The Timing of CMBS Losses

CRE Finance World, Winter 2013

The Timing of CMBS Losses An update on past industry studies in light of recent CMBS performance and the Great Recession David Nabwangu Xuedong Yang Senior Vice President, CMBS Data Analyst, CMBS DBRS Inc. DBRS Inc. A frequency; in that the bell-shaped pattern observed in with two important differences. 2,Upon first examination of recent CMBS performance statistics,we observed the same bell-shaped pattern in frequency of defaultwith respect to loan seasoning as the aforementioned past studiess CMBS losses continue at historically elevated frequencyand severity, two important observations emerge in theindustry loss data and suggest a break from past experience.The first relates to the effects of loan seasoning on default past landmark CRE studies is not evident given recent experience First, a significant portion of the defaults seem to be more front-loaded and may not hold in the future. The second related point, is that than the past studies, with a greater frequency of loans defaulting refinance risk is real, and will likely prove to be an influential concern in years two to five. The cause of the front-loading of our instances for legacy CMBS through 2017. of default is due to the recession that began in 2008 and the larger number of loans issued between 2005 and 2007, which Our primary observation is that both the shape of the default were subject to the environment created by the recession early frequency curve, as well as the frequency of refinance (versus on in their lifetimes. term) defaults, is mostly driven by the life of a loan relative to an event of recession or other adverse macroeconomic event. Second, while the past studies (especially Lancaster et al.) tended to trail off after year seven, our observations demonstrate a steady For this article, we define default as all loans with one of the increase in default frequency between years eight and ten; caused following characteristics: (1) a reported special servicer transfer by the prevalence of refinance defaults in our data set, that occurred date, (2) greater than 30 days delinquency or (3) a reported loss during and after the Great Recession. to the trust (where no delinquency or special servicing transfer date was available). Refinance (balloon) default is defined as any Indeed, after we controlled our dataset for recession, we found no default that occurs within 180 days of a loan’s original scheduled strong bell-shaped relationship between frequency of default and maturity date. Term defaults are defined as all other defaults. loan seasoning. Two landmark studies of CRE proposed that defaults occur in a The importance of this finding is evident when we break out timing bell-shaped pattern as loans season. Lancaster, Butler, Mayeux of loss by vintage and compare it to its relationship with the and Frerich, (Wachovia, April 2006) published data comprised adverse economic circumstances that began in 2008 with effects of fixed-rate CMBS conduit loans originated between 1995 and that linger today. 2005, which suggests a bell-shaped relationship between the frequency of loan default and seasoning: “Defaults increase at The series of charts below indicate the years since origination a steep incline and peak in years five and six, and then steadily (the first x-axis) in addition to the actual year of default (the second decrease over the remaining years in a symmetric bell-shaped, x-axis) with the stress period of 2008 through 2011 shaded. manner.”1 This symmetric bell-shaped curve is also observed in Nearly all charts show a steep increase in default frequency during the earlier Snyderman/Esaki studies, which were based on life this time, irrespective of loan seasoning or the year of origination. insurance loans originated from 1972 to 2002. Figure 2 Figure 1 Timing Default by 2007 Vintage Timing of CRE Default CRE Finance World Winter 2013 60


CRE Finance World, Winter 2013
To see the actual publication please follow the link above