BID® Daily Newsletter
Feb 13, 2007

BID® Daily Newsletter

Feb 13, 2007

CRE LOAN UNDERWRITING


We recently reviewed more than 200 commercial real estate loans originated from 2002 to 2005 in order to compare the quality of projected property cash flows against estimated cash flows at the time of underwriting. These credits came from 75 different banks in 17 different states. Underwriting among banks differs greatly. In approximately 25% of the cases, banks failed to create a sensitivity analysis of cash flows. Despite the fact that this is the largest determinant of default, it was not analyzed. By excluding such analysis, lenders are left with little in the way of understanding how future changes to interest rates, lease rates, absorption, vacancy and capitalization will affect the property and related credit risk. For banks that did produce cash flows, it is no surprise that 43% of the time, actual cash flow came in lower than forecasted (at least when we did the analysis at the end of 4Q). In a perfect world, this number would be 50%, as half the time a bank would overestimate and half the time they would underestimate. Given that many underwriters are influenced by the borrower (many use the developer's own projections of cash flows as a base case), it is not unreasonable to assume that the distribution would be asymmetrical. What is notable is that in a material amount of the cases (8%), cash flows were 50% to 100% below the expected case. For comparison, we only saw cash flows 50% to 100% above the base case in less than 1% of the loans reviewed. This research indicates that not only is bank underwriting naturally biased, it is also skewed. The largest driver in this skew was that properties did not lease up as originally planned (or more supply came to market than forecasted, slowing lease up and appreciation). Single family development, hospitality and multifamily suffered the greatest negative differences between actual and projected cash flow, while self-storage, retail and office suffered the least (all in order). While we found a difference by property type, we did not find much of a difference by structure. The underwriting of fixed rate loans was just slightly more accurate (by 6%) than the underwriting of floating rate loans. This is probably because of the added variability inherent in floating rate exposure. We also looked at origination in a period of falling rates (such as 2002 and 2003) versus a period of rising rates (2004 and 2005). This indicated a slight positive bias to the latter (when rates are falling, there is less of an underwriting bias). Structural elements like maturity, prepayment structure, covenants and the inclusion of an interest only period also made little difference. While this research indicates that banks need to strive to be more accurate in their loan underwriting, the important point is to know that there is an inherent bias in projecting CRE cash flows (and as a result, pricing). Banks that are aware of this can devote more resources to monitoring, validating cash flow assumptions and improving understanding of market dynamics in an attempt to mitigate underwriting risk. In times of fast appreciation and strong cash flows, underestimating does not result in a material difference in loan performance. However, as rates stabilize or rise, credit quality becomes more of an issue. This difference can quickly translate into actual losses if banks are not careful.
Subscribe to the BID Daily Newsletter to have it delivered by email daily.