Sample Size
Another factor not considered in the blanket acceptance of industry standard operating expense studies is the size of the sample. As discussed in Part 1 of this post, Dollars & Cents appears to have a threshold where a certain number of responses for an expense category are necessary for it to be reported. Fortunately, most studies of this type report the number of properties that have been used to provide that information, so we have an ability to judge whether the data is sufficient to be statistically significant. Strangely, few people seem to notice this stat.
Same goes for IREM’s Income/Expense analysis. Selecting another random page, each grouped set of columns has a “Buildings” column that appears to report the number of buildings where respondents provided that line item on their financial statements. This example has 92 buildings as part of the the sample… a really good number! The individual income and expense line items range from 1 to 90, but about 90 percent of them show 70 observations or more, so we can see that the sample is good enough to be credible and statistically significant. I can’t say there’s much gray between the black and white of the printed word and page for this study… which is a good thing. Same goes for BOMA’s Experience Exchange reports.
Let’s take an example. The IMI Financial Operational Benchmark Study for Marina Operators is an indispensable source for marina operating data, but the number of marinas used as part of the sample are much lower than that reported by similar studies by ULI, BOMA and IREM. Their 2000 study had 13 respondents from Region III (13 Midwest states) and 13 from Region IV (13 West Coast states). With California in Region IV, do you think that 13 marinas are statistically significant enough to be reliable? Would you accept the Region III results with only 13 marinas reported from 13 states or an average of 1 per state? I’d say the odds are against them even being a reliable and reflective sample of the population, as they say in statistics speak. I’m not trying to bash the study… I’m simply pointing out that you need to know how well it represents what it is supposed to represent to rely on it. When there aren’t many “observations” in a sample, don’t treat it like the gospel.
Do the Math
Getting back to our example from Part 1, it looks like our Sherlock Holmes sleuthing is paying off. Let’s say we select the Average column as best reflective of the property we’re comparing it to. Have you tried to see how well the numbers add up? Going back to my random page selection from Dollars & Cents, Total Advertising and Promotion is $0.98 per square foot of GLA (gross leasable area) on average. Underneath that and indented are components of Advertising and Promotion. Advertising is $0.39, Promotion & Special Events are $0.21, Christmas Decor/Events are $0.09 and Marketing Administration is $0.17. Doing the math, that’s $0.86, not $0.98 per square foot.
That leads me to two thoughts. First, almost all of the “Total” columns suffer from the same problem. Second, Merchants Association was blank, which was part of Total Advertising and Promotion. You’ll find lots of these situations… but don’t conclude the difference is Merchants Association! You’ll see why in Part 3 when I go over a real example and the response of one of the industry study designers to my questions.
“Totally…”
I can’t help but read this title using Crush’s voice from Finding Nemo. I’m sure he’d read the paragraph above and say “Like… Dude… when do I use Total?” Well, Crush, Total isn’t really a total. I’m sure you just as confused as poor Crush. You see, the Total columns is based on the the number that is reported from the sample respondents. It is not the sum of the line items that comprise it, which is somewhat misleading given that they are presented in the Dollars & Cents study as indented below the total they would logically represent. What this means is that if you want to use total advertising and promotion to compare to your subject’s total advertising and promotion line item, that’s the way to compare it. You don’t add the sub-items up because you’d find in most cases that they add up to a different total.
So let’s take this a step further. You’ve got a financial statement in front of you with advertising, promotion, Christmas Decor/Events and Marketing Administration, or their equivalents. Should you compare these to the individual line items or add them up and compare them to the Total advertising and promotion expense? Your guess is as good as mine. It’s not explained well at all in the industry standard studies.
The Operating Expense Ratio
Given the issues discussed above and those presented in Part 1, you might want to fall back on the operating expense ratio (net operating income divided by effective gross income) as the most reliable line item. Now that’re you’ve studied and are ready for the final example, do you think the operating expense ratio is stand-alone, is based on the sum of the total columns or the sum of the individual subitems for each total column? If you find out the answer, let me know. It has to be calculated somewhere because it is the ratio of two numbers, but whether it is calculated from each sample financial statement submitted or is calculated from column items is anyone’s guess.
Stuffing Santa’s Stocking
So maybe I’ve made you a big fan of using the average in an industry standard operating expense study. Before you start getting cocky, ask what the average is an average of. Studies like BOMA’s Experience Exchange Report are great because they give you tabular statistical data on what comprises the average. You can see the average square footage per office user in a particular category, the occupancy rate and lots of other details. That’s very useful for pinpointing how well the property you wish to compare to the study reflects the properties that are reported. If your number is significantly different, maybe consider one of the other categories like high, low, Upper Decile, Lower Decile, or whatever is used in the study. ‘Food for thought.
Conclusions
To base important financial decisions on industry standard operating expense studies is a little like being spun around and having to play darts blindfolded. If you understand the caveats of each study, you’ll at least be pointed in the direction of the dartboard. You may find that to score a bulls eye, you’ll have to contact the designer of the study and have them answer your questions and give you advise on how best to interpret their numbers. That’s what I did and you’ll see the result in Part 3.
John Simpson, MAI
