The first half of 2007 was a busy time in bank commercial real estate lending. The problem with this is that a lot of 5Y interest only loans were made back then. Of course, a large amount of 10Y loans (both amortizing and 25Y amortizing) were also originated. This gives us an interesting set of data to use as we compare how a loan???s structure can play a role in defaults. We took a look at over 900 loans to see if any conclusions could be drawn.
Back in 2007, times were good and commercial property cash flow was increasing (due to higher rents). Higher rents meant higher property values and banks were tripping all over themselves to make loans and get a piece of the action. Looking at our set of 900 loans, we found the 5Y interest only is the most problematic, due to a high balloon balance, aggressive underwriting, property price declines and limited refinancing options. Looking at the performance of 5Y loans that have matured year to date in 2012, it turns out that only about 42% were paid off without a problem. The rest are either having problems, already been modified or have been extended. We compared this subset to a group of 5Y loans with a 25Y amortization and found that while performance is slightly better (38% of these structures have been paid off without a problem), the difference is within our margin of error (so we refrain from drawing conclusions).
However, when we compared a group of 5Y balloon loans with the group of 10Y balloon and 10Y amortizing structures, we found a performance difference where we could draw conclusions. The 10Y fully amortizing loans that came due in 2012 had about a 90% performance rate (that is 90% paid off without any issues), while the 10Y balloon structure had a 70% rate. We quickly add that most of this difference is a result of the vintage or the year of origination, as 2002 was a time when we were a little less than halfway through the last real estate bubble (based on price appreciation). In fact, 2002 was a time when asset prices were about comparable to where they are today in many sectors and in many geographic areas.
We also looked at loan performance of both the 5Y and 10Y loan terms over the past 5 years and found much the same relative performance. The high-level takeaway is that a 10Y loan structure looks to be safer because it gives the borrower more time and options to work out of problems. Ironically, many banks would rather have a 5Y balloon loan than a 10Y balloon loan, a structure that the data shows will work against you in terms of both payment performance and profitability. All things equal, this difference in delinquency performance translates for this data set into roughly about 2.20% more in pricing that a bank should have demanded to originate a 5Y balloon loan compared to a 10Y fully amortizing one. Of course, the caveat is that all things are rarely equal and the quality of cash flows and DSC are more important and will have a larger impact on pricing than the structure.
As a side note, it is interesting that of the 5Y loans that have run into problems refinancing, almost 50% of the extensions have been between 2Ys and 5Ys, with a mean of 3.3Ys. About 33% of the loans were extended 5Ys or more.
Over time we will have more data, which will allow us to draw better conclusions. That will eventually allow us to find out what we really want to know - whether it would have been better to make a 10Y fully amortizing loan in 2006 or a 5Y balloon loan in 2011? Only time will tell.
Until then, the data shows community banks should consider the 10Y over a 5Y structure and include strong credit covenants that force negotiations along the way as needed should credit falter. So far, given our current data set, doing so will place banks in the best position to see a successful loan conclusion.