Ratings and Modeling
Risk Rating Methodologies
River will use methodologies developed by Kroll, Moody’s, S&P, Fitch, and Morningstar to determine the implied ratings for each tranche in a warehouse facility. These rating methodologies are publicly available and borrowed from by banks, credit funds, and asset managers when underwriting transactions. Rating agencies are typically not engaged unless the warehouse facility is meant to be publicly traded otherwise referred to as ABS or held by an insurance company.
Rating agencies are not formally engaged as they are slow, costly, and require borrower participation. Additionally, they tend to “check the box” rather than provide additive value in isolating or quantifying risk.
Lastly, lenders typically earn greater yield by not engaging a rating agency due to the benefits of private transaction pricing power.
Eligibility Criteria and Concentration Limits
Criteria will be put on the loans to limit the funding of assets that have inappropriately high risk or to create a pool that performs better in turn receiving a higher advance rate.
For example, if loans with terms greater than 2 years have a default rate of 10% and loans less than 2 years have a default rate of 1%, an eligibility criteria can be created to say loans greater than 2 years are ineligible for the facility. Alternatively, if the originator finds it important to do a small subset of these loans to further test performance and iterate on the product a concentration limit can be established to loans greater than 2 years of say 10%.
A concentration limit can give flexibility to the originator and protection to the investor.
Stress Testing
The core of the modeling will be stressing the pool loss assumptions to determine probability of repayment to lenders in each tranche and the level of overcollateralization needed.
A simple unit economic view of this stress is below. It looks at the revenue generated by the pool of assets in a base case and subtracts defaults and lost fees to get total collections or cash on cash of the pool. Based on the 3x Stress scenario, lending at a 95% advance rate would still allow the investor to break even.
Base Case
2x Stress
3x Stress
Asset Pool Fee Rate
20%
20%
20%
Asset Pool Default Rate
5%
10%
15%
Lost Fees
1%
2%
5%
Total Collections
114%
108%
100%
95% Advance Rate
Pass
Pass
Pass
Note: This simplification excludes funding interest, costs, and timing; we still run cash-flow waterfalls with loss magnitudes and timing vectors, recoveries, prepayments, and rate/utilization stresses by asset class.
Typical stress levels:
1× Expected Loss: Base Case 2× Stress: Typical recessionary environment 3× Stress: Deep/extended downturn
4-5× Stress: Severe, multi-standard-deviation tail event
These multiples are based on decades of historical performance data across the underlying assets as well as performance data from alternative data sources. River anticipates the srUSDS to be 3-5x Stress and jrUSDS to be 2-3x.
1x Expected Loss
This is based on actual portfolio performance, usually:
12–60 months of static pool data
Delinquency migration
Roll rates
Recoveries
Loss curves
Expected loss (EL) is the “base case.”
2x Stress: Typical Severe Recession Scenario
Across major credit categories, 2x EL aligns with what happened during:
2008–2009 Global Financial Crisis for prime & near-prime consumer loans
2020 Pandemic for unsecured SME credit
2001 Dot-Com + mild recession for personal loans
Historical notes:
U.S. credit card charge-offs increased ~70–120% during the GFC.
Unsecured installment loans often experienced 1.8x to 2.2x EL increases.
Auto loan losses increased roughly 1.5x to 2.0x.
SBA & SME portfolios rose roughly 2x.
3x Stress: Severe but Historically Rare
A 3x stress corresponds to:
Deep, prolonged recessions
Significant unemployment spikes
Sustained credit tightening
Major liquidity disruption
Historical analogs:
Subprime segments in 2008–2010 often experienced 2.5x–3.5x stresses.
Markets with strong borrower deterioration (e.g., early fintech vintages 2015–2016) saw ~3x multiple swings before stabilization.
A 3x assumption is generally tied to A/AA level like protection.
4-5x Stress: “Tail Event” or Multi-Standard Deviation Shock
A 4-5x stress is designed to withstand extremely severe, multi-standard-deviation environments.
Reference points:
In consumer finance, 4x EL typically corresponds to >5σ events if losses are normally distributed.
Historically, only a handful of collateral types have ever approached 4× EL increases, usually tied to structural shocks rather than credit deterioration (e.g., 2020 pandemic shock for certain unsecured portfolios, 1998 Russian crisis for emerging markets, or sudden regulatory changes).
Rating agencies treat 4-5x as an extreme tail, not something observed frequently in practice.
Typically used to size AAA senior tranches.
Covenants and Performance Triggers
Triggers and Covenants are incorporated to wind down a warehouse facility if performance is trending in a poor direction or there are issues with the originator of the assets. These include
Performance triggers: delinquency, cumulative net loss, repayment %, minimum excess spread
Corporate/operational covenants: minimum liquidity, TNW, leverage, servicing standards, reporting cadence, KYC/AML, audit and collateral exam rights.
Events of default: payment/covenant breaches, fraud, insolvency; servicer EoD enables transition to backup servicer.
Performance Triggers typically include advance rate step downs if the cohorts don’t perform to expectations. For example, advance rates will step down to provide more overcollateralization if cohorts underperform early on to provvide more overcollateralization in the structure.
Additional Key Risk Factors
River will look at a number of other risks that are important in determining the quality of the opportunities. Those include:
Originator/servicer strength: operational, financial, and governance quality directly affect collateral performance.
Legal structure: verify true sale, bankruptcy remoteness, and perfected security interest of the SPV assets.
Counterparty risk: assess exposure to banks, trustees, hedging agents, and liquidity providers.
Macroeconomic and sector cyclicality: consider how downturns or secular changes could impair obligor performance or asset resale values.
Data integrity and surveillance: ensure ongoing performance monitoring and trigger management throughout the deal life.
Last updated