Readings
• Required: Green and Malpezzi, A Primer on U.S. Housing Markets and Housing Policy, pp. 31‐60 (on measuring prices), and pp. 146‐
156 (on measuring regulation).
• Background: Davis, Lehnert and Martin, The Rent‐Price Ratio for the Aggregate Stock of Owner‐Occupied Housing. Review of Income and Wealth, 2008.
• Background: Green, Malpezzi and Mayo, Metropolitan‐specific estimates of the price elasticity of supply of housing, and their sources. American Economic Review, 2005.
• Background: Malpezzi and Maclennan, The Long‐run Price Elasticity of Supply of New Residential Construction in the United States and the United Kingdom. Journal of Housing Economics, 2001.
• Background: David Shulman, The Outlook for Commercial Real Estate: Asset Prices Ahead of Fundamentals. UCLA Anderson Forecast, June 2011.
Quadrant III: The Supply of New Construction
144
145
Demand Shocks with Inelastic Supply
Rent (or
Price) per
Unit of
Space
Demand Shocks with Elastic Supply
Rent (or
Price) per
Unit of
Space
Supply
Supply
P0
P0
Demand Increases
Demand Increases
Demand
Falls
Q0
Demand
(original)
Demand
Falls
Number of Housing Units
Q0
146
Demand
(original)
Number of Housing Units
147
1
Elasticity
S
PS
ε

Q P
/
/ SP
Q Δ
Δ
• Economic jargon for "responsiveness." The proportionate change in output given a proportionate change in price. • Mathematical representation of price elasticity of supply:
• There are many elasticities, e.g. supply vs. demand elasticities; w.r.t. price, income, population…
• Be familiar with this and with the corresponding graphical representation.
• Be familiar with how elasticity varies with, inter alia, time, and the extent (size) of the market.
148
149
Calibrating Quadrant III: empirical estimates of the supply determinants of commercial real estate are scarce!
• There are now a few dozen good studies of (e.g.) the price elasticity of supply of housing.
– See Green, Malpezzi and Mayo (2005), Malpezzi and Maclennan (2001), and references therein.
• It’s hard to think of a single good estimate of the price elasticity of supply of, say, office space.
– How steep is the supply curve?
• We’ll look at housing first; then nonresidential supply.
150
151
2
What’s the “price” of housing, anyway?
Simple medians and averages
• Every house is more or less unique:
– Size
– Quality
– Services
– Location, location, location
• If I pay $1,000/ month rent and you pay $1,200, who’s paying a higher price?
• If our units are nearly identical, you do.
• If your apartment is 1500 square feet and mine is 500, maybe I do.
• If your apartment is in midtown Manhattan and mine is in Madison, maybe I do.
• How do we handle many characteristics at once?
• These have one great virtue: simplicity.
• They also tell us something about levels as well as changes.
• No/few controls for (e.g.) size, quality differences.
– These matter more for longer time spans.
– These matter more when the market turns weird.
– These matter more when locations have very different housing stocks.
• Examples: National Association of Realtors median sales price of existing homes.
Hedonic price indexes: regression analysis and valuation
• Hedonic models adjust for the quality of housing services from a given unit.
• Needed: data on properties that have sold, and their characteristics.
• Hedonic price indexes can be constructed from place to place, and/or over time; can estimate price levels
as well as price changes.
Regression Analysis and Valuation
ln(Sales Price)  Xβ  Nα  Lλ  Cγ  ε
^
^
^
^
^
 V  e X β N α L λ  C γ
• Where X are structural characteristics, N are neighborhood characteristics, L is location, and C are contract conditions.
• These regression models are called “hedonic price indexes” by economists.
• Example: See Primer, Table 2.5, metro data for Washington, D.C.
3
Repeat sales price indexes
User cost
• Repeat‐sales indexes are becoming popular for estimating price changes, if appropriate data are available.
– Case‐Shiller price indexes
– OFHEO price indexes (from Fannie and Freddie data)
• Needed: data on sales prices of units that sell at least twice.
– These must be arms‐length transactions.
– These must be units that have not depreciated much (or an adjustment must be made).
– These must be units that have not been upgraded (or an adjustment must be made).
• Determine what a “user” of the house really pays (or would pay), net of financing, taxes, maintenance, inflation, and so on.
• User cost measures are most often time series but can be cross section.
Example of Repeat Sales Sample
Figure 2.14
Spliced Quarterly Real House Price Index
17 Properties that Sold Twice
FHFA Index 1975 to 1986; Case‐Shiller 1987‐2010
$1,000,000
240
Natural
Logarithm
of Sale Price
7
7
6
5
11
14
2
16
$100,000
9
9
12
4
13
11
16
6 17
28
8 1 17
14
15
200
180
12
415
10
1
5
220
3
13
160
10
3
140
120
100
80
Jul-89
Aug-90
Dates of Sale
Sep-91
Oct-92
75‐1
76‐1
77‐1
78‐1
79‐1
80‐1
81‐1
82‐1
83‐1
84‐1
85‐1
86‐1
87‐1
88‐1
89‐1
90‐1
91‐1
92‐1
93‐1
94‐1
95‐1
96‐1
97‐1
98‐1
99‐1
00‐1
01‐1
02‐1
03‐1
04‐1
05‐1
06‐1
07‐1
08‐1
09‐1
10‐1
$10,000
Mar-86
May-87
Jun-88
Source: Village of Shorewood Hills, WI.
Slope of line is rate of appreciation.
Spliced FHFA/CS Real House Price Index
Prediction, minus 2 SEs
Prediction from Log Trend, 1975‐1995
Prediction, plus 2 SEs
4
Davis Estimates of Gross Rent‐Price Ratio, U.S. Stock of Owner‐Occupied Housing
Housing problems simplified
0.065
0.060
0.055
Based on Case‐Shiller‐
Weiss Prices
0.050
0.045
0.040
Based on Federal Housing Finance Agency Prices
0.035
• “Every country has only two housing problems, in the end. And once you fix those two problems, everything else falls into place.”
• “The difficulty is, these are the two housing problems:
– Housing prices are rising too fast; and
– Housing prices are not rising fast enough.”
Larry Hannah, former World Bank Lead Economist, Urban Development.
0.030
160
Nobody buys a house in the United States…
What does it mean to look at a national price index, anyway?
You buy a house in a metropolitan area.
162
5
Inflation‐Adjusted FHFA House Price Indexes
Selected Midwest Markets
500
Inflation‐Adjusted FHFA Housing Price Indexes
Twelve Most Volatile Markets (thru Q1 2010)
500
450
450
400
400
Milwaukee
350
Nassau
Salinas
350
Boston
Average of 147 MSAs
300
Santa Cruz
Chicago
250
Madison
300
250
Napa, CA
New York
Minneapolis
200
150
St Louis
200
Cincinnati
150
San Luis Obispo
San Jose
Cleveland
100
Ann Arbor
50
San Francisco
100
Detroit
0
Peabody
Cambridge
50
Santa Barbara
Q1 1980
Q1 1981
Q1 1982
Q1 1983
Q1 1984
Q1 1985
Q1 1986
Q1 1987
Q1 1988
Q1 1989
Q1 1990
Q1 1991
Q1 1992
Q1 1993
Q1 1994
Q1 1995
Q1 1996
Q1 1997
Q1 1998
Q1 1999
Q1 2000
Q1 2001
Q1 2002
Q1 2003
Q1 2004
Q1 2005
Q1 2006
Q1 2007
Q1 2008
Q1 2009
Q1 2010
0
165
House Prices & Regulation
Inflation‐Adjusted OFHEO House Price Indexes
Twelve Least Volatile Markets (through Q1 2010)
450
Topeka
400
Fort Worth
350
Rockford
300
Wichita
Little Rock
250
Fort Wayne
200
Dallas
150
Memphis
Lincoln
100
Indianapolis
50
Beaumont
Tulsa
0
Median House Value, 1990 Census
(Thousands)
350
500
SF
300
SJS
250
HON
LA
NY
NWK BOS
SDI
200
HRT
150
SAC
PRV
CHI
100
DTN
GRY
50
ALN
BAL
ALB PHL
FTL
ATL
MIA
ORL
DAL MIN DEN PHX RCH ROC
SYR
MIL GNC
CLE CTE
BUF
TPA
SLK COL MEM
STL DET
NO GRR POO
IND CIN
KCM
HOU
AKR
TDO
BIR
TUL
SAT
PGH
MOB
OKC
YNG
0
10
15
20
25
30
MSA-Specific Regulatory Index
166
Source: Malpezzi, J. Hsg Research, 1996
Summarized in Primer, pp. 146-56
What’s 167
this?
6
What does it mean to measure “regulation?”
• Building permit caps, sewer moratoria, large lot zoning, minimum lot sizes, floor area ratio, setback requirements, height limits, subdivision codes, environmental impact reviews, transferable development rights, retention ponds, rent controls, land use controls, impact fees, building codes, minimum street widths, curb and gutter requirements, traffic mitigation…
• Malpezzi and Ball, Malpezzi (1996), Gyourko, Saez and Summers (2006): regulations are correlated.
– Implies that omitted regulations will be “picked up” by included regulations.
– Implies that we can’t take the partial derivative of an element of a regulatory index. • How do we interpret MSA indexes of regulation when many sub‐
MSA jurisdictions have regulatory authority?
Measuring Regulation by MSA
• Index based on answers to the following:
– Recent changes in approval time for single family housing development
– Time required for rezoning and permitting for a small residential subdivision
– Ditto, for a large subdivision
– Single family zoning, compared to demand
– Multifamily zoning, compared to demand
– Percent of zoning changes approved
– Index of adequate infrastructure (roads and sewers)
• Index values range from 7 (permissive) to 35 (most stringent)
• Other indexes constructed for state regulations, rent control.
169
Housing Supply Elasticities
and Development Regulation
• Studies by Richard Muth, Jim Follain: the long run price elasticity of supply of housing from new construction is high (10, 20 or more).
• Studies by Jim Poterba, Topel and Rosen: the price elasticity of supply is more like 2 or 3.
• Malpezzi and Maclennan: the U.S. elasticity is high; Poterba and Topel and Rosen don’t use enough data. But the long run is, well, long.
• Green, Malpezzi and Mayo, and Tsur Somerville and Chris Mayer: the supply elasticity varies a lot by MSA. Appears to be related to regulation.
Supply Elasticity, Lagged RHS Vars
Supply from new construction: divergent views
DAL
30
T PA
25
PHX
ATL
OKC
COL
20
CTE
15
DEN
10
T UL
SAT
POO
STL
5
M IN
DET
M IL
BUF
CHI
NO
0
HOU
IND
GRR KCM
SLK
HRT
CIN
AKR
BAL
BIR M EM
ROC
ORL SYR
PRV
SDI
FT L
T DO
PHL
LA
BOS
ALB
PGH
M IA
SJS
SF
HON
-5
10
15
20
MSA Regulatory Index
25
30
What’s this?
Source: Green, Malpezzi & Mayo
(Linear Fit)
171
7
Office Appreciation and Regulation
1989 to 1996
0.15
LA
Annual Appreciation
0.10
JKL
0.05
CHI
TPA
STL CIN
BAL
PHL
CLE RVR
NO
INDPOO
SEA
DAL
SACGNC
PHX
OKC
ATL
HOU
BIR
FTL
MIL
KCM
PGH
COL CTE
NSH
LSVORL
MIN
SLK
SAT
0.00
-0.05
HON
SF
SDI
DET
NY
NFK
WPB
BOS
NAU NWK
SJS
MIA
DEN
AUS
-0.10
14
16
18
20
22
24
26
28
30
32
MSA Regulatory Index
172
(Linear Fit)
Regulation raises real estate prices; is that a good thing for investors?
• What’s the other thing that accompanies high returns?
174
8
Std Dev of Real Avg House Price Change
(1979-96) and Regulation (1989)
18%
NHA
Exploratory Regression, Explaining Standard Deviation of Annual Agency Housing Price Changes, U.S. Metro Areas
ATC
Std Dev Real Agency House Price Change
16%
TRN
14%
12%
10%
8%
6%
4%
2%
14
SJS
Std Dev of
Real Changes in
Income Per Capita
Std Dev of
Annual Changes
in Employment
M-C-G Regulatory
Index
Intercept
FWA
LA
RDG
SF
SRS
APL
SIL
RVR
BEL
SMA
MOD
PRV
SAT
SDI
RAC SAC
HRT
SEA
DAV PEO
SBR
ALN ALB
LAN
NWK BOSNY
VALNAU PME
ANN SNS
RKF
STC
SCZ
AUS
MIL
OKC
SYR
BILPRO
LVL
PGH
TDO
POO
CDRPHL BUF
SLK
DES
CSC
FCL
GBY TUC
EUG
WIL
DTN MEM
GSC
LSV
CHI LAN
MAD
COS
GRY
SAG
HOU
KAL
MEL
DAL AKR
SLM
INC
ROC
NO
DET GRR
YRK
ABQ
TAC
BIR
NSHFRO VIS
DAB
CLE
BAL HALWPB
TUL
CTN
FTL
DEN
WCH
BAT
PHX
MIA
HMOHBG
LRA
KNX
LEXNFK
EVN
CIN
CSC
KCM
COL
STL IND
BAK
RLGRNOSRA
CTE
AUG
ATL
RCH
OMH
GNCORL
FLT
MIN
JKL
TPA
16
18
20
22
24
26
28
30
32
Malpezzi Chun & Green Regulatory Index
©1999, S. Malpezzi
(Linear
Standardized
Coefficient
-0.10
t-Statistic
Prob > |t|
-1.1
.2877
0.26
2.7
.0073
0.42
5.3
.0001
-0.00
-2.9
.0046
Adjusted R-squared: 0.21
Degrees of freedom: 125
Fit)
Supply elasticities are even more difficult to measure for commercial real estate
Commercial Real Estate Price Indexes
based on NCREIF data
• Data
• “Identifying restrictions:” we don’t know as much about demand for office space as we know about the demand for housing
• Still, a fertile field, given some time?
• A quick test: what can we infer about supply elasticities, if asset prices are rising? Flat? Falling?
• National Council of Real Estate Investment Fiduciaries tracks “institutional” prices back to 1978.
– Problems with sample selection, and “appraisal smoothing.”
• Fisher, Geltner and Pollakowski, Journal of Real Estate Finance and Economics, 2007: Get NCREIF’s raw data, base the index on transactions.
– Lose a lot of data, but the data you have is much better. Should reduce “smoothing bias.”
– Still subject to sample selection bias!!!
– Updated data available from MIT’s website.
• Next slide splices the original index for early years with the more recent transaction index. Serious problems remain but the best we have over any long time frame.
178
179
9
Real Commercial Real Estate Price Index
Supply elasticity of commercial real estate
Original NCREIF Capital Index 1978‐1983; MIT Transaction Index 1984‐2010
150.00
140.00
• At least in the aggregate, until (say) 2005 or 2006, CRE prices were broadly consistent with fairly elastic supply.
• While scale/timing was different, CRE asset prices saw a boom and bust cycle post 2005 that was broadly parallel to housing’s. (Housing prices peaked early in 2006, CRE prices peaked about a year later).
• Boom and bust in recent years: evidence of inelastic supply, or something else?
• Note: nobody’s yet estimated supply elasticities of CRE directly, by market; or looked at determinants of same.
130.00
120.00
110.00
100.00
90.00
80.00
70.00
60.00
78‐1
79‐1
80‐1
81‐1
82‐1
83‐1
84‐1
85‐1
86‐1
87‐1
88‐1
89‐1
90‐1
91‐1
92‐1
93‐1
94‐1
95‐1
96‐1
97‐1
98‐1
99‐1
00‐1
01‐1
02‐1
03‐1
04‐1
05‐1
06‐1
07‐1
08‐1
09‐1
10‐1
50.00
180
181
Back to the lab, again…
• We need a lot more work on supply elasticities
– By market
– By property type
– Understanding dynamics (the time path of adjustment)
– Understanding the role of expectations
182
10