Last month we discussed the meteoric rise in community bank commercial real estate loan probabilities of default. When underwriting, that is just part of the equation. The other major component is estimating loss given default, also termed loss severity.
Over the past 30Y (as far back as our data goes), losses given default equate to an average centered around 30% and 50% of the principal amount of a given loan. The severity of loss largely depends on a loan's structure, the type of collateral and guarantees. This amount also includes non-accrued interest, transaction costs of selling a property, legal fees and other associated workout costs. Here is the problem â€" all major loan pricing models that we have found (including our own and those at use at national banks) base their loss severity on the same 30Ys of data. During this period of time, there was only one brief chronological period (between 1989 and 1992) when property values declined year after year - a scenario that appears highly likely in 2009.
For most of that 30Y history, property values also had healthy appreciation. In fact, the average community bank loan had an average real property appreciation of almost 9% per annum. That's pretty healthy. That means that our 40%-ish loss severity that most community banks have experienced is against a historic backdrop of very positive price appreciation.
While we are taking steps to update our Loan Pricing Model, our initial analysis (based upon the current path of GDP, unemployment and consumer spending), shows commercial properties values will fall 20% on average for community bank loan collateral over the next 3Ys. If true, this would drive the loss given default up by an average of 50%. Keep in mind that this is on top of higher probabilities of default. This combination is toxic, as it increases the expected loss rate. For example, a loan on a retail property with a 1.25x debt service coverage ratio and 75% LTV back in 2007 most likely went from a 0.9% expected loss to now a 1.6% expected loss. Remember, our base case analysis assumes that the historic econometric relationships between GDP, consumer spending, unemployment and commercial capitization rates hold true. There are many that feel this relationship no longer holds and the results might be worse.
Following years of aggressive property price increases, a period of deleveraging, global economic problems and tighter credit, many real estate analysts now estimate that property price declines could easily be in the 50% range from 2007. If true, loss severity will increase causing loan losses to further skyrocket. A typical "deep recession" scenario now shows that 1 every 5 commercial loans could go into default, resulting in expected losses in the 75% range. This would mean that expected loses climb to as much as 10% - a loss level that many banks would be unable to support without massively increasing capital. Loans that are particularly sensitive to higher loss levels include those that have balloon maturities coming due between now and 2010. Health care, retail and special purpose properties are all at the higher end of this potential range.
As we learn how to handle dramatically higher probabilities of default and loss given default rates, having a plan to increase reserve levels to the 3% to 5% range would be a good idea. We don't know how bad credit performance will get, but having a prolonged period of negative appreciation is a real possibility.