Transcript Charakteristika und Determinanten von Performance und

ERES Doctoral Session:
Capitalization rates as risk indicator for (non-)efficient properties?
Elaine Wilke
Real Estate Management Institute
EBS Universität für Wirtschaft und Recht
Söhnleinstraße 8d
65201 Wiesbaden
Agenda:
Capitalization rates as risk indicators for (non-) efficient properties?
1.0
Introduction / Research objective
2.0
Data and Research Method
3 .0
Results
4.0
Conclusion
Elaine Wilke
REMI, 2010-06-23
2
1.0 Introduction
• Risk of (non-)efficient properties
• so far only research and publications from the investors’ perspective
• But: most sustainability aspects influence the operating and occupancy costs of the occupier;
these costs are not taken into consideration when calculating the NOI
• how are the results going to change if also the occupiers’ goals are considered?
• investors: higher net rents and/or higher returns
• occupiers: reduced operation and occupancy costs
• both perspectives interact in the valuation process:
 sustainable (efficient) properties realize higher Capital Values than non-sustainable (efficient) properties,
as sustainability (efficiency) reduces the property specific risk
• Hypothesis:
The capitalization rate as all risks yield interacts as indicator for the risk of (non-) efficient properties
 higher risk premiums for non-efficient properties!
Elaine Wilke
REMI, 2010-06-23
3
Capitalization rates as risk indicators for (non-) efficient properties?
1.0
Introduction / Research objective
2.0
Data and Research Method
3 .0
Results
4.0
Conclusion
Elaine Wilke
REMI, 2010-06-23
4
2.1 Data
 Data from Investors and Occupiers:
 IPD Investment Property Databank UK
 IPD Occupiers Databank UK
 Sample size n = 47 objects (in both databanks for 2007 and 2008)
Elaine Wilke
REMI, 2010-06-23
5
2.2 Research Method
 Partial Least Squares (PLS) – method
(according to Wold)
 causal relationships between
(in)dependant (latent) variables
 also available for smaller sample sizes
 any measurement levels
 separate calculation for 2007 and 2008
 Illustration of the cause-effect relations
(directions and sings) between the hypothesis
and the latent variables.
Elaine Wilke
REMI, 2010-06-23
6
2.3 Base Model
Block 2
Model latent exogenous variable
Reflective Structure
Block 3
Model latent exogenous variable
Reflective Structure
Block 1
Block 4
Model latent exogenous
variable
Formative Structure
Model latent
endogenous variable
Reflective Structure
Elaine Wilke
REMI, 2010-06-23
7
2.4 Indicators
Indicator
Short
Lease contract
Calculation
Individual conditions of the lease contract
Net rent per sqm (NR)
NR
Net rent per sqm as relative difference to IPD mean at date of last rent review
Remaining years of lease contract
RL
Remaining years of individual lease contract
Property quality
Age
Condition
Overall quality of the property
A
Age of building at date of valuation
C
1 = very good condition
2 = good condition with minor improvements
3 = bad condition with major improvements
Property efficiency
Location
Total operating costs per sqm
Overall property efficiency as relative difference to the relevant IPD mean
L
OC
Rentable area
SQM
Vacancy rate
V
Property specific risk
Capitalization rate
1 = London
2 = big cities (biggest 8 cities in the UK excl. London)
3 = small cities
Total operating costs per sqm for the property as relative difference to the relevant
IPD mean
Total rentable area of the property
Economic vacancy rate (in % of income)
The property specific risk
CR
relative difference between the "risk-free rate" (10 years UK gilts) and the
calculated capitalization rate of the valuer for the individual property
Elaine Wilke
REMI, 2010-06-23
8
2.4 Indicator - Total Operating Costs per sqm
IPD International Total Occupancy Cost Code (ITOCC):
Elements of the IPD Total Operating Costs per sqm:
 consolidated service charge
 insurance
 internal repair and maintenance
 M&A repair and maintenance
 external/structural repair and maintenance
 minor improvements
 internal moves
 reinstatement
 security
 cleaning
 waste disposal
 internal plants and flowers
 ground maintenance
 water, sewerage
 energy
Source: IPD ITOCC 4th edition
Elaine Wilke
REMI, 2010-06-23
9
Capitalization rates as risk indicators for (non-) efficient properties?
1.0
Introduction / Research objective
2.0
Data and Research Method
3 .0
Results
4.0
Conclusion
Elaine Wilke
10
REMI, 2010-06-23
3.1 Base Model (2007)
R²
Elaine Wilke
11
REMI, 2010-06-23
3.1 Base Model (2007)
Lease contract
Property efficiency
Property quality
Property specific risk
AVE
0,514
0,000
0,743
1,000
Composite
Reliability
0,363
0,000
0,851
1,000
R Square
0,247
0,000
0,000
0,377
Cronbachs
Alpha
Communality
-0,152
0,514
0,000
0,231
0,674
0,743
1,000
1,000
Model quality criteria:
 AVE > 0.5
 Composite Reliability > 0.7
 Cronbach‘s Alpha > 0.7
 Modification of the model!
Elaine Wilke
12
REMI, 2010-06-23
3.2 Model – modified (2007)
R²
Property efficiency
Property quality
Property specific risk
Composite
Cronbachs
AVE
Reliability
R Square
Alpha
Communality
0,00
0,00
0,08
0,00
1,00
0,75
0,86
0,00
0,67
0,75
1,00
1,00
0,28
1,00
1,00
 AVE > 0.70
 Composite Reliability > 0.7
 Composite Reliability > Cronbach‘s Alpha
 Cronbach‘s Alpha ~ 0.7
Elaine Wilke
13
REMI, 2010-06-23
3.2 Model – modified (2007)
Discriminant Validity
A
C
CR
V
Property
efficiency
0,00
0,00
0,00
1,00
Property
Property
quality
specific risk
0,90
0,00
0,83
0,00
0,00
1,00
0,00
0,00
A
C
CR
V
Property
efficiency
-0,16
-0,33
-0,06
1,00
Property
Property
quality
specific risk
0,90
-0,52
0,83
-0,30
-0,49
1,00
-0,28
-0,06
Property efficiency
Property quality
Property specific risk
Property
efficiency
0,000
-0,275
-0,058
Outer Loadings > 0.7
 support validation of reflective model
Cross Loadings < Outer Loadings
 no multicollinearity
Property
Property
quality
specific risk
0,000
0,000
0,867
0,000
-0,485
1,000
degree to which two measures designed to
measure similar or conceptually related
constructs:
√AVE > Latent Variable Correlation
 no relation
Elaine Wilke
14
REMI, 2010-06-23
3.2 Model – modified (2007) - Bootstrapping
 Estimating the distribution of the statistic by using the bootstrapping method
 The calculation is based on 300 cases and 500 samples
Original
Sample (O)
A <- Property quality
0,899
C <- Property quality
0,833
CR <- Property specific risk
1,000
V -> Property efficiency
1,000
Sample
Mean (M)
0,900
0,833
1,000
1,000
Standard
Deviation
0,012
0,024
0,000
0,000
Standard
T Statistics
Error
(|O/STERR|)
0,012
74,650
0,024
35,420
0,000
0,000
0,000
0,000
 standard errors with values < 0.04 suggest a low level of uncertainty
 T Statistics indicate a good fit of the model explaining the degree of variability of the dependent
variable
Elaine Wilke
15
REMI, 2010-06-23
3.2 Model – modified (2007)
Property efficiency
Property quality
T-Statistics
5,542
13,454
f²
q²
0,0883
0,0879
R² = 27,53%
GoF = 0,401
 bigger role in explaining than in predicting as q² <f²
 The effect size (f²) with a value of 0.088 has to be interpreted according to Cohen as a
low to medium effect with f² <0.15.
 Q² with 0.267 implies predictive relevance
 The calculated GoF (.401) is higher than the marginal value of > .275 indicating that the
model strongly fits the set of observations.
Elaine Wilke
16
REMI, 2010-06-23
3.3 Model – modified (2008)
R²
Property efficiency
Property quality
Property specific risk
Composite
AVE
Reliability
0,00
0,00
0,69
0,81
1,00
1,00
 AVE ~ 0.70
 Composite Reliability > 0.7
Cronbachs
R Square
Alpha
Communality
0,05
0,00
1,00
0,00
0,67
0,69
0,11
1,00
1,00
 Composite Reliability > Cronbach‘s Alpha
 Cronbach‘s Alpha ~ 0.7
Elaine Wilke
17
REMI, 2010-06-23
Capitalization rates as risk indicators for (non-) efficient properties?
1.0
Introduction / Research objective
2.0
Data and Research Method
3 .0
Results
4.0
Conclusion
Elaine Wilke
18
REMI, 2010-06-23
4.0 Conclusion
Type of Hypothesis
Indicators -> latent variable
Supported
H1: Property specific risk-> CR
H10: A->Property quality
H11: C->Property quality
H15: V->Property efficiency
H7: Property
efficiency
Exogenous -> exogenous variable
Exogenous -> endogenous
variable
Not supported
H8: Lease contract-> RL
H9: Lease contract-> NR
H12: L->Property efficiency
H13: OC->Property efficiency
H14: SQM->Property efficiency
quality->Property H5:
Property
quality->Lease
contract
H6: Property efficiency-> Lease
contract
H3: Property quality-> Property H2: Lease contract-> Property
specific risk
specific risk
H4: Property efficiency-> Property
specific risk
Elaine Wilke
19
REMI, 2010-06-23
4.0 Base Model
Block 2
Model latent exogenous variable
Reflective Structure
Block 3
Model latent exogenous variable
Reflective Structure
Block 1
Block 4
Model latent exogenous
variable
Formative Structure
Model latent
endogenous variable
Reflective Structure
Elaine Wilke
20
REMI, 2010-06-23
4.0 Conclusion
 so far only minor consideration of property characteristics in the derivation of the cap rate
 changes in the (economic) environment dominate the choice of the risk premium
 no integration of use efficiency
 occupiers’ perspectives are ignored
 future consideration of (non-)efficiency ?
 review of the interrelations with bigger sample sizes
Elaine Wilke
21
REMI, 2010-06-23
Capitalization rates as risk indicators for (non-) efficient properties?
Thank you!
Elaine Wilke
22
REMI, 2010-06-23