BID® Daily Newsletter
May 11, 2010

BID® Daily Newsletter

May 11, 2010

LOOKING BELOW THE SURFACE


We heard all the criticism regarding President Obama's Supreme Court nominee Elena Kagan. Not to be political, but we agree that experience is important, as we urge all Americans to look below the surface. Do we really need another Harvard educated, NY lawyer on the highest court in the land (no offense to graduates of Harvard, NY or banks that make land loans)? More importantly, what experience does she really have ruling over important matters? This is why we believe the right pick is -Paula Abdul. She is CA born, educated at CA State and has much deeper experience passing judgment from the bench. She more fair than Kara, more compassionate than Simon and carries more gravitas than Randy. In short, a solid pick.
Another area where we need to look below the surface is when working with Loss Given Default (LGD). As a quick reminder, LGD is the loss a bank takes on a loan (after foreclosure and property disposition that includes all lost principal, interest, fees and costs associated with working out and selling the loan less any proceeds derived from the sale of the property). If you are using a risk-adjusted loan pricing model (or credit stress model) like ours, the model will feed users the LGD after asking for the terms of the loan, guarantee, location and collateral. Users of our models know that LGD can range from an average of 5% for cash/securities collateral to 70% on some special use properties. For community banks, the range of expected LGD is usually around 25% to 50% of the loan's original principal amount.
While good information, it is important to look even deeper to understand what drives collateral quality. In looking at collateral performance (correlation to loss) since 2004, in order, the largest drivers outside of cash flow quality have been: 1) geography, 2) year of origination, 3) property type and 4) starting LTV. While that is admittedly only a short period of analysis, the point is nonetheless an important one to make and to understand.
Geography plays a huge roll. For example, states that have seen fast appreciation like WA, have LGDs that are 20% to 30% higher than an average growth state like OR. By zip code, the relationship works in much the same way, with speculative, high development areas such as AZ (around a 50% LGD), getting hit harder than infill-metro type regions like DC (10% LGD).
Next, and most often overlooked, is the year of origination. Here banks need to be cognizant of where they are in the business cycle and adjust pricing to compensate. Amortization, appreciation, seasoning of leases and other factors, all help contribute lower LGDs for loans originated before the run up in commercial real estate prices in 1999/2000 and again in 2004 to 2007. For example, for loans originated in 1999, LGDs for the generic community bank loan has been running around 8%. Meanwhile, LGDs for loans originated between 2001 to 2004 have been in the 30% range. Loans originated from 2005 to 2007 have been running close to 55%. So far, loans originated in 2008 have seen LGDs start to peak at 60%.
Property-type is also important, but that really goes back to the quality of cash flows. Properties with multiple revenue contracts, such as multifamily, tend to do dramatically better than single contract industrial properties.
Finally, LTV, the ratio that we as bankers often spend the most time analyzing, has very little impact on LGD. The difference between a 60% LTV loan at origination and a 70% LTV loan matters about 33% less than where the property is located or when the loan was originated.
Unfortunately, this market has given us superb data to calibrate LGD. By studying what impacts LGD, bankers can become more adept at pricing and managing credit. In the meantime, consider our Court pick. We admit Paula isn't all there sometimes, but neither are some of the other justices either we would argue.
Subscribe to the BID Daily Newsletter to have it delivered by email daily.