Unboxing BendDAO #05 Understanding Risks - Part 3

deep-dive on the Borrowing Side

In the previous newsletter, we explored how improper liquidity management from depositors can expose utilization risk to BendDAO. However, as BendDAO is a two-sided market with both depositors and borrowers, we also need to examine the credit risk of borrowers to fully capture the risk that BendDAO bears. In this post, we will delve into BendDAO's credit risk management to understand how the protocol manages the risk associated with lending.

Table of Contents

1. What is Credit Risk Management?

2. BendDAO Credit Risk Management

3. BendDAO Loan Risk Analysis

Loan Grade and Liquidity Risk

4. Conclusion

---

What is Credit Risk Management?

Credit risk management involves assessing the likelihood of default and the potential loss that could result from an obligor or counterparty default. Every lending protocol, including BendDAO, is subject to credit risk in its lending operations. However, most Web3 lending protocols lack the proper tools to keep track of credit risk.

BendDAO Credit Risk Management

To manage credit risk associated with lending, we must assess the borrower's side of the risk. This involves understanding the utilization rate to ensure no imbalance within the liquidity pool. By looking at two dates, Post bank run and the most recent calendar month, we can see how BendDAO’s reserve has changed over time.

From the end of August to the end of January, BendDAO's liquidity pool increased by over 50%, with a corresponding increase in the amount borrowed. However, the ratio of ETH lent out from the pool has decreased compared to August. While this suggests that BendDAO may flourish without borrower risk, it is not quite the case.

BendDAO measures borrower behavior risk based on the floor price of the NFT collection used as collateral, along with the amount of the loan borrowed. The metric that ties this in is called the health factor. The higher the health factor, the less likely the loan will go into default.

However, for a more comprehensive look at risk, we need to understand the health of individual loans, which can be analyzed with the borrower's payback history and the desirability of the underlying collateral. To fully capture how borrowers can expose risk to the protocol, loans have to be assessed into different grades, adding relevant factors on top of the health factor.

We graded all loans based on the type of collection, health factor, and the amount of defaults for individual wallets, giving a grade from 1 (least likely to default) to 5 (most likely to default). Each loan has the potential to move up/down the grades depending on the factors that we input.

BendDAO Loan Risk Analysis

To compare the worst condition for the pool and the most recent, we analyzed BendDAO's loan receivable at the end of August 2022 (post-bank run) and at the end of January 2023 (most recent full-calendar month).

By January 31, 2023, we can see that the majority of loans have been concentrated in Grade 2 loans. Although there was a decrease in the number of Grade 1 loans, there was also a decrease in the number of both Grade 4 and Grade 5 loans, balancing out the soundness of the loan receivable. Overall, we can say that the health of the loans has improved compared to August. But how does this impact depositors, and why is the grade of the loan an important metric for depositors to consider when participating in the protocol?

Loan Grade and Liquidity Risk

Grading the loans surfaces additional signals to determine the likelihood of depositors getting their deposit back(Such a practice has been the norm when analyzing credit risk in TradFi). In its simplest form, it shows you an expectation on the percentage of the loan that can go awry; the lower the grade, the higher the chance of default. But to get a vivid picture of what loan grades can signal, let’s imagine a scenario, possibly the worst case scenario(called ”stress test”), where the largest number of depositors wouldn’t get their deposit back.

The stress test would look like this: all loans are defaulted, and a bank run is triggered. Liquidation of defaulted collaterals are quickly taking place through the auction. As expected, the collaterals from higher grade loans are quickly getting recovered at its full price, and as time goes by, collaterals from the lower grade loans are left unsold. Seeing this occur, those with the largest deposit are the quickest to act, as they have the highest stake. They take out their deposits, claiming the proceeds from the sale of collaterals from higher grade loans. Then what’s left is a set of lower grade loans, with low chance of recovery. This is what the majority of depositors who deposited relatively lower amount is left to swallow.

What should we takeaway from here? While such extreme case is unlikely to occur, the % of lower grade loans indicate exposure to higher borrower risk. Additionally, the higher the number of depositors who deposited smaller amount, compared to those who deposited the most, the likelihood of larger number of people getting impacted is higher. Knowing this, let’s look at the periods we reviewed above (August 2022 and January 2023).

The size of the box for each grade, represents the loan amount that are categorized as the respective Grade. From August 2022 (above) to January 2023 (below), we can clearly see that the amount of Grade 5 loans had decreased. But the distribution of depositors per their deposit amount remains the same(The shape of the blue graph looks similar). You can see that there’s now smaller number of people exposed to the risk of swallowing the bad loans.

We can interpret this as BendDAO’s borrower risk indeed getting reduced from August 2022.

Conclusion

This week, we discussed how accounting for borrower risk is essential in understanding the overall risk that BendDAO assumes in its lending system. If borrowers face a higher risk of default and liquidation, it also puts depositors at risk, especially if the liquidity pool is concentrated in a few wallets. However, grading loans based on default likelihood can serve as a directional tool for depositors to withdraw their funds if necessary. As the NFT space grows, assessing credit risk of wallets will be crucial for minimizing risk in web3 lending protocols. Although the practice is still in its early stages, it is expected to become more sophisticated as the industry matures, as it enables stronger risk management, which in turn will invite new stream of liquidity.