Subprime Loans, Making Sense of a Growing Problem

Subprime mortgages are very quickly becoming every lender’s nightmare and as SunGard Bancware Erisk Monthly points out, taking a balanced view on the issue is an important step to understanding how the chaos surrounding high-risk, high-interest loans began and what the future holds for the spiraling situation.  

As Shahram Elghanayan explains, a step in the right direction for a lot of banks is using “quantitative risk and capital modeling to adopt a risk-adjusted ‘through-the-cycle’ view of the spreads available in subprime.” Although this certainly does not guarantee that subprime portfolios will turn a profit every time out of the gate, it does give the bank a heads up about potential future profits, once the calculated losses, risk capital costs, funding and operating costs are accounted for.

Elghanayan notes that risk capital costs are “a particularly important factor because they reflect the volatility of the portfolio’s loss rate – how far losses might depart from their historical average or ‘expected’ rate as economic conditions deteriorate.”

As an experiment, the monthly publication recently dissected several portfolios of subprime auto loans.  The purpose of the experiment was to try to uncover the truth behind the type of through-the-cycle credit spreads banks should be making. It was also aimed at trying to find the main causes of long-term success and failure with the subprime loans.

For instance, a BB- rated institution with a portfolio of $3bn in subprime auto-loan assets that had an average probability of default of 15.1% and a loss given default of 55% was found to have a credit risk capital cost of about 7.88% of total assets (using some ‘middle of the road’ risk factor assumptions).

As the two factors around which the study revolved around (probability of default {PD} and facility loss given default {LGD}) were changed, so were the results. For example, as the assumed correlation between the two rose from 0.20 to 0.60, the credit capital results rose from 6.65 percent to 9.73 percent.

It is this type of correlation that has plunged many home owners further into debt. Many subprime loans offer reasonable rates in the beginning but the often drastic increase in interest often hits homeowners hard, raising the probability of default factor significantly.

Elghanayan also writes that lower credit scores are often a key factor in increasing risk measures and, as a result, an increase in required capital. When maintaining PD and LGD correlation at a level of 0.40, the study found the people with a credit score of 600-619 required a credit spread of 7.22%. Those with scores of as low as 560-580 required a spread of 12.13%.  The credit spread used in the publication’s research includes the cost of maintaining risk capital at a typical bank hurdle rate plus the cost of expected losses, but does not include funding and non-interest expenses.

Elghanayan notes that a fat spread over funding costs with any given subprime segment could indicate either a profitable business activity or simply be caused by extremely high capital costs.  

He adds that “rather than blindly pursuing market share or notionally high spreads, banks must learn to use this data to identify the subprime business that offers the best risk-adjusted margins – and the best long-term return for shareholders.”

To subscribe or visit go to:  http://www.riskcenter.com