFT after it has been liquidated
Utilization Risk: Risk of lending more/less than the appropriate ratio
By looking at the potential income stream without understanding these risk factors, you are missing half the picture. We’ll unpack each of these factors so that you can understand the full concept, and in our next newsletter, we’ll provide you with the numbers you can use to actually manage the risk.
Subscribe to Unboxing Labs to continue to get these insights and analyses of different DAOs!
Table of Contents
1. Default Risk
Case 1: Strategic Default
Case 2: Cash Flow Default
How does BendDAO hedge default risk?
How is this similar or different from TradFi?
2. Defaulted Asset Liquidation Risk
How does BendDAO hedge liquidation risk?
How is this similar or different from TradFi?
3. Utilization Risk
How does BendDAO hedge utilization risk?
How is this similar or different from TradFi?
1. Default Risk
BendDAO’s P2Pool model works very similarly to how a commercial bank operates in TradFi (Read our previous newsletter to see the comparison). This means BendDAO is also exposed to similar risks. One of the major risks of such an operation is default risk, that is, the risk of loans not being repaid by the borrower.
Why do these defaults happen in the first place? There are two primary scenarios observed in BendDAO’s situation:
when the price of the collateralized NFT drops suddenly (strategic default),
and when the borrower’s financial status worsens (cash flow default).
Identifying default cases is important because this becomes the key to understanding what kind of data we should look for and how BendDAO hedges default risk (stay tuned for the analysis based on more data). Let’s go over how these two different scenarios can influence the financial statement through accrual basis bookkeeping. You’ll get a sneak peak of the deeper analysis to be presented i n our next newsletter.
Case 1: Strategic Default
If the price of the collateralized NFT drops in price, the incentive for the borrower to pay back the loan significantly decreases (it becomes cheaper to buy another comparable NFT instead of paying back the loan). So for this reason, it’s important to capture the value of the collateralized NFT and keep track of them. In other words, BendDAO needs to regularly re-evaluate the collateral price of the NFTs on its Balance Sheet and identify the risk exposure.
Let’s take a look at one example. For Loan ID No. 8. On July 31, 2022, a user borrowed 21.3 ETH against an NFT with a floor price of 68.4 ETH. However, on August 31, 2022, the loan amount increased to 21.82 ETH (with added interest) and the floor price decreased to 67.45 ETH.
In this case, the change of floor price (0.95 ETH) is recorded in the income statement as “Loss on Re-evaluation.” This later influences the balance sheet, which becomes an indicator for investors to identify the default risk. To create the most financial statements, Unboxing Labs’s system runs re-evaluations on a daily basis, reflecting price changes in NFT assets and the turnover rate of BendDAO loans.
Case 2: Cash Flow Default
Separate from the risk from changes in the underlying collateral price, there may be scenarios where the borrowers’ ability to repay has changed. To manage this risk, we can look at the wallets involved with loan receivable, along with the balance of user wallets at the point of liquidation.
We can also zoom out and see the distribution of the borrowers is concentrated to a small number of borrowers or spread out. If BendDAO is heavily dependent on a few wallets, it means the impact of user default will be much more severe than when the loan receivable is more distributed.
How does BendDAO hedge default risk?
While defaults cannot be prevented all the time, BendDAO has several policies that minimize the chance of default. There are two “safety nets” that BendDAO has set in their protocol as a DAO to hedge against the default risk:
whitelisting of NFTs to be used as collateral,
and LTV (Loan to Value) selection.
A. Whitelisting of NFTs
Selecting NFTs to be used as collateral is a sensitive matter because the value of the NFT collateral is directly related to the default risk. In the case of an NFT being liquidated, BendDAO has to make sure the liquidated NFT can be sold via auction sale.
BendDAO follows a voting mechanism as a DAO to add in a new collection. They consider variables like the collection’s market cap (Supply x Floor Price), Liquidity (Trade Volume, Transaction Volume) and more, when deciding to whitelist the collection.
Moonbirds were their latest voted collection back in September 2022 to support a total of 9 NFT collections as collateral.
B. Changing LTV (Loan-to-Value)
LTV tells you how much a borrower can borrow depending on how much his or her collateral is worth. For example, if the LTV of an NFT collection was 50% and you wanted to borrow 100ETH, the NFT being used as collateral would have to be worth at least 200 ETH.
Setting the LTV is crucial to risk because if the value of the NFT falls close to how much one initially borrowed, the borrower has less of a reason to hold the NFT and to pay back the loan plus interest, causing a default.
BendDAO currently operates their protocol with LTV between 40~50% with the exception of BAYC (60%), which is a conservative approach considering the collaterals being used are blue chips. Setting a conservative LTV refrains a borrower from defaulting on their NFT.
Note that while whitelisting and LTV selection help control for risk, it does not get rid of the risk. The risk that remains after controlling for it with the two factors is tied to the liquidation risk.
How is this similar or different from TradFi?
In TradFi, institutions refer to credit scores to determine whether an individual is qualified for a loan. The system calculates various factors into account, such as a person’s payment history, the amount of money owed, and the length of credit history. These credit ratings determine whether someone is approved for credit as well as the maximum loan amount he/she can borrow.
Even after filtering out people with bad credit ratings, there still might be a chance of defaults. And TradFi institutions set up a policy to guarantee recovery through warranty. The default history will also be permanently recorded in the creditor’s credit report, thereby impacting the credit profile.
However, this system is inapplicable to DeFi at the current stage. Users are anonymous on blockchain and can own multiple wallets at once. In other words, an individual does match a single wallet. With the absence of a credit rating system, the best BendDAO can do to hedge risk is to adopt NFT collection whitelisting and LTV policies. However, when DeFi matures and enables open finance, we might be able to implement a credit policy in the future to solve information asymmetry in financial activities.
2. Defaulted Asset Liquidation Risk
Just like default risk, all lending protocols are exposed to defaulted asset liquidation risk. This risk is the risk of not being able to recover an unpaid loan by selling NFT after it has been liquidated. BendDAO is exposed to this risk when a user defaults and the collateral NFT is sold on auction.
A metric called the “recovery ratio,” which is the liquidation price/loan amount, is used to check the status of such a liquidation risk. When an NFT goes into default, BendDAO takes the NFT into auction and sets the price of the NFT equal to the amount of the remaining debt, caused by the defaulted loan. This means that the liquidation price would be at least equal to the loan amount, setting the recovery ratio to be at least 100%.
Thankfully, there hasn’t been a case where BendDAO was not able to sell off a defaulted NFT through an auction. As of January 31, 2023, there has been a total of 171 NFTs sold via auction all of the NFTs were sold within an average of 36 hours, implying that the liquidation risk didn’t impose a threat. Yet, as BendDAO adds in more collections and the lending pool’s volume gets bigger, there is always a chance that this liquidation risk exposure will cause damage to the protocol. This is why there is still a need for the protocol to hedge against the liquidation risk.
The average liquidation count and the duration on the auction is different for collection. As we add in more collections, we can predict that the amount of liquidation count itself can increase and hence the chance of duration on auction increasing as well.
How does BendDAO hedge liquidation risk?
In order to hedge liquidation risk and properly recover loan amounts, BendDAO devised two ideas:
the health factor
and the auction system.
A. Health Factor
According to BendDAO, the health factor is an index showing “the safety of a collateral NFT against the borrowed ETH and its underlying value.” The higher the health factor, the less likely a liquidation would happen. When the health factor goes below 1, the collateral can get liquidated and put into an auction by a bidder.
B. Auction System
As mentioned above, what’s special about BendDAO’s auction system is that NFTs (with a health factor < 1) are manually put to an auction by bidders. Only when the first bidder finds a liquidatable NFT and bids a price, it is put to an auction and is, therefore, fully “defaulted.”
While this incentivizes BendDAO users to actively monitor the health factor, there are two problems with the system. Firstly, there might be NFTs not put to an auction even after the health factor reaches below 1. Secondly, the auction might fail to work when market conditions worsen, and the price of the NFT suddenly falls below the remaining loan amount. This is also the case when the interest rate largely increases, and the remaining loan amount outweighs the actual price of the NFT.
In fact, during the bank run of August 2022, BendDAO implemented a short-term parameter improvement (95% to 70%) to encourage NFT liquidation.
How is this similar or different from TradFi?
Traditional Finance is very similar to BendDAO in that TradFi handles collateral asset disposal through a court auction. The remaining debt is claimed by the debtor to be paid off. What’s different from BendDAO is that, if the remaining debt is not paid off after it has gone through auction or the debtor, the debt is claimed by the entity that gave out the loan, which is the bank. Even though the system seems unjust and the bank would take a loss from paying off the loan, it would prevent a bank run by filling in the lack of liquidity that could occur.
3. Utilization Risk
Utilization rate (U) is an indicator of the availability of capital in the pool. As of January 31, 2023, the U is 26%. This is a metric that indicates how active BendDAO loans are. A U that is too big or too small will indicate exposure to risk.
For example:
U = 0
would mean borrowing is inactive.U = 100
would mean there would be no capital left in the cash reserve and the depositors would not be able to withdraw the amount they deposited, causing a bank run.
In fact, here is a great post by BanklessDAO covering BendDAO’s bank run in August 2022, for those of you who might be interested.
The logic behind utilization risk exposure is based on the balance between the demand to deposit and borrow. In fact, there were two cases where the utilization risk was exposed when the U was high and low:
Low U case: “There was a time” when BAYC/MAYC were able to be staked to receive interest in APE coins. This led to a decrease in BAYC/MAYCs being used for loans, causing fewer loans overall, driving the low U.
High U case: In September 2022, Ethereum’s protocol changed from Proof-of-Work (PoW) to Proof-of-Stake (PoS), calling it the ETH merge. The demand to hold ETH grew because being able to stake ETH gave a high interest back to the depositors. This led to ETH being deposited less to BendDAO, driving the high U.
How does BendDAO hedge utilization risk?
In order to control the utilization rate, the borrow rate and deposit rate have to be controlled.
In the case of underutilization, the borrowing rate has to be increased to drive demand for borrowing
In the case of overutilization, the deposit rate has to be increased to drive demand for depositing.
However, these rates are shaped by the collective decision of DAO (we won’t go into the details around the interest slope, as that lies outside the scope of this article), so it has significant friction to move in any direction. There were actual cases of such voting in August 2022 when a bank run occurred in BendDAO (In order to prevent the depositors from withdrawing, the DAO came together to increase the base rate from the initial 10% to 20%), and again prior to Ethereum Merge, where the DAO came together to change the optimal utilization rate from 80% to 45%, and doubling the slope from 8% to 16%. But as you can see, these are rare incidents.
How is this similar or different from TradFi?
TradFi also sets the interest rate for a similar reason. However, the Central Bank is in charge of managing the interest rates rather than individuals or the bank. The Central Bank considers all the variables in the market to decide on the “market interest rate,” and then respond to the risk factors that could occur before lending/borrowing.
Conclusion
By understanding the three risk factors explained above (default risk, the defaulted liquidation risk, and utilization risk), you will now have a clear picture of how risk comes within BendDAO’s revenue-generating model. Moreover, you can also imagine how BendDAO controls and mitigates different risk factors.
But it is better to show than tell. In the Part 2 article of the risk series, we’ll examine actual numbers with real-life risk scenarios. Let’s break down the financial statement of BendDAO to analyze every risk-related data in depth. This will provide you with extensive insight into understanding a DAO’s financial health.
Want to be the first to read the article? Subscribe and join Unboxing Labs.