How to Measure Nonprime Loan Returns vs. Risk

Many financial institutions have excess liquidity and are experiencing net interest margin (NIM) compression. Ideas to enhance yield through riskier investments or expanded credit boxes are discussed more often. Whatever the motivation, it is important to approach this choice in a dispassionate, quantitative manner, using the same risk-pricing framework used elsewhere in the institution. It is rarely correct to declare entire, relatively common asset classes off-limits unless the institution lacks the framework to understand them.

Define your Risk Measurement Metric (RMM)

Every day, institutions decide whether to hold risk-free assets like US Treasuries or seek higher expected yields in riskier loans. Intentionally or not, they are judging that the extra after-cost yield compensates for the extra risk. Standard deviation, return variance or Beta are common RMMs, but the essential point is that every institution has a defined one. Institutions without an explicit method need to “discover” the RMM they are already using implicitly.

The Asset Reward for Risk Ratio (ARRR)

Start with a simple principle: a risky asset’s expected yield, minus servicing cost, should exceed the risk-free yield.

R(e)1 − SC > Rf

To compare two opportunities, divide the excess return by each asset’s RMM:

ARRR = ( R(e) − SC − Rf ) / RMM

If Asset 2’s ARRR is higher than Asset 1’s, Asset 2 offers the better risk-return profile. When the RMM is return standard deviation, the ARRR is simply the Sharpe ratio. In LoanStreet’s worked example, a nonprime auto pool offered a superior return for risk versus a prime auto pool because its ARRR was higher.

Institutions can also set a Minimum Reward for Risk Ratio (MRRR) and screen out any opportunity that falls below it.

Questions to answer before entering nonprime

  • Can we form an informed view of the expected return?
  • Can we reliably estimate the risk on an apples-to-apples RMM with our existing assets?
  • For participations: do we assess the originator’s underwriting and the servicer’s capabilities favorably?
  • For originating: do we have the resources to service efficiently?

Institutions that cannot estimate return and risk should avoid the asset class. But many that can are needlessly constraining themselves. Buying participations in nonprime loans lets them earn higher risk-adjusted returns without large investments in nonprime origination and servicing.

[Online Event] LoanStreet's 7th Annual Virtual Forum, Nov 18th & 19th @ 1PM