A look at 134 commercial real estate loans that were just sold last month with recent appraisals reveals some interesting data points for banks. The sale price was higher by about 10% than the appraised value. In looking at the details, appraisers placed a heavier weight on current capitalization rates versus investors that tend to look more forward. This begs the question – how accurate are commercial real estate appraisals and should banks be basing loan amounts on them?
This recent evidence is in line with past performance over the past 25 years, as our research indicates that, on average, if done within 6 months of an orderly liquidation event, appraisals are accurate to within a 10% variance. This is good news for banks. Further, as can be expected, appraisals fair worst in markets that rapidly adjust. In market downturns, appraised values tend to be an average of 14% higher than actual values. In strong, rapidly appreciating markets, such as the environment we are in now, appraised values tend to be about 3% higher.
If you take the history over the past 25 years, in both up and down markets, the mean is that appraised values are about 3% higher than sale prices. In terms of property type, appraisals on industrial properties tended to be more accurate (2% mean), followed by office (3.7%), then by retail (5%) and multifamily (5%).
What this means for banks is that this error factor needs at least to be understood if not taken into account when underwriting commercial properties. In general, appraisals tend to be fairly accurate, but not perfect. They tend to lag the market and are biased in the direction of the market. While this is common sense, the extent to which this is true and the factor of error is not commonly known in banking circles. If the market is improving, a bank's appraisal is most likely too conservative. Conversely, in recessions, appraised values tend to overestimate the property’s true value. As such, to get more accurate, banks may consider loosening their underwriting in improving times and tightening in decreasing times. In addition, given the appraisal error volatility by property type, banks may want to adjust loan-to-value ratios accordingly.
How do you know when and how much to make an adjustment? In terms out, the data can point us in the right direction in order to be more quantified in our underwriting. Statistically, after running multi-factor correlations on a variety of factors, it turns out that four of these data streams account for 52% of the movement. These four are: GDP, unemployment, interest rates and commercial real estate total return.
It turns out, given the relationship of some of these variables to past results; banks may be changing underwriting standards exactly at the wrong times. For example, the total return on commercial real estate and appraisal errors are positively correlated. The coefficient indicates that for every 1% increase in returns, appraisal error increases 2%. Thus, in times of high return, banks should become more conservative, not less.
This is the exact opposite of how many banks operate. When times are good and growth is rapid, banks tend to loosen underwriting standards and lower allowance for loan loss. This might be a mistake as unless a bank is adjusting for appraisal error, a bank would be lowering reserves right at the time when they should be increasing them.
To a similar extent, high growth in GDP, rising unemployment and low, but rising interest rates also induce significant errors. In today’s environment, low GDP and unemployment changes, mean more accurate appraisals. In addition, and no surprise, the largest errors were right at the point that the economy starts to deteriorate or improve.
Given that credit quality can be the most significant driver of profitability, the more accurate a bank can be in its underwriting, the more it can leverage its balance sheet and manage risk. Over the next couple of years, you will see more and more banks change their loan-to-value limits so that they become sensitive to the business cycle. In addition, we will see banks more sensitive to where they are in the cycle and will adjust loan pricing accordingly. In this manner, banks will be able to better correlate reserves with risk and produce smoother, more predictable earnings.
Submitted by Chris Nichols on March 03, 2014