This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Slide 1
Dr. Charles C. Carter and Dimitri M. Teddone, MAI Store
Location in Shopping Centers: 15 Years After
Slide 2
Frank Lloyd Wright: The Elevator & the Automobile at the
Turn of the Century With the advent of the elevator, downtown
shopping districts with its department stores dominated the retail
shopping experience near the end of the nineteenth century.
Electric streetcar lines then connected the downtown retail
district with residential areas like the spokes of a wheel.
Chicago: Marshall Fields; Detroit: Hudsons; New York: Macys;
Minneapolis: Daytons; Portland: Meier & Frank Frank Lloyd
Wright (1867-1959) knew what to expect when he said: The city of
the future will depend on the race between the car and the
elevator, and anyone who bets of the elevator is crazy.
Slide 3
Slide 4
Rise of the Shopping Center Automobile use expanded during the
1920s and 1930s & residential uses expanded rapidly into the
suburbs. Levitt-towns first started on Long Island, NY after World
War II & these became the model of residential uses. Suburban
subdivisions consisting of single-story, detached houses on single
lots became the norm spreading out from metropolitan areas
throughout the U.S. Two- or three-story row houses along boulevards
near the downtown were no longer the custom. Shopping centers came
to revolutionize retail shopping during the latter half of the 20
th century.
Slide 5
Slide 6
Slide 7
Slide 8
Fall of the Downtown Retail District Downtown retail districts
of the 1950s were eventually drained of retail uses as suburban
shopping centers took over the retail trade. Early suburban centers
were built by department stores, i.e., Hudson Co. built Northland
Center in Detroit (1954) & Dayton Co. built Southdale Center in
Minneapolis (1956). Office uses came to dominate the downtown after
the 1960s Suburban residences met the demand for more living space
after the time to cover distances (using cars) shortened. ULIs
Shopping Center Development Handbook, 2 nd ed. graphs the increase
of US shopping centers from 1950 to 1985.
Slide 9
Slide 10
Growth of Shopping Centers From 1950 to 1955 the graph shows a
slow increase in the number of centers; from 1955 to 1980 the rate
grows at 6 times as much. By 1985 about 50% of all retail sales
were made at shopping centers, not including automobile sales &
gasoline. The rate of retail sales in shopping centers since 1985
has declined: 1974: 25%, 1985: 50%, 1996: 52% In 1998 87% of all US
shopping centers were less than 200,000 SF. GLA (smaller than
regional or super-regional centers). In the late 1970s & early
1980s regional shopping returned
Slide 11
Slide 12
Slide 13
Shopping Centers Slow to the downtown, e.g., Faneuil Hall,
Boston, Harborplace, Baltimore, & South Street Seaport, New
York By the1990s shopping center growth had slowed considerably,
evidence that the market had become saturated Big box or category
killer retailers such as Lowes, Home Depot, Costco, Best Buy, &
K-Mart became popular during the 1980s; online retailing has
expanded since about 2000 By 2005 construction of new, enclosed
shopping centers had just about stopped. Today shopping centers are
very much with us, and, unlike the downtown retail district, are
not likely to disappear soon.
Slide 14
What Did Shopping Centers have that the Central Business
District Didnt? Studies of microeconomics of shopping centers
started in earnest about 1989 and continues today. Research
eventually showed that the structure of rent, space, leases, and
store location under a unified plan & management results in
positive demand externalities that a CBD cannot reproduce. What are
positive consumer demand externalities? They get customers to spend
their money by providing a place & circumstances where
consumers want to shop. Economists call this internalizing positive
externalities; the ULI calls this process synergistic & we can
leave it at that.
Slide 15
Original Data: 1992-93 Shopping Centers, 8 Regional &
Super-Regional Original data was gathered from 9 regional
(300K-900K GLA) and super-regional (500K-1.5M GLA) shopping centers
scattered throughout the U.S. (avg. 828.6K GLA) 6 were
single-level, 2 double-level, & 1 three-level (3-level mall not
used) anchor stores ranged from 2 to 6 per center (avg. 3.56)
Parking ratios ranged from 5.35 to 5.62 stalls per 1,000 sq ft of
shopping area (standard = 5.5 stalls per 1,000 sq ft) Years
completed ranged from 1952 to 1990, the center built in 1952 had
been completely renovated in 1987 (avg. 16 yrs.) Vacancies ranged
from 4.4% to 0.4%, avg. 3.1%
Slide 16
Original Data: 1992-93 Shopping Centers, 8 Regional &
Super-Regional Numbers of non-anchors (including vacancies) ranged
from 117 to 158 (avg. 129) Of the 1,010 leases 689 were deemed
usable (69%) Leases were unusable: 1) lease either starts or ends
in the year sales were reported (insufficient info. on
renewed/rolled over tenants) 2) some tenants did not report sales
(i.e., service tenants w/out overage rent) 3) cinemas & out-pad
tenants (the former often operate on a different schedule & the
latter are outside the mall)
Slide 17
Slide 18
Slide 19
Findings in First Study 1) Highest customer traffic takes place
at the malls center (usually at or near the food court), &
customer traffic tapers off with distance from the center. 2) Store
size increases, & rent per SF decreases, w/distance from the
malls center (rents per SF @ +4.0% /100 ft) 3) Stores of the same
type that promote comparison shopping (e.g., womens apparel &
mens apparel) will generally be dispersed as opposed to clustered
4) Rents decline at different rates with distance from the mall
center, smaller store types paying higher rents/SF at the center
whose rents decline faster than larger store types.
Slide 20
Slide 21
Slide 22
Slide 23
Slide 24
Conclusion: There is an Optimal Tenant Mix & Optimal Tenant
Locations Stores locate according to what is termed a bid-rent
process in microeconomics, so that store types position themselves
to receive maximum sales (and pay maximum rents) Store location
& size are two of the mall characteristics that, together, go
to maximize rents (lease structure does too) Assuming leases are
properly structured, the retail market is just-saturated, and zero
vacancies, store sizes and locations within the mall are the
remaining variables to obtain optimal tenant mix. These variables
are best understood as the result of trial & error, Adam Smiths
invisible hand.
Slide 25
Slide 26
Follow-Up Data: 2006 Shopping Centers, 8 Regional &
Super-Regional Follow-up data was gathered from 8 regional
(300K-900K GLA) & super-regional (500K-1.5M GLA) shopping
centers located along the U.S. Atlantic Coast (avg. 481.5K GLA) 6
were single level & 2 were double level centers anchor stores
ranged from 2 to 4 per center (avg. 3.25) Years completed ranged
from 1961 to 1992, with 2 centers completely renovated (ones built
in 1961 & 1985 in 1993 & 1995, respectively) (avg. 25.6
yrs.) Leases were unusable: 1) 3 shopping centers had too many
vacancies (46%, 30%, & 20%) for results to be creditable
(previous research shows this)
Slide 27
Follow-up Data: 2006 Shopping Centers, 8 Regional &
Super-Regional 2) For 3 centers site plans were not readily
available (from which to measure distances from stores) For the
remaining 2, vacancies were 8.6% and 10%, parking ratios were 5.5
& 6 spaces per 1000 GLA, & the number of usable non-anchor
stores were 50 (w/33K GLA vacant) and 41 (w/47K GLA vacant) Of the
91 leases 54 were deemed usable (59%) Comparisons: 481.5K avg. GLA
(2006) vs. 828.6K avg. GLA (1992); 3.25 anchors avg. (2006) vs.
3.56 anchors avg. (1992); 25.6 avg. age (2006) vs. 16 avg. age
(1992); 17.2% avg. vacancies (2006) vs. 3.1% avg. vacancies (1992);
number of usable leases 54 (2006) vs. 689 (1992)
Slide 28
Slide 29
Slide 30
How to Get More Observations: Bootstrapping A bootstrapping
technique was used to provide more observations (lease data to use
in statistical analysis). By this nonparametric method samples from
the observations at hand are made as if these samples were from a
population. In bootstrapping, sampling with replacement takes place
using the data on hand (the 54 store leases) to create a larger
number of observations from which to run statistical analysis. From
this method 1,200 non-anchor observations were produced, i.e., 60
random samples of 20 leases from the 54 store leases = 1,200 lease
observations
Slide 31
VariablesMeanStd. Dev.Min.Max. Square Feet
(SF)3084276159015,709 Rent per SF ($/SF)
$43.64/SF$24.92/SF$4.09/SF$118.30/SF Feet from Mall Center
267.17130.7830540 VariableStore TypeObservatio n Frequency MeanStd.
Dev.Min.Max. SFWomens Apparel 11.6%5,2151,665.823,0157,773
SFJewelry7.5%1464.5205.401,2051,673 SFFood
Court13.3%784254.645901,226 TRNT /SFWomens Apparel
$18.17$13.91$8.4441.36 TRNT / SFJewelry$57.29$17.13$40.46$76.13
TRNT / SFFood Court$82.07$21.55$60.10$118.30 Store Characteristics
for 2006 Follow-Up Data
Slide 32
Findings in Follow-Up Study 1) Store size increases, & rent
per SF decreases, w/distance from the malls center (rents @ +5.3%
/100 ft, square footage +8.26% / 100 ft) 2) Generally, store types
still line up with distance from the malls center (higher rents
& smaller stores w/distance) the same as they did in the
earlier study. 3) Store types have changed significantly from 1992
to 2006, e.g., fewer mens shoe stores & more electronics
stores, which makes testing for location effects of comparison
store shopping difficult.
Slide 33
Comparisons:1992 & 2006 Data MeanStd.
Dev.Differencew/Inflation 1992 SF2,417.06 ft 1992 SF
Jewelry1,239.57 ft679.90 ft 1992 SF Fast Fd874.69 ft859.13 ft 1992
SF Wom A3,855.74 ft2,305.82 ft 2006 SF3,084 ft2,761 ft+ 27% 2006 SF
Jewelry1,464.5 ft205.64 ft+ 18% 2006 SF Fast Fd784 ft254.64 ft -
10% 2006 SF Wom A5,215 ft1,665.8 ft + 35% 1992 $Sales$363.72/SF
1992 $Jewelry$656.37/SF$367.37/SF 1992 $Fast Fd$503.18/SF$268.42/SF
1992 $Wom Ap$252.01/SF$130.19/SF 2006 $Sales
Comparisons: 1992 & 2006 Data MeanStd DevDifference % 1992
Fast Fd ft70.5 ft115.4 ft 1992 Wom Ap ft312.7 ft519.8 ft 2006 Ft
Cent267.16 ft130.78- 9% 2006 Jewelry ft 2006 Fast Fd ft 2006 Wom Ap
ft
Slide 36
Observations on Store Types, 1992 vs 2006 Data 1992 Tenants% of
total2006 Tenants% of total Fast Food11%Fast Food12%by number of
stores not GLA Jewelry6.2%Jewelry6.9% Specialty Food6.3%Specialty
Food6.9% Mens Apparel6.3% Womens App18.5%Womens App14% Family
Apparel7%Men & Wom Ap8.6% Womens Shoes5%Men Wom Shoes8.6% Mens
Shoes6.7% Home Furn6.2%Home Furnish3.4% Cards & Gifts6.5%Cards
& Gifts5.2% Leisure & Enter22%Leisure & Ent8.6% Books2%
Services14% Electronics10%
Slide 37
Did Rent per Square Feet Keep Up with Inflation? Definitely
not, but it would be premature to say exactly why and how shopping
center rents are falling historically. The only obvious conclusion
is that regional & super- regional shopping centers are not
doing as well as they had been to internalize positive demand
externalities compared with other retailing. Still I dont think any
other form of retailing surpasses the shopping center experience
(if only we could impose sales taxes on online shopping). Retailing
in urban areas is best done using regional & super- regional
models such as Harborplace, Baltimore.
Slide 38
Future Research Similar research should be done using recent
data from like regional & super-regional shopping centers. The
best follow-up data would be for those 8 centers used in the 1 st
study. The 1992 data was from the best shopping centers then owned
by a state pension fund. Choice of malls in 1992 by regional
managers was meant to impress the pension fund managers. The
follow-up data in 2006 was most probably from a cross- section of
centers that the owners or lenders thought might be having
financial difficulties.
Slide 39
My Address & Phone Number Dr. Charles C. Carter 9034 S.W. 7
th Street Boca Raton, FL 33433 954-708-3654
[email protected]