Maxwell O. Ramsland, Jr., MAI, CRE
Ramsland & Vigen, Inc., 302 W. Superior Street, Duluth, MN 55802
Shannon M. Luepke
Ramsland & Vigen, Inc., 302 W. Superior Street, Duluth, MN 55802
Malls and their modus operandi are fascinating in their complexity. The first enclosed mall, Southdale in Edina, Minnesota was built in 1956, and 36 years later within five miles of Southdale the Mall of America was built in Bloomington. These two malls demonstrate a complete cycle in mall design and efficiency, and exhibit considerable contrasts in design, layout, materials, etc.
The typical regional or super regional shopping center, and more recently the life style center, has changed over the past 50 years. The catalyst for this transformation has been for obvious reasons: reduced costs, better customer and tenant efficiency, and most importantly, to maximize their revenue producing capability. For the smaller inline tenants as rents increased their demised square footage decreased; gone were the back rooms and storage areas, store fronts shrank as did store depths. The typical inline space went from approximately 2,500 square feet to ±1,800 square feet. Changes have also taken place for the anchor department store. In the 1960s and 1970s, the typical anchor space was near or exceeded 200,000 square feet, often approaching 300,000 square feet. By the late 1990s, anchor store sizes decreased to 100,000 to 170,000 square feet on average, with the mean being approximately 150,000 square feet.
What has not Changed at the Mall is the Profit Motive 
Most, if not all retail activity is driven by optimizing retail sales. As measured by the PriceWaterhouseCooper/Korpacz Real Estate Investor Survey, mall inline sales in excess of $600 per square foot of GLA represent a Class A+ mall, the very best, while average sales approximate $300+/- per square foot of GLA. With the successful mall’s first objective being to maximize retail sales, tenant space size and layout are designed to increase selling areas and minimize non-selling areas. As store sizes evolved, leasing patterns and rental rates also evolved, with rent being directly tied to sales productivity. For example, if the overall sales level of a particular mall is generating sales of $400 per square foot, and the average base rent is 7% of sales, the base rent then approximates $28.00 per square foot. Some, of course, would be higher, e.g. smaller stores near the mall’s center court, food court, kiosks, etc., and some would be lower, e.g. large inline tenants, book stores, etc. This brings up the existence of the secondary mall occupant: the anchor department store.
While the average mall might contain 300,000 square feet, and have 150 inline tenants, anchor department stores, on the other hand, are single tenant operations usually containing 100,000 to 170,000 square feet each. Anchor department store sales average about $180.00 per square foot, and rents are traditionally between 2% to 3% of sales, resulting in rents of between $3.00 and $6.00 per square foot when they are owned by the mall and leased to a particular retailer (Macy’s, Dillard’s, Sears, etc.).1 Many anchor department stores, however, are independently owned, and they are often bought and sold independent of the mall.
Critical Information Sought by Real Estate Analysts  
When a mall sells, the critical information sought by real estate analysts is the net operating income (NOI) and the sales price. For example, a mall that sells for $100,000,000, and has a net operating income of $7,500,000, will have an overall capitalization rate (OAR) of 7.5% (NOI of $7,500,000/$100,000,000 sales price = .075). Industry professionals constantly monitor the marketplace for mall sales, for the resulting OAR of a sale is a key component in valuing a mall and a leading indicator of existing valuation trends. From the above example the derivative of the transaction can be transformed to units representing the income and price per square foot in units of GLA. The corresponding NOI is then: $25.00 per square foot of GLA and the sale price is $333.33 per square foot. The OAR remains the same: 7.5% ($25.00/$333.33 = .075). Reducing the sale’s components to units of the GLA allows the analyst to make comparisons based on the industry’s most relevant common denominator.
Department Store’s Capitalization Rate versus that of a Mall’s Capitalization Rate  
Recently, a discussion of mall sales and anchor department stores took place at an informational meeting between representatives of Dillard’s department store and the Williamson Central Appraisal District (WCAD) in Georgetown, Texas. When the question of a department store’s capitalization rate versus that of a mall’s capitalization rate came up, Alvin Lankford, RPA, the WCAD’s Chief Appraiser, observed that those companies publishing national real estate investment survey reports feature surveys that are essentially perceived to represent only the inline portion of the mall. He went on to ask, “What impact on the mall’s OAR does the presence of one, two or more anchor department stores have when they are owned by the mall and included in the sale?”
An Excellent Question 
Mr. Lankford is correct that PriceWaterhouseCooper/Korpacz and RERC investor surveys do track capitalization rate information for national shopping centers, but both surveys are silent on anchor department store capitalization rates or any residual impact they may have on a mall’s overall capitalization rate. A published proxy for the department stores might be in the category of the national power center, which typically has a high ratio of single story big box properties, many of which approach the size of anchor department stores. But, power centers do not enjoy a mall’s possibly dominant location, nor do they have the department stores’ physical characteristics of larger sizes, and/or second or third stories, a condition anathema to most big box retailers. So, does the mall’s superior location offset the department stores unique physical attributes? Anchor department stores when they sell, and if leased, typically have capitalization rates of 100-200 basis points higher then those of similarly aged national shopping centers. Higher department store capitalization rates are frequently attributed to the slower growth expectations of retail sales, obsolescence due to size, multi-stories, existing trade dress, and physical recapitalization issues.
However, the inescapable conclusion is that there exists a symbiotic relationship between a mall and its anchor department stores. One needs the other to maximize sales and profits. However, the difference of a 300,000 square foot, 150 inline tenant shopping center versus a 600,000 square foot shopping center with two or three anchor department stores presents an entirely different set of investment and/or risk considerations for the investor, appraiser or assessor.
Getting back to Mr. Lankford’s question: if our hypothetical 300,000 square foot inline mall had a 150,000 square foot department store included in the sale, what impact on the mall’s NOI and OAR would the anchor department store theoretically have on the mall diagnostics?
Potential Theory
Our theory would suggest that more square footage, a lower NOI per square foot, and a lower price per GLA, ceteris paribus, would result in a higher OAR. Expanding our earlier example and adding an anchor department store, the hypothesis bears this out, as shown below.
Mall:                      300,000 sq. ft. x $25.00  =   $7,500,000
Anchor:                150,000 sq. ft. x $ 4.00   =         600,000
Total/Average:    450,000 sq. ft. x $18.00  =    $8,100,000
The mall now has a 450,000 square foot GLA, and an average NOI of $18.00 per square foot or, in other words, 50% more square feet of GLA, and a 28% reduction in rent per GLA. The theoretical adjustment can be calculated as follows:
Adjusted OAR   =       .075  x  450,000   x  $18.00
                                                    300,000      $25.00
                           =          .075   x  ( 1.5  x   .72 )
                           =          .075   x   1.080
                           =          .0810
                                    Say: 8.1%
The adjusted NOI would increase the resulting overall capitalization rate (OAR) approximately 60 basis points (.0810 – .0750), or an approximate 8% increase in the OAR. Employing the same methodology, the corresponding OAR rates for two department stores is 8.67%, and three stores is 9.30%; again approximately 60 basis points per anchor department store. Thus, the theory could conclude that as the total income increases, the NOI per square foot decreases, the sales price per square foot decreases, and the corresponding OAR increases.
This is the simple theory of the department store influence on mall OAR. The follow-up question is: Is the theory supported by market evidence? The authors’ contention is that there is ample market based evidence to conclude that the existence of one, two, three or more anchor department stores in a mall will have an incremental affect on a sale’s OAR.
The Model 
The process to test the department store capitalization rate evidence will include three variations on an econometric model. The empirical evidence will include 53 observations of shopping center sales that have taken place between 2000 and 2006. The independent variables include the date of sale, sold GLA (includes inline space and the anchor department stores, if any), the effective age of the mall (i.e., the year built with adjustments for upgrades), the number of anchors, and location. The dependent variable is the overall capitalization rate (OAR).   

Presented below are two graphs of the relationship of the model’s NOI per GLA and sales price per GLA experienced as an OAR. The first graph identifies the relationship of price per square foot of NOI and sales price. The second graph demonstrates the dichotomy of the model color coding the components of the mall, e.g., no anchor, one anchor, etc., and the OAR. It may be obvious to the observer that the lower NOI values represent malls with anchors, and the higher values represent essentially inline space. The model has a tendency to self-compensate for size, rental rates and other disparities. The graphs are presented below.    

First Graph Analysis of Data   

   Second Graph Analysis of Data   


The mall sales observations are summarized below.  It is interesting to note that the unadjusted data tends to confirm the theory by itself. This is not surprising insofar as the average age is approximately 20 years. 

          Difference in
    No. Ave. NOI/Sq. Ft. Ave Cap Rate Basis Points
No Anchor   24 $18.71 8.68%  
1 Anchor   10 $12.83 9.07% 0.39
2 Anchors   7 $8.18 9.22% 0.53
3 or more Anchors   12 $7.09 10.84% 2.15
Total/Means   53 $13.58 9.31% 1.02


The universe of data was tested using three options. The independent variables in all models included date of sale, the GLA sold, effective age, anchor department stores, and a subjective variable for location. The dependent variable is the OAR. The models examined three variations on the identification of the anchor department store. All variables were determined to be significant, but to varying degrees.
Model 1: The model employed an independent step variable identifying one, two or three or more anchor department stores. This variable produced a T-score of 2.6, and was significant to 1%. Overall, the model had an adjusted R2 of 56.2, a standard error of .013427, and a 1% significance (a 99% confidence level). While this model is significant, it presupposes that all additional department stores have an equal weighing in the formula. This may or may not be true. The theory section suggests that it may be a correct hypothesis.
Model 2: This model employed an independent variable for one department store, another for a second store, and a third for three or more stores. In the aggregate, the T-factor was equal to that in Model 1, but individually only the three or more anchor department store variable was independently significant. The R2 equaled 54.0, standard error was .01375, and the significance was at the 1% level. The coefficients for the department store start out lower and increase as the number of department stores increase, thus suggesting that additional department stores influence is not uniform.
Model 3: This model bifurcated the department stores into two independent variables, namely 1 and 2 department stores, and 3 or more department stores. The anchor store T-factor was again equal to that of the earlier models, with 3+ anchors being significant. The R2 equaled 57.1, standard error was .013286, and the significance was at the 1% level. Statistically, Model 3 is the preferred model.
The anchor department store issue is significant insofar as our hypothesis suggested that the advent of one anchor had the greatest percentage effect on the NOI, reducing the original NOI from $25.00 per square foot of GLA to $18.00 per square foot of GLA; a change of 28%. The change is not corroborated by the department store coefficient alone. Other variables must contribute to this condition providing the department store OAR increment theory is to be proven true.
The other independent variables were also tested. Individually: date of sale, sold GLA, effective age, and location all tested significant at <5%. Overall, all the models were significant at the 1% level.   The dependent variable is the OAR.  The means and assumptions of the independent variables are as follows: 
Date:                          February 2004 (tested at February 2004)
Sold GLA; Mean:    492,270 sq. ft.
Tested at:            300,000 sq. ft. no anchor
                          450,000 sq. ft. one anchor
                          600,000 sq. ft. two anchors
                          750,000 sq. ft. three anchors               
Effective Age, Mean: 20 years (tested at 10 and 20 years)
Location:  From a range of 1 being a weak location to 4 being a stronger location. In all likelihood, considering the ages of the universe, data equates to a Korpacz range of A-A+ (tested at 4).  
The model’s output, assuming the above variables and an effective age of 10 years, and three variables of department stores, resulted in the following OARs. 



  Model 1  Model 2   Model 3   Average

No anchor   

  6.94%   7.12%   7.05%   7.04%

1 anchor   

  7.71%   7.49%   7.48%   7.56%

2 anchors   

  8.48%   7.83%   7.91%   8.07%
3+ anchors   9.25%   9.65%   9.50%   9.47%

 The corresponding OAR for a 20 year old property is as follows: 

Category    Model 1    Model 2    Model 3    Average
No anchor    7.42%    7.44%    7.47%   7.44%
1 anchor    8.19%    7.81%    7.90%   7.97%
2 anchors    8.95%    8.15%    8.33%   8.48%
3+ anchors    9.72%    9.97%    9.92%   9.87%

Then, comparing the theory with the mean outputs from the 10 and 20 year age analyses, the following graph demonstrates these relationships.   


                              Theory      10 years       20 years      
No anchor            7.50%           7.04%                 7.44%
1 anchor              8.10%   60    7.56%      52     7.97%      53
2 anchors            8.67%   57   8.07%       51     8.48%      51
3+ anchors         9.30%    63   9.47%    140     9.87%     139
       Means           8.39%             8.04%                 8.44%

From the empirical evidence presented herein, it is suggested that the theoretical impact of having an anchor department store in a mall sales is largely corroborated by market based evidence. Our theory discussion suggested a difference of approximately 60 basis points for malls with up to three department stores. The market based evidence suggested 50 basis points for adding one or two anchors each, and the increase for the third plus is 100 basis points. The overall conclusion suggests that anchor department stores add approximately 50 to 240 basis points to a mall’s sales OAR as opposed to sales without an anchor department store included in the sale.   

It is noted that traditional wisdom concludes that the existence of an anchor department store will bring stability to a mall’s financial statement, reduce the presumption of risk, and result in lower overall capitalization rates. This perception is evidenced in institutional studies measuring the difference between anchored malls and non-anchored malls, e.g., specialty malls, strip malls or power centers. All of the malls included in this study are anchored malls. The distinction is whether or not the anchor(s) are included in the sale.   

It is further noted that the studied analysis is time, age, size, and location sensitive. The coefficient for time is negative, demonstrating that OARs are decreasing at approximately .5% per annum over the study period. This decrease roughly coincides with the decline in rates measured by Korpacz for institutional grade properties over the same period. While all of the properties in the model could not be considered of institutional grade, the data conclusively suggests that the advent of an anchor department store, in all probability, should result in an increase in the transaction’s reported overall capitalization rate.   

The authors wish to acknowledge James B. Robinson, CMI, National Director of Property Taxes at Dillard’s Department Store for initiating this discussion, and Alvin Lankford, RPA, and the appraisal staff of Williamson Central Appraisal District, Georgetown, Texas.   

1Department store rents tend to be based on a base rent plus a decreasing rate of overage rent beyond a given breakpoint. Thus, as sales increase, rents increase, but rents increase at a decreasing rate. In other words, the higher the sales, the higher the rent, the lower the rent as a percentage of sales. The inverse is also true.




About Jim MacCrate

Real estate appraiser and valuation consultant for more than 30 years specializing in reviewing real estate appraisals, risk management and quality control.
This entry was posted in Appraisal Process, Capitalization Rates, Discount Rates, Real Estate Valuation Methodology and tagged , , , , , . Bookmark the permalink.


  1. G Gervasi says:

    That was one of the best analysis of malls and anchor store cap rates I have read in 10 years James. Very well written and explained too. Thank you for the information.
    Please continue to publish this kind of excellent work to keep us informed and educated in the Real Estate Valuation field.

Comments are closed.