Credora: Where We Are (Part 1 of 3)

Credora is a technology company that enables superior credit risk management in crypto and ultimately traditional markets. Our product facilitates comprehensive and real-time counterparty credit risk assessments using provably private and neutral infrastructure.

Capturing real-time data distinguishes Credora from traditional lending businesses. Credora aggregates KYB information, financial statements, and real-time portfolio information on borrowers, running only a specific set of computations on granular data while keeping it completely private. For trading firms, trade and portfolio information is incredibly sensitive. We analyze this information directly from the venues, and cryptographically validate privacy.

The advantages of real-time data are particularly clear in credit shocks like the one experienced as a result of FTX and Alameda. We think it is an important time to reflect on the industry, our performance, and our expectations for the future.

Ecosystem

Over the past year, Credora sought partnerships with credit protocols to create a more transparent credit infrastructure. Protocols allow capital allocators to better understand their ultimate exposure. In comparison to the CeFi lenders, the construction of a loan portfolio is publicly visible. Alongside credit protocol partners, Credora quantifies the risk of an individual borrower through the distribution of credit metrics. The Credora rating methodology is public, and currently, we are on the 6th version of it.

Credora operates a Maple Solana pool (operating as a Pool Delegate), and provides credit metrics for multiple protocols (operating as a Credit Oracle), including Clearpool, Ribbon, Atlendis, and dAMM.

As a Pool Delegate, we rely on scores and ratings from Credora as the primary inputs for capital allocation decisions, cementing the value of data-driven lending. As a Credit Oracle, Credora helps attract capital to protocols by allowing borrowers to distribute scores externally.

Institutional capital allocators tend to prefer to underwrite based on more granular information, which is exclusively available on Credora. Therefore, many of them consume information through the Credora platform, and then allocate via protocols as the effective contract and settlement infrastructure. This user flow is a core component of the vision we are building towards, as Credora provides a credit data layer on top of multiple protocol infrastructures (who dictate the terms of a loan, manage the settlement, and promote transparency).

It’s worth taking a step back and comparing this developing market structure to the industry mechanics that dominated prior to recent events. CeFi lenders (Genesis, BlockFi, Celsius, etc.) were the main sources of credit for the industry, allocating primarily to trading firms, given borrowing demand is particularly strong and consistent. These lenders operate as principal lenders, borrowing unsecured from retail and institutional capital allocators, and deploying capital using their own internal risk models. The 3AC event first shed light on their lacking underwriting practices, in certain instances relying on a single NAV attestation as justification for large lines of unsecured credit.

We can identify three core weaknesses of the principal lender model:

1. No available credit assessment of the principal lender. The market tolerated credit exposure without any due diligence of the lender’s practices.
2. Weak credit assessments done by the principal lender, including a reliance on relationships and internal attestations from borrowers.
3. In certain market environments, the principal business model (spread capture) incentives incremental risk taking. Idle capital carries a cost.

Methodology

An overview of the Credora credit methodology provides more insight into how the risk of a borrower is quantified to encourage a more dynamic and transparent underwriting process. We are constantly taking feedback from lenders and borrowers looking for new data sources to improve further. The methodology is public to ensure that capital allocators have an understanding of what is being scored in the assessment. Credora’s role is defining the model, and systematically following it across all borrowers.

It is important to note that the methodology is an expert-driven model. Factors are identified by credit experts, and weighted according to expected correlation to creditworthiness. Very simply, there is not enough data today to have anything other than an expert-driven model, but given the richness and consistency of the data we collect, we ultimately expect to rely more heavily on statistical analysis and machine learning methods to determine factors and weights.

The main sections and primary factors are listed below. More detail is available in our documentation.

Due Diligence & Operations (20%)
  Due Diligence
  Borrow History
Financial Analysis (40%)
  Quality of Financials
  Performance
  Operating Margin
  Maximum Drawdown
  Interest Coverage
  Current Ratio
Risk Monitoring (40%)
  Visible Liquidity
  Equity
  Leverage

In thinking through how a borrower’s rating evolves over time, it’s worth noting the data sources Credora relies on in scoring each factor. The Due Diligence & Operations section relies on relatively static information, gathered through the initial onboarding process (KYB) and credit assessment process. Financial Analysis relies on recurring but static documentation, uploaded on the Credora platform monthly or quarterly. Risk Monitoring is effectively continuous information contextualized by recurring information. For example, Visible Liquidity benchmarks real-time asset values against reported assets from financials.

In summary, there are infrequent changes in Due Diligence & Operations factors (e.g. age of the entity), periodic changes in Financial Analysis factors, and there are very frequent changes in Risk Monitoring factors.

Transparency is a core belief of ours. We aim to balance the utility of external scores, transparency regarding the methodology, and information privacy (preventing reverse engineering and calculation of specific financial variables). In the coming weeks, we will publish a more granular breakdown of our credit methodology, including the active curves that are used for scoring various factors.

Join Us

We are actively looking to expand our credit and development teams. If you are data and model oriented, and believe that credit risk can be accurately quantified and assessed in real-time, we want to hear from you!

Additionally, we invite industry capital allocators to reach out and discuss our methodologies.