On the pulse of the property world Italian lease events: changes in tenants’ behaviour in the first year of the crisis Lease expiries, breaks and void periods in 2008 Luigi Pischedda - Country Manager Italy, IPD Duccio Martelli - University of Rome “Tor Vergata” Claudio Giannotti - University LUM Casamassima (Bari)
On the pulse of the property world Italian lease events: changes in tenants’ behaviour in the first year of the crisis Lease expiries, breaks and void.
Dec 17, 2015
Welcome message from author
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
On the pulse of the property world
Italian lease events: changes in tenants’ behaviour in the first year of the crisis
Lease expiries, breaks and void periods in 2008
Luigi Pischedda - Country Manager Italy, IPDDuccio Martelli - University of Rome “Tor Vergata”Claudio Giannotti - University LUM Casamassima (Bari)
www.ipd.com
On the pulse of the property world
Agenda
Italian lease events 2008
Databank features
Executive summary
Expiry analysis
Break Analysis
Void periods
Model
Introduction
Hypothesis
The regression function
Preliminary results
Strengths and limits of the model
Next steps
www.ipd.com
On the pulse of the property world
Sources: IPD
Databank features
The IPD Italian Tenancy Universe 2008
As at the end of 2008, the databank included circa 8,000 leases, for a total rental value of over Euro 2bn.
The total market value of the 1,762 properties was circa Euro 21bn.
32%
6%
25%
2%
35%
17%6% 1% 3%
73%
retail office industrial residential other
retail office industrial residential other
Sector weights by number Sector weights by rental value
www.ipd.com
On the pulse of the property world
Sources: IPD
Executive summary
1. Lease expiries
Weighted by number
Weighted by rental value
Renewals (%)
Re-let (%)
Vacant at year end (%)
52
27
22
56
4
40
What was the result of leases expiring in 2008?
www.ipd.com
On the pulse of the property world
Sources: IPD
Executive summary
2. Break clauses
Weighted by number
Weighted by rental value
Break not exercised (%)
Exercised and re-let (%)
Exercised and vacant at year end (%)
50
43
7
39
46
15
What was the result of leases containing a break in 2008?
www.ipd.com
On the pulse of the property world
Sources: IPD
Lease expiry analysis
Sectors, % weighted by number
0%
20%
40%
60%
80%
100%
Retail
Retail (
2007
)
Office
Office
(200
7)
Indu
stria
l
Indu
stria
l (20
07)
Other
Other
(200
7)
All pro
perty
All pro
perty
(200
7)
Renewed Re-let Vacant
www.ipd.com
On the pulse of the property world
Sources: IPD
Lease expiry analysis
Sectors, % weighted by rental value
0%
20%
40%
60%
80%
100%
Retail
Retail (
2007
)
Office
Office
(200
7)
Indu
stria
l
Indu
stria
l (20
07)
Other
Other
(200
7)
All pro
perty
All pro
perty
(200
7)
Renewed Re-let Vacant
www.ipd.com
On the pulse of the property world
Sources: IPD
Break analysis
Sectors, % weighted by number
0%
20%
40%
60%
80%
100%
Retail
Retail (
2007
)
Office
Office
(200
7)
Indu
stria
l
Indu
stria
l (20
07)
Other
Other
(200
7)
All pro
perty
All pro
perty
(200
7)
Renewed Re-let Vacant
www.ipd.com
On the pulse of the property world
Sources: IPD
Break analysis
Sectors, % weighted by rental value
0%
20%
40%
60%
80%
100%
Retail
Retail (
2007
)
Office
Office
(200
7)
Indu
stria
l
Indu
stria
l (20
07)
Other
Other
(200
7)
All pro
perty
All pro
perty
(200
7)
Renewed Re-let Vacant
www.ipd.com
On the pulse of the property world
Sources: IPD
Void period analysis
Segments, % weighted by number, 2008
Total 0-6 6-12 over 12
All Retail 100 37 63 0
Standard Units 100 15 85 0
Shopping Centres 100 53 47 0
All Office 100 26 59 14
Offices - Milano 100 10 60 29
Offices - Milano Hinterland 100 44 56 0
Offices - Roma Central 100 3 10 87
Offices - Roma Other 100 53 47 1
Offices - North West 100 1 82 17
Offices - North East 100 1 99 0
Offices - Centre and South 100 80 0 20
All Industrial 100 0 100 0
All Hospitality 100 98 2 0
All Other 100 29 71 0
Residential 100 0 100 0
Other 100 74 26 0
All Property 100 32 58 10
In percentage terms, by number of units, number of months void
www.ipd.com
On the pulse of the property world
Sources: IPD
Total 0-6 6-12 over 12
All Retail 100 83 17 0
Standard Units 100 85 15 0
Shopping Centres 100 78 23 0
All Office 100 85 15 0
Offices - Milano 100 86 14 0
Offices - Milano Hinterland 100 72 28 0
Offices - Roma Central 100 100 0 0
Offices - Roma Other 100 96 4 0
Offices - North West 100 83 17 0
Offices - North East 100 65 35 0
Offices - Centre and South 100 56 44 0
All Industrial 100 100 0 0
All Hospitality 100 0 100 0
All Other 100 97 2 1
Residential 100 85 8 8
Other 100 99 1 0
All Property 100 87 13 0
In percentage terms, by number of units, number of months void
Void period analysis
Segments, % weighted by number, 2007
www.ipd.com
On the pulse of the property world
• The global economic and financial crisis which hit the Italian real estate market in 2008 is reflected in the lease events occurred in the same year.
• The comparison of 2008 figures against 2007 results show a deteriorated lease market in terms of lease behaviour following the expiry of the contract or a break.
• The void period figures of vacant units in 2008 has deteriorated: while in 2007 most of the vacant units at year end (87% of the total number) was in the 0 to 6 months interval, in 2008 nearly 60% of vacant units had a void period within the range 6-12 months.
Preliminary considerations
www.ipd.com
On the pulse of the property world
Agenda
Italian lease events 2008
Databank features
Executive summary
Expiry analysis
Break Analysis
Void periods
Model
Introduction
Hypothesis
The regression function
Preliminary results
Strengths and limits of the model
Next steps
www.ipd.com
On the pulse of the property world
• Tenant risk has reached higher importance especially in recent years, due to markets falls and consequently the increase in the number of break option exercises
• The aim of this model is to investigate the critical factors affecting the probability of leaving a building vacant by the tenant before the natural expiration date of the contract in the Italian real estate market
• The model which is going to be presented is a preliminary version: all comments and suggestions are very appreciated
• Nowadays only very few models try to estimate the probability of leaving a building vacant; probably this is mainly due to the absence of available databases
• The study is based on IPD Italian databank and thus the factors taken under consideration are mainly economic and contractual
The forecasting model
Introduction
www.ipd.com
On the pulse of the property world
• We adopted a methodology which is mainly used in medical researches to identify factors affecting diseases: the logistic (or logit) regression
• Since tenant risk is a kind of risk like any other, it is possible to apply this model to real estate market
• The logistic regression model is a non parametric model where the output variable is binary (assume values 0 or 1) depending if that variable has the characteristic investigated or not
• Besides, the logistic model combines the easiness of the construction to the immediateness of results interpretations
The forecasting model
Hypothesis
www.ipd.com
On the pulse of the property world
• The regression has been applied on a sample of about 250 contracts which could be closed prior their natural expiration date during the years 2007 and 2008
• Like for all regressions implying qualitative variables, the model identify a comparable building, which in this case has been defined as an office located in Rome, with a break option dated 2007
• Taking into consideration collinearity problems (an independent variable is mostly explained by another) and estimation problems (absence of the event analyzed in some considered variables), the function used in the model is the following:
The forecasting model
The regression function
www.ipd.com
On the pulse of the property world
The forecasting model
The regression function
Variable name Description
yesbreakDummy variable with value 1 if the building has been left by the tenant during the year; 0 if the tenant has not exercised the break option
sectorDummy variable which can assume the following values: Hospitality & Other, Industrial, Office, Residential, Retail
sqm100 The size of the building in square meters (actually sqm/00)
efferv100Indicates the difference between the effective rent paid by the tenant and the estimated rented value that could reasonably be expected to be obtained on a new letting (on annual basis for 100 sqm)
areaDummy variable which can assume the following values: Center, North-East, North-West, South
city Indicates whether the building is located in a big city or not.
renovDummy variable which assumes value 1 if the building has been recently renovated or 0 otherwise.
yearIndicates the year during which the break option can be exercised. It can assume the following values: 2007, 2008
yesbreak = β0 + β1*sector + β2*sqm100 + β3*efferv100 + β4*area +
+ β5*city + β6*renov + β6*year
www.ipd.com
On the pulse of the property world
Preliminary results are shown in the following table:
The forecasting model
Preliminary results
How to read the table
Exponentials of coefficients represent the number of chances (more or less) compared to the comparable building
The most critical variables seems to be sector of activity, year of exercise and location
Although the pseudo-R2 seems to be low, we have to remember that the model considers just economic and contractual variables
www.ipd.com
On the pulse of the property world
• Strengths and limits of the model can be summarized in the following points:
It identifies the key factors affecting the tenant risk, but probably it overestimates the intensity of the relations
It excludes some variable, not because they are not important, but just because no break options were exercised during the period 2007-2008 (i.e. industrial sector, specific regions, …)
It is based on a two-year sample
• Although the model is not limits exempted, it can represent a good starting point for future developments
The forecasting model
Strengths and limits of the model
www.ipd.com
On the pulse of the property world
• In order to solve these issues, next steps regard the possibility:…
…to extent the time series used for the analysis
…to transform the results from “numbers of chances” to “probabilities” of break exercises for every building, no matters if this one is included or not in the sample
…to extend the analysis to other real estate markets, in order to identify differences in risky factor affecting the chances/probabilities of leaving a building vacant
The forecasting model
Next steps
www.ipd.com
On the pulse of the property world
The forecasting model
Operative implications
• In particular, once those issues are solved (or limited), we think that the model can be used especially by landlord for two main applications:
I. Identifying risky situations before they really happen (and thus renegotiate the contract before the break option exercise)
II. Pricing the contract correctly, linking the rent to the real riskiness of the investment (taking into consideration both exogenous and endogenous variables, but also risky factors related to the tenant)
Related Documents
