Where to start?
How to choose a model for analyzing DAO Loans
Challenges with data availability, quality and relevancy for DAOs means that the DAO lending market and borrower credit risk analysis is nascent [1]. Default data, a key building block towards on-chain credit ratings, is simply not there for example.
Given the differences between on-chain and non-crypto business models and risk profiles, there isn't the option of using default data and a ratings scale from TradFi to which you’d map shadow scores for DAOs.
A statistical modeling approach backed by AI could be considered but it’s too early. It would only be possible where data is available in sufficient quantity, quality and relevancy.
We have to start somewhere so for the launch of Debt DAO we’re offering this Borrower Analysis Framework to the community. It’s a first stab at putting together a standard for analyzing DAO credit risk. When the market develops we'll see templates for specific sectors and business models. Right now it’s too early.
The framework is a combination of a qualitative, judgemental approach and a review of a DAO’s financial data. It will also lean on quantitative analysis where required for the financial risk assessment of Treasury assets and collateral.
We envisage that as the market develops borrowers will be grouped into risk buckets whereby borrowers with similar stand-alone ratings should achieve similar borrowing rates. We predict that this will eventually lead to some kind of ratings scale.
The assessment process may follow a crowdsourcing approach to some extent until a level of consensus standardization is reached.
[1] The risk that a lender won’t get paid back in full is universally known as Credit Risk. There are other risks involved of course which need to be considered and measured where possible but they are all abstracted here under the overall Credit Risk of a Borrower and/or the Loss Given Default of a secured loan.