Content uploaded by Luiz Andrés Paixão
Author content
All content in this area was uploaded by Luiz Andrés Paixão on Dec 30, 2020
Content may be subject to copyright.
Content uploaded by Luiz Andrés Paixão
Author content
All content in this area was uploaded by Luiz Andrés Paixão on Dec 30, 2020
Content may be subject to copyright.
Hedonic price methods and real estate price index: an explanatory study for apartments
market in Belo Horizonte, Brazil from 2004 to 2015.
Luiz Andrés Paixão
(luiz.paixao@ibge.gov.br)
Instituto Brasileiro de Geografia e Estatística (IBGE)
Abstract
Brazil has not yet an official real estate price index. However, due to the great
valuation of real estate prices in the last years and the 2008’s subprime crisis, the demand for
such indexes has increased. A 2011 presidential decree stipulated for the Brazilian Institute
of Geography and Statistics (IBGE) the assignment to create and spread a real estate price
index for Brazil. As a result, attention has turned to study both methods and database
affordable for a future Brazilian official real estate price index.
Most official statistics institutes around the word have used the hedonic price regression to
estimate the real estate price index, since this methodology is appropriate to deal with
composed goods. In this paper, we test some different hedonic model methods to estimate
monthly, quarterly and annual price indexes for apartments in Belo Horizonte, Brazil, from
January 2004 to December 2015. The data set comprises all apartments transactions in the
analyzed period from Belo Horizonte's real estate transmission tax.
Our goals are: i) to measure and compare the different hedonic methods; ii) to present
some results that will contribute to the discussion towards the development of an official real
estate price index in Brazil. The empirical results corroborate the idea of intense apartment
prices valuation in Belo Horizonte, mainly between 2007 and 2011, when the annual price
growth taxes remained above 20%. However, in the two last years of the analyzed period,
the annual growth tax has decreased below 5%. These results shed light on the potential use
of both hedonic methods and administrative data base to construct an official real estate price
index for Brazil.
1. -Introduction
Subprime crises turned attention around the world on the real estate price dynamics
question. In Brazil the recent large valuation of real estate price adds more attention to the
subject. Academics, news, real estate agents, Brazilians government agencies and statistical
institutes started discussing the importance of having an adequate measure of real estate price
over time. Federal Government Decree number 7.565 from 21 September, 2015, established
IBGE (Brazilian Institute of Geography and Statistics) as the responsible to create and
disseminate an official real estate price index for Brazil. Since then, IBGE and others
governmental agencies have implemented studies regarding database and methodology to
construct a future Brazilian official real estate price index (Nadalin and Furtado, 2011; Santos
and Salazar, 2011).
The Brazilian academy has been studying real estate price index and its application
in Brazilian context. In recent years, Rozenbaum (2009), Paixão (2015a) and Simões (2017)
are examples of doctoral thesis related to this subject. Rozenbaum (2009) used administrative
data to construct a hedonic quality adjusted price index for the city of Rio de Janeiro. Simões
(2017) also measure the hedonic quality adjusted price index for Rio de Janeiro, using real
estate agencies data, while Paixão (2015), also used administrative data to construct a hedonic
quality adjusted price index for Belo Horizonte’s city.
Some researches in this field were published in Brazilian academic journals.
Gonzalez (1997) estimated a simple time-dummy hedonic model to construct price index for
apartments rents in Porto Alegre. The same approach was used by Rozenbaum and Macedo-
Soares (2007), to estimated real estate valuation in Rio de Janeiro’s district of Barra da
Tijuca, and by Paixão (2015a), to estimated real estate price indices for Belo Horizonte.
Albuquerque et al. (2018) used repeated sales method to construct an index for the city of
Brasilia, the capital of Brazil.
Some University’s agencies like FIPE, from USP, and IPEAD, from UFMG, released
real estate price indices using stratified median methods. The widespread FIPE-ZAP real
estate index is calculated by FIPE from real estate’s advertised data collected in ZAP’s web
site platform. The Brazilian Central Bank also estimated and published a monthly stratified
median real estate price index, constructed from real estate loans data, called IVG-R. Despite
the importance and relevance of those indices for the society, government, academics and
real estate agencies some gaps still remain. None of those indices used the hedonic quality
adjustment methodology, recognized as the best to deal with the nature of real estate’s market
data (Diewert, 2009; Hill et al, 2018). Besides that, only IPEAD uses administrative data
which cover overall transacted dwellings market.
In this paper we try to construct quarterly hedonics quality adjusted real estate prices
indices for Belo Horizonte’s city using administrative data. For this task we will use the
hedonics methods proposed by Hill’s (2013) and Hill’s et al (2018). We will use the same
methods used by Hill et al (2018) in their analysis of Sidney and Tokyo markets, to produce
comparable results.
The reminder of the paper is structured as follows. The next section explains the
different hedonic price methods used to measure quality adjusted housing price indices. The
third section is focused in the database and introduces the Brazilian’s city of Belo Horizonte.
The objective of the fourth section is to test the several hedonic methods in a Belo Horizonte’s
real estate market database. Finally, our main results are summarized in the conclusion.
2. Hedonic Quality Adjusted Real Estate Price Methods
2.1 – The Hedonic Price Model
Estimating housing price indices is a complex task. Housing is a type of complex
good (or service), that is, a good where each unit or model differs from the others in
qualitative terms. A complex good can be described as a bundle of many characteristics (or
attributes), so each unit or model is a peculiar bundle of attributes. Computing index prices
for complex good necessarily means controlling the change in the good price by the change
in composition of its characteristics.
The hedonic price model establishes a functional relationship between the price of the
good and its characteristics. In a hedonic perspective a good is a basket of characteristics ,
as represented below:
The price of a good follows a hedonic function as describes in (2):
Although only the price of the good can be observed in the market, the hedonic
function establishes that the price of a good is determined by the composition of basket of
characteristics. Therefore, each attribute () has an unobserved price (implicit price) that is
represented by the first derivative of the hedonic function with respect to .
The seminal paper of Rosen (1974) validated the hedonic price model in theoretical
terms. Empirically, Waugh (1928) was pioneer in apply hedonic regression in vegetables
market study. Court (1939) used hedonic price regression to construct automobile price
indices. Griliches (1958, 1961) constructed hedonic quality adjusted price indices for
fertilizers and automobile markets respectively. From Griliches contributions, the application
of hedonic model widespread in the academic world, covering many types of different goods
and services like computers, refrigerators, fruits, musical instruments, paints etc. However,
it was in the real estate market that the hedonic approach achieved its largest projection.
2.2 – The Hedonic Quality Adjusted Price Indices: the real estate case
Griliches (1971) argues that a complex good’s price change can be divided in two
dimensions. The first is the observed price change of the good in the market. The second is
the unobserved price change of the basket of characteristics. To estimate the unobserved price
change, it is necessary to use the hedonic model regression as a quality adjusted factor.
Discounting the price change of the attributes bundle from the observed price of the good
results in a “pure” estimate of a complex good price change.
There are several ways to construct a quality adjusted price index from the hedonic
methodology. Court (1939) and Griliches (1961) already advanced some questions, like the
possibility of using both cross-section regressions or time-dummy approaches. Tripllet
(2004) created a taxonomy of the several hedonic methods used to compute quality adjusted
price indices for technological goods. Hill (2013) applied this Tripllest’s taxonomy to the
housing market case.
Hill et al (2018) using a Hill (2013) approach compiled the hedonic methods used by
the European national statistics institutes. The first category embraces all indices which
requires cross-section regressions models and, as a result, involves data imputation. The
second category is based on time-dummy regressions.
2.2.1 – Imputation Approach
2.2.1.1 – Repricing Model
The first imputation method described by Hill et al (2018) is the repricing method.
Like Hill et al (2018) in this study we adopted quarterly price indices as default and hedonic
quality adjusted price indices could be constructed for any period. Defining the base period
is the first task in the repricing model. Then a hedonic regression is estimated for this period.
The price implicit estimated in hedonic regression is used to impute prices for each
subsequent quarterly. The base period can be fixed or be updated at regular time intervals.
Hill et al (2018) recommends estimating one regression for the whole base year:
Where is the natural logarithmic of housing price in the base year (, is the
quarterly of the sale, is the dwelling sold and is each characteristic of the dwelling.
Then the implicit prices estimated () are used to estimated prices for each subsequent
quarterly. The repricing method is, therefore, a sort of Laspeyres index. The quality
adjustment factor is defined as the ratio of the imputed prices for adjacent
quarters, and for example.
To construct the repricing method price index (RP) a quality unadjusted price index
() defined as a ratio between geometric mean prices of adjacent quarters
is calculated as follows:
Finally, RP is the ratio between quality unadjusted price factor and quality adjusted
price factor.
The main attractive feature of RP relies in the fact that it is not regression intensive.
In the end, it requires only one regression (Hill et al, 2018). However, to achieve good results,
Hill et al (2018, 224) suggested that “the base year under the repricing method should be
updated at regular time intervals”. Italy and Luxembourg national statistics institutes are
benchmark examples since both updated the base every year (Hill et al, 2018).
2.2.1.2 – Average Characteristics Method
The average characteristic method (AC) requires, as any method, a definition of a
base period. After that, the average characteristic of the dwellings sold in the base period are
computed. The next step consists in estimating hedonic regressions for each subsequent
period, quarterly in our case. Then the imputed prices are calculated, applying the estimated
quarterly implicit prices on the average characteristics of the base period.
Following Hill et al (2018), the European national statistical institutes calculate de
the basket of average characteristics for a whole year (base year), . In this line, the
European national statistical institutes adopted a Laspeyres version of AC1. The base is
updated every year. Then a hedonic regression is estimated for each quarter ( ) of the
following year (:
The quality adjusted price index estimated by AC is given as follows:
2.2.1.3 – Hedonic Imputation Method
Real estate is a threshold situation of complex goods. Each unity of real estate differs
from the other. Added to this, the set of dwellings sale on one period differs of the set of
dwellings sale in other periods. Therefore, it is not possible to construct a basket of dwellings
to follow over time. The hedonic imputation method is a way to estimate the price each
dwelling sold in would have in another period, for example. According to Hill et al
(2018, 225-6):
Once a hedonic model has been estimated, it allows one to ask
counterfactual questions such as what a particular dwelling actually
sold in say period would have sold for instead in period .
Like the AC, one regression as (8) is estimated for each period. The regression in
is used to impute the price in for each observed transacted dwelling in . Likewise,
the regression in is used to impute the price in for each observed transacted dwelling in
. To construct the index, Hill (2013) recommended to use the regression estimated price
in instead of the observed price for each observed transacted dwelling in
1
. Such procedure
is known as double imputation. From the hedonic imputation method geometric Laspeyres
(GL), geometric Paasche (GP) and Tornqvist prices indices can be extracted.
Few European national statistical institutes use hedonic imputation methods. Hill et
al (2018) follow the German version of double imputation Tornqvist (DIT). From a set of
regressions like (8) the GL, GP and DIT are estimated as follows:
1
Hill and Melser (2008) demonstrated in real estate case that double imputation is a way to minimize the
omitted variable bias.
2.2.3. – Time-Dummy Approach
2.2.3.1 – Simple Time-Dummy
The time-dummy approach consists in constructing price indices from the estimated
parameter of a dummy time variable in a hedonic regression. Usually, the first period in the
series is used as the base. The simple time-dummy model (TD) requires only one regression
and it is the simples and most intuitive hedonic method. The typical TD regression is
illustrated as follows:
Where is a time dummy for each period and is the prince index estimated for
the period .
Despite its simplicity, there are some pitfalls in using TD (Hill, 2013). First, the TD
does not allow the implicit price changes over time. As a result, the longer the series, the
worst will the TD prince index estimations be. For national statistical institutes, TD is not
recommended because it does not follow the temporal fixity criterium, as defined by Hill
(2004). According this criterium, once an index has already been disseminated by the national
statistical institute it should remain unchanged when new data becomes available. Using the
single regression TD approach, when new data are added, a new estimation of (13) changes
all parameters previously disseminated.
2.2.3.2 – Rolling Time Dummy Method
The hedonic rolling time dummy method (RTD) consists in estimated hedonics time-
dummy regressions for subperiods instead of only one regression for the whole period. The
limiting case occurs when a regression is estimated for each pair of adjacent periods.
Although when data points are scarce its recommended to estimate a regression including
more than one subperiod.
France and Portugal, for example, estimated price index from an RTD with 2 quarter
windows. In other words, both countries are using an adjacent period RTD. Other countries
like Cyprus and Croatia estimated a 4 quarter windows RTD. The RTD price index is
calculated from an RTD regression like (13) as showed below:
3. The data
3.1 – Belo Horizonte: an overview
Belo Horizonte, the capital of the State of Minas Gerais, is an important economic, politic
and cultural center in Brazil. According to 2010’s Brazilian Census (IBGE, 2010), Belo
Horizonte had a population of almost 2,4 million and was the 6th most populous city in Brazil.
The Metropolitan Area (MA) of Belo Horizonte, in turn, had a population of 5,4 million and
was the 3th most populous MA in Brazil.
Belo Horizonte was a planned city, conceived to replace Ouro Preto as the capital of
Minas Gerais, and was founded in 1897. Nowadays, the planned area corresponds to
downtown and its nearby districts bounded by Contorno Avenue. Like Aguiar et al (2014,
119) resumes:
This planning created a center-periphery radial model for the city, which
concentrated urban services and urban infrastructure in particular areas, and
reinforced social disparities.
From an administrative point of view, the space of Belo Horizonte is divided in
Districts (487), Planning Units (82) and Regionals (9), see Figure A1 in appendix.
Following Villaça (1998), historically, the Central-South city’s Regional (Regional
Centro-Sul) concentrated the elite’s neighborhoods. Nowadays, some bordering Central-
South Regional districts in the West Regional (Regional Oeste) are also occupied by Belo
Horizonte’s elite. Nonetheless,a few elite’s districts are in Pampulha Regional (Regional
Pampulha), in the north of the city, surrounding Pampulha’s lagoon.
3.2 The Data
In Brazil all real estate transactions are subjected to The Real Estate Transfer Tax
2
(RETT) and its collection is in charge of municipalities. We used Belo Horizonte’s
municipality’s RETT as our dataset, covering the 2004 to 2015 period, collected by
IPEAD/UFMG. The RETT dataset contains the value of transaction, type of building, area,
age, quality of building finishing material, zoning and location (district). The type of building
includes apartments, houses and commercial real estate. In this paper we analyzed only
apartments market. Further we shall expand the analysis for house and commercial real
estate’s markets.
Tables 1, 2 and 3 resume the data. There were 266.529 observations in our dataset
for the whole period. The mean apartment price was around R$ 218.149 (approximately U$
57 000) and the standard deviation was 235.011, indicating a high dispersion of this variable.
Apartments sold in Belo Horizonte were fairy big and new, the mean area and age was 120
m2 and 13 years, respectively. Most of apartments were classified as normal in terms of
quality of building finishing material and this variable classified the quality in 5 categories.
Ordering from the top there were the following categories: luxury, high, normal, low and
popular. Most observations where located in the Center-South (Centro-Sul) and West (Oeste)
Regionals.
2
Imposto de Transmissão Imobiliária Inter-Vivos (ITBI).
Mean Median
Standard
deviation
Mean Median
Standard
deviation
Mean Median
Standard
deviation
2004 17.767 90.682 60.122 91.427 122,0 102,9 70,6 12,4 8,0 11,7
2005 39.606 93.363 61.746 99.219 118,7 99,1 68,9 13,7 10,0 12,0
2006 37.614 107.572 70.196 117.326 120,4 102,1 69,2 14,3 10,0 12,2
2007 19.664 125.848 83.000 129.249 120,9 103,2 68,4 14,3 10,0 12,2
2008 19.224 155.117 100.000 156.457 121,4 101,7 70,6 14,5 10,0 12,5
2009 18.272 186.106 130.000 172.571 118,4 98,9 69,8 14,0 10,0 12,8
2010 21.177 227.110 160.764 201.441 114,4 94,4 68,5 12,1 8,0 13,1
2011 19.257 297.875 220.000 247.637 118,9 98,3 70,3 11,7 7,0 13,3
2012 18.408 362.735 274.412 284.955 123,8 107,0 70,9 11,8 6,0 13,5
2013 20.364 389.815 303.982 283.168 120,5 104,4 67,9 11,1 4,0 13,7
2014 19.516 418.895 330.000 291.744 120,6 105,7 66,9 10,1 2,0 13,3
2015 15.727 420.250 334.000 290.847 117,0 102,1 64,9 10,9 3,0 13,8
2004-2015 266.596 218.149 140.000 235.011 119,7 101,4 69,0 12,8 8,0 12,8
Sources: IPEAD/UFMG; author's calculation
Table 1 - Descriptive statistics for apartments in muicipali ty of Belo Horizonte: 2004-2015
Value (reais)
Area (m2)
Age (years)
Year
Observation
4 – Quality adjusted price indices for Belo Horizonte
4.1 – Model specification of the hedonics regressions
In section 2, we described the hedonic price model and its usefulness to design some
quality adjust price indices for real estate’s market. Since we adopted a log-linear estimation,
the dependent variable was the natural logarithmic of the apartment’s price. The set of
independent variables was composed by numeric and dummies variables. The set of numeric
variables was composed by area, age and squared age. It is expected that the higher is the
apartment, the higher it will be the price. Age is a proxy of depreciation, so we expected a
negative relationship between age and apartment’s price. The effect of age, however, is not
linear. Repairs and improvements reduce the depreciation age’s effects. In addition, some old
apartments are valued in the real estate market, the vintage effect phenomenon, which also
reduces the age’s negative effect in hedonic models (Goodman and Thibodeau, 1995).
The dummy variables set contains three subsets of variables. Apartment’s physical
conditions (quality of building finishing material) and apartment’s location characteristics
TABLE 2 - Distribution of quality buiding finishing materials: Belo Horizonte, 2004-2015
Quality of building finishing materials Mean Standard deviation
Popular 0,02 0,15
Low 0,21 0,41
Normal 0,59 0,49
High 0,16 0,37
Luxury 0,02 0,13
Sources: IPEAD/UFMG; author's calculation
TABLE 3 - Apartaments transactions - mean for Regional - Belo Horizonte: 2004-2015
Regional Mean Standard deviation
Centro-Sul 0,21 0,31
Leste 0,09 0,28
Nordeste 0,09 0,29
Noroeste 0,11 0,31
Oeste 0,20 0,40
Pampulha 0,15 0,35
Venda Nova 0,05 0,21
Barreiro 0,04 0,18
Norte 0,03 0,18
Sources: IPEAD/UFMG; author's calculation
(zoning and UP location). Analyzing the quality of building finishing material, Table 2 shows
that the worse category (popular) and superior category (luxury) represented a small share of
the market. The most usual was the basic category (normal), thereby we expected a positive
signal for luxury and high estimated parameters and negative signal for low and popular
estimated parameters.
Zoning is an important apartment’s price determinant. We chose as basic category
ZAP (intensification use preferential zone), which corresponds to areas where the
municipality encourages new buildings. ZPA and ZAR (restricted intensification uses zone)
corresponds to areas where news buildings are not allowed or encouraged due to the natural
or topographic conditions. Despite the new building restrictions, these areas are not valued
in housing market because of its lack of affordable natural conditions. ZA (dense zone)
represents high density areas where the municipality discourages new buildings. Commonly
ZA’s are in the most valued Belo Horizonte’s districts and since it represents a supply
restriction in a high demand context, we expected a positive estimated parameter on this
variable. ZE (urban’s infrastructure equipment zone) represents areas with great urban
equipment (like bus stations). ZEIS (special social interest zone) correspond to original
spontaneously occupation’s areas (like informal slums), which were formalized by
municipality. Both ZE and ZEIS represent poorly valued areas by the real estate’s market
agents, so we expected a negative signal for the estimated parameter on these variables. ZHIP
(over central zone), ZCBH (Belo Horizonte’s central zone), ZCVN (Venda Nova’s central
zone) and ZCBA (Barreiro’s central zone) represent central areas. ZHIP is Belo Horizonte’s
downtown and ZCBH represent the other districts, besides Belo Horizonte’s downtown - into
Contorno Avenue boundary. Both represent valued land’s locations – where firms and
families wish to be located – consequently we expected a positive signal for estimated
parameters on these variables. The central areas of Barreiro and Venda Nova’s distant
suburbs are represented by ZCBA and ZCVN. Since both zoning represent central areas in a
minor magnitude, the same effect described above for ZCBH and ZHIP could occur.
As described in section 2, the construction of hedonic quality adjusted price indices
is a regression intensive process. To illustrate the control variables included in the Z set of
dependent variables we estimated a hedonic price model with quarter fixed effects. We
estimated an ordinary least square (OLS) regression, as below:
Where represented quarter’s fixed effect. The local fixed effects were contained
on Z’s characteristics set. The regression (15) corresponds to a simple time-dummy method.
The results are expressed in tables 4, and 5.
All estimated parameters had the expected signal and were significant at 5%. The
marginal effect of area in apartments price was 0,67% which means that estimated shadow
price of an additional m2 for a mean price apartment was almost R$ 1.444. Each additional
year depreciated the apartment in almost 1,8%. Nonetheless, due to squared age effect, each
additional year added 0,016% on apartment value. Considering both age effects, the implicit
price of each additional year on the mean price apartment was R$ 4.001. Quality of building
finishing material classified as luxury added almost 10% on apartments price, while popular
discounted around 10% on Belo Horizonte’s apartment price.
Zoning was an important price’s determinant. The affluent neighbor’s downtown
areas into Contorno Avenue’s boundaries of ZCBH added 55,79% on apartments price. The
low supply of vacant land combined with a large demand for both commercial and residential
Variabel Estimated parameter Standard deviation t P-value
Constamt 10,49568 0,00514 2042,01 0,0000
Area 0,00664 0,00000942 705,04 0,0000
Age -0,01833 0,00012124 -151,19 0,0000
Age^2 0,00015915 0,00000295 53,89 0,0000
Luxury 0,10088 0,00414 24,39 0,0000
High 0,05128 0,00148 34,59 0,0000
Low -0,06656 0,00131 -50,69 0,0000
Popular -0,09517 0,00369 -25,79 0,0000
ZPA -0,053524 0,00857 -6,21 0,0000
ZAR -0,05477 0,00162 -33,83 0,0000
ZA 0,13924 0,00231 60,23 0,0000
ZHIP 0,07185 0,00564 12,74 0,0000
ZCBH 0,44336 0,00359 123,34 0,0000
ZCVN 0,06234 0,0196 3,18 0,0015
ZCBA 0,19982 0,01279 15,62 0,0000
ZE -0,17845 0,00704 -25,5 0,0000
ZEIS -0,23219 0,0289 -10,61 0,0000
Spatial fixed effect = yes
Time fixed effect = yes
Ajusted R2 0,9236
F 30399,1 0,0000
Sources: IPEAD/UFMG; authors calculation
Tabela 4 - Hedonic price model for Belo Horizonte's apartment market: 2004-2014
Belo Horizonte - 2004/2015
real estate caused this great valuation on ZCBH’s apartment prices. In contrast, apartments
located in ZEIS’s area suffered a considerable discount (26,14%). As mentioned above these
are areas with failed urban infrastructure, and most of them where slums originally. Finally,
it is noteworthy the ZCBA’s marginal effect, 22,12%, signaling a great demand for
apartments in Barreiros’s center.
Variable Estimated parameter Standard Deviation t P-value
Sagrada Família -0,0264 0,0041 -6,3667 0,0000
Floresta 0,0171 0,0040 4,2710 0,0000
Pompéia -0,1302 0,0071 -18,2411 0,0000
Santa Efigênia -0,0689 0,0049 -14,1919 0,0000
Santa Inês -0,0904 0,0069 -13,0247 0,0000
Cabana -0,3098 0,0066 -47,2627 0,0000
Jardim América -0,1329 0,0037 -35,6164 0,0000
Barroca 0,0205 0,0033 6,1366 0,0000
Betânia -0,2489 0,0053 -46,8326 0,0000
Buritis 0,0083 0,0035 2,3820 0,0172
Barro Preto -0,5034 0,0062 -81,4118 0,0000
Centro -0,1541 0,0051 -30,4518 0,0000
Prudente de Morais 0,1510 0,0043 34,7497 0,0000
Serra 0,0394 0,0043 9,1459 0,0000
São Bento 0,2052 0,0078 26,2196 0,0000
Belvedere 0,4503 0,0104 43,4269 0,0000
Anchieta 0,1415 0,0034 41,6524 0,0000
Glória -0,3642 0,0057 -64,0957 0,0000
Padre Eustáquio -0,0811 0,0040 -20,4647 0,0000
Camargos -0,3228 0,0049 -66,1975 0,0000
PUC -0,1169 0,0050 -23,2561 0,0000
Abílio Machado -0,2440 0,0056 -43,2920 0,0000
Caiçara -0,0622 0,0044 -14,1295 0,0000
Pampulha -0,1864 0,0088 -21,0600 0,0000
Santa Amélia -0,1895 0,0043 -43,9341 0,0000
Ouro Preto -0,0931 0,0043 -21,6291 0,0000
Jaraguá -0,1081 0,0041 -26,1436 0,0000
Castelo -0,1394 0,0036 -38,2734 0,0000
Table 5 - Estimated parameters for UP dummy (cont)
Regional Noroeste
Regional Centro-Sul
Regional Oeste
Regional Leste
Regional Pampulha
The location dummy variables were all statistically significant. UP Savassi was the
basic category. The signals were what we expected, except for the downtown’s UPs of Barro
Preto and Centro. Barro Preto, specifically, had the bigger negative marginal effect (-39,59%)
among all. Part of this puzzling results should be the effect of the positive marginal effects
of Barro Preto’s zonings codes (ZHIP and ZCBH). In contrast, Belvedere, locate in the
extreme south of Belo Horizonte, was the most valued UP, its price was, on average, 56,86%
above the basic category.
4.2 – Quality adjusted price indices for Belo Horizonte
We estimated quarterly housing price indices for Belo Horizonte using the various
methods discussed in sections 2.2.2 and 2.2.3 and an UP’s stratified median method (MIX-
UP). Since some methods were based on previous regression or mean characteristics, the
2004 data was used to compute reference baskets used in 2005. For this reason, results are
represented for years 2005-2015. Following Hill et al (2018) we estimated three forms of
repricing model (RP): i) (RP-X, which uses shadow price from 2004 (no updating base year);
Cachoerinha -0,2334 0,0053 -44,2551 0,0000
Concórdia -0,2021 0,0083 -24,4453 0,0000
Cristiano Machado -0,0216 0,0034 -6,3133 0,0000
São Paulo -0,3012 0,0048 -62,5666 0,0000
Planalto -0,2793 0,0054 -51,8408 0,0000
São Bernardo -0,3022 0,0065 -46,2063 0,0000
Primeiro de Maio -0,2980 0,0091 -32,7033 0,0000
Jaqueline/Tupi -0,4589 0,0056 -82,5165 0,0000
Barreiro de Baixo -0,3479 0,0045 -77,3910 0,0000
Cardoso -0,4978 0,0057 -87,8971 0,0000
Europa -0,3088 0,0059 -52,3095 0,0000
Venda Nova -0,2741 0,0043 -63,3277 0,0000
Source: IPEAD/UFMG; author's calculation
Regional Venda Nova
Regional Nordeste
Regional Barreiro
Table 5 - Estimated parameters for UP dummy (end)
Regional Norte
ii) RP-5, which updates the shadows prices every five years; iii) RP-1, which updates the
shadows prices every year.
The average characteristic indexes (AC) were estimated with a base update every
year. In AC’s case, the base is the one year lagged average characteristics. The double
imputation indexes were calculated estimating, for some quarter set of observations,
counterfactual housing’s basket for previous and posterior periods. The double imputation
Laspeyres (DIT), Paasche (DIP) and Tornqvist (DIT) were estimated as presented in section
2.2.1.3. The rolling time dummy indices were estimated for 2 (RTD2) and 4 (RTD4) quarters
window. Finally, we estimated an UP’s stratified median (MIX-UP) index, to compare the
quality adjusted hedonic housing price index with a simpler and more intuitive price index.
Table 6 resumes the quarter housing price indices estimated.
From the chosen period (2005-2015), the large appreciation of housing prices in
Brazil was supported by estimated indices. Even so, the magnitude of the appreciation differs
between methods. Using the DIP index, Belo Horizonte’s house prices rose 383,6% in
contrast with DIL which shows an increase of 435,7%.
RIP-X, which uses shadow price form 2004, was apartheid from RP-5, which changes
the base every five years, and RP-1, which updates shadow prices every year (Figure 1).
From our Belo Horizonte’s database, the RIP-X seems not be an appropriate index due to its
fail to control shadow prices change over time. RP-5 and RP-1 lines were close each other,
highlighting the importance to update the base year from time to time. Since RP-1 is the more
flexible RP index it will be used in the remainder of the paper.
Year Quarter RP-X RP-5 RP-1 AC DIL DIP DIT RTD4 RTD2 MIX -UP
2005 Q1 1,0000 1,0000 1,0000 1,0000 1,0000 1,0000 1,0000 1,0000 1,0000 1,0000
2005 Q2 1,0159 1,0159 1,0159 1,0122 1,0133 1,0132 1,0133 1,0151 1,0148 1,0160
2005 Q3 1,0300 1,0300 1,0300 1,0274 1,0287 1,0287 1,0287 1,0284 1,0280 1,0294
2005 Q4 1,0361 1,0361 1,0361 1,0361 1,0380 1,0381 1,0381 1,0374 1,0380 0,9913
2006 Q1 1,1176 1,1176 1,1176 1,1176 1,1183 1,1176 1,1179 1,1176 1,1184 1,0758
2006 Q2 1,1338 1,1338 1,1335 1,1306 1,1320 1,1282 1,1301 1,1305 1,1307 1,1339
2006 Q3 1,1522 1,1522 1,1544 1,1536 1,1543 1,1511 1,1527 1,1529 1,1522 1,1512
2006 Q4 1,1773 1,1773 1,1775 1,1767 1,1783 1,1752 1,1767 1,1759 1,1760 1,2444
2007 Q1 1,2737 1,2737 1,2755 1,2738 1,2758 1,2721 1,2739 1,2734 1,2728 1,2002
2007 Q2 1,3417 1,3417 1,3411 1,3404 1,3408 1,3351 1,3379 1,3380 1,3373 1,3514
2007 Q3 1,3781 1,3781 1,3768 1,3757 1,3767 1,3702 1,3734 1,3742 1,3740 1,4270
2007 Q4 1,4379 1,4379 1,4364 1,4344 1,4338 1,4301 1,4319 1,4325 1,4310 1,4483
2008 Q1 1,5394 1,5394 1,5397 1,5358 1,5357 1,5324 1,5340 1,5340 1,5305 1,5949
2008 Q2 1,6406 1,6406 1,6419 1,6380 1,6344 1,6311 1,6328 1,6303 1,6309 1,7305
2008 Q3 1,6941 1,6941 1,6951 1,6978 1,6891 1,6831 1,6861 1,6849 1,6841 1,8012
2008 Q4 1,7785 1,7785 1,7688 1,7438 1,7341 1,7390 1,7365 1,7451 1,7367 1,8468
2009 Q1 1,9184 1,9184 1,9172 1,9345 1,9065 1,9186 1,9125 1,9272 1,9180 1,9341
2009 Q2 2,0426 2,0252 2,0257 2,0425 2,0136 2,0299 2,0218 2,0346 2,0234 2,0699
2009 Q3 2,0891 2,1232 2,0977 2,1446 2,1267 2,1227 2,1247 2,1439 2,1286 2,1578
2009 Q4 2,2967 2,2535 2,2511 2,2620 2,2500 2,2437 2,2469 2,2718 2,2595 2,4103
2010 Q1 2,4744 2,4665 2,4741 2,5631 2,5172 2,4793 2,4982 2,4981 2,4937 2,4125
2010 Q2 2,4880 2,5573 2,5652 2,6623 2,6152 2,5917 2,6034 2,6342 2,6337 2,5304
2010 Q3 2,7862 2,7400 2,7484 2,7969 2,7748 2,7385 2,7566 2,7880 2,7802 2,8651
2010 Q4 2,9135 2,8074 2,8161 2,9123 2,8878 2,8011 2,8441 2,8545 2,8871 3,0194
2001 Q1 3,2437 3,1066 3,1162 3,2142 3,1937 3,0847 3,1387 3,1050 3,2067 3,1962
2001 Q2 3,4362 3,3175 3,3171 3,3540 3,3481 3,2471 3,2972 3,2563 3,3594 3,3921
2001 Q3 3,6513 3,5211 3,5426 3,5843 3,5391 3,4267 3,4824 3,4522 3,5629 3,7063
2001 Q4 3,7284 3,5977 3,6223 3,6786 3,6425 3,5187 3,5801 3,5345 3,6464 3,7071
2012 Q1 4,0576 3,8488 3,8256 3,9331 3,8747 3,7464 3,8100 3,7766 3,8917 4,1396
2012 Q2 4,1077 3,8928 3,8712 4,0276 3,9735 3,8167 3,8944 3,8407 3,9710 4,1294
2012 Q3 4,1714 3,9783 3,9776 4,1165 4,0657 3,8808 3,9722 3,9430 4,0697 4,2109
2012 Q4 4,3150 4,0642 4,0950 4,2527 4,2365 4,0178 4,1257 4,0922 4,2239 4,6228
2013 Q1 4,5945 4,4102 4,3938 4,5606 4,5262 4,2977 4,4105 4,3651 4,5180 4,5493
2013 Q2 4,6607 4,4861 4,5009 4,6730 4,6725 4,3679 4,5176 4,4663 4,6086 4,5502
2013 Q3 4,7261 4,5590 4,5954 4,7992 4,8126 4,4874 4,6471 4,5974 4,7413 4,4635
2013 Q4 4,9766 4,6991 4,7265 4,9220 4,9485 4,6035 4,7729 4,7176 4,8628 4,9431
2014 Q1 5,2788 4,9844 4,9813 5,1751 5,2173 4,8410 5,0256 4,9716 5,1357 5,2863
2014 Q2 5,2140 4,9427 4,9356 5,1372 5,1858 4,7813 4,9794 4,9178 5,0871 5,1202
2014 Q3 5,1184 4,8852 4,8634 5,0797 5,1412 4,7105 4,9212 4,8619 5,0297 4,8198
2014 Q4 5,1985 4,9858 4,9885 5,1283 5,2087 4,7432 4,9705 4,9148 5,0874 5,0084
2015 Q1 5,4096 5,1672 5,1846 5,3642 5,4633 4,9590 5,2050 5,1404 5,3163 4,9899
2015 Q2 5,4479 5,1016 5,1188 5,2688 5,3796 4,8807 5,1241 5,0599 5,2280 5,1586
2015 Q3 5,3027 5,0147 5,0316 5,2519 5,3478 4,8320 5,0833 5,0190 5,1943 4,9577
2015 Q4 5,4044 5,0481 5,0651 5,2719 5,3570 4,8362 5,0900 5,0308 5,1950 5,1515
Sources: IPEAD/UFMG, author's calculation
Table 6 - Esimates of Price Indices for Apartments in Belo Horizonte (2005 Q1=100), acoording to different methods
Sources: IPEAD/UFMG; author’s calculation
The double imputation Laspeyres (DIL) and double imputation Paasche (DIP)
showed evidence of drift (Figure 2), as Hill el al (2018) also have noted for Sydney data. For
Belo Horizonte apartment’s market DIP estimated the smallest price variation while DIL
estimated the biggest. In agreement with price index theory, DIL, as a Laspeyres index, tends
to overestimate the price change and DIP, as a Paasche index, tends to underestimate it.
Double imputation Tornqvist (DIT), in any case, does not exhibit a drift. DIT, as a Tornqvist-
Geometric index, is a geometric mean of DIP and DIL. Bearing in mind that the Tornqvist
indices are recognized as superlatives, the DIT becomes an attractive alternative method to
compute housing indexs prices. Since DIP a DIL exhibited a drift behavior, the imputation
methods will be reduced to DIT in the following analysis.
The rolling time dummies (RT) were estimated for 4 quarter window (RT4) and 2
quarter window (RT2) from our database. On the RT4 index the quarter base changes once
a year, and on RT2 the base changes every quarter. RTs methods are attractive because the
index corresponds to the estimated regression time dummy parameter. RT2 indices stayed
above RT4 for the whole period. Both RT indices will be kept in the following analyses.
The Figure 3 compares the Belo Horizonte’s hosing price indices estimated by
different hedonic methods and by stratified median.
0
1
2
3
4
5
6
2005Q1
2005Q2
2005Q3
2005Q4
2006Q1
2006Q2
2006Q3
2006Q4
2007Q1
2007Q2
2007Q3
2007Q4
2008Q1
2008Q2
2008Q3
2008Q4
2009Q1
2009Q2
2009Q3
2009Q4
2010Q1
2010Q2
2010Q3
2010Q4
2011Q1
2011Q2
2011Q3
2011Q4
2012Q1
2012Q2
2012Q3
2012Q4
2013Q1
2013Q2
2013Q3
2013Q4
2014Q1
2014Q2
2014Q3
2014Q4
2015Q1
2015Q2
2015Q3
2015Q4
FIGURE 1 - Repricing Indices for Apartments in Belo Horizonte
(2005Q1=100)
RP-X RP-5 RP-1
Source: IPEAD/UFMG, author’s calculation
Source: IPEAD/UFMG, author’s calculation
0,0000
1,0000
2,0000
3,0000
4,0000
5,0000
6,0000
2005Q1
2005Q2
2005Q3
2005Q4
2006Q1
2006Q2
2006Q3
2006Q4
2007Q1
2007Q2
2007Q3
2007Q4
2008Q1
2008Q2
2008Q3
2008Q4
2009Q1
2009Q2
2009Q3
2009Q4
2010Q1
2010Q2
2010Q3
2010Q4
2011Q1
2011Q2
2011Q3
2011Q4
2012Q1
2012Q2
2012Q3
2012Q4
2013Q1
2013Q2
2013Q3
2013Q4
2014Q1
2014Q2
2014Q3
2014Q4
2015Q1
2015Q2
2015Q3
2015Q4
FIGURE 2 - Double Imputation Indices for Apartments in Belo Horizonte
(2005Q1=100)
DIL DIP DIT
0,0000
1,0000
2,0000
3,0000
4,0000
5,0000
6,0000
2005Q1
2005Q2
2005Q3
2005Q4
2006Q1
2006Q2
2006Q3
2006Q4
2007Q1
2007Q2
2007Q3
2007Q4
2008Q1
2008Q2
2008Q3
2008Q4
2009Q1
2009Q2
2009Q3
2009Q4
2010Q1
2010Q2
2010Q3
2010Q4
2011Q1
2011Q2
2011Q3
2011Q4
2012Q1
2012Q2
2012Q3
2012Q4
2013Q1
2013Q2
2013Q3
2013Q4
2014Q1
2014Q2
2014Q3
2014Q4
2015Q1
2015Q2
2015Q3
2015Q4
Figure 3 - Estimate of Price Indices for Apartments in Belo Horizonte (2005Q1=100)
RP-1 AC DIT RTD4 RTD2 MIX -UP
DIT, RP1 and RT4 exhibited a very close behavior. AC and RT2 do the same,
although the latter stayed above the former in the most recent quarters. Partly, the different
behavior between AC and DIT is expected, since the first is a Laspeyres type of index and
the second is a Tornqvist type. The MIX-UP line was more volatile than the other indicex
lines due to the lack of characteristics control related to this method.
Hill et al (2018) recommended analyzing the volatility of indices in more details. The
authors present two volatility measures: the root mean squared error (RMSE) and mean
absolute deviation (MAD). Also, they calculated the minimum (MIN) and maximum (MAX)
value for each index. All these indicators are computed both on a year-by-year and quarter-
by-quarter basis. The indicators volatility formulas are specified below:
The results are summarized in a Table 7.
MIX-UP is more volatile than the other indices. As Hill et al (2018) pointed out, it is
expected for stratified median indices (like MIX-UP) to exhibit more volatility since they are
not adjusted for changes in the quality of median over time. The hedonics quality adjusted
indices exhibited relative low volatility, the magnitudes were between those which were
estimated for Sidney and Tokyo by Hill et al (2018). From volatility indicators perspectives
our results suggest, for Belo Horizonte’s housing market in 2005-2015, that the hedonic
quality adjusted housing price indices were accurate, except for the MIX-UP cases.
4.3 – Apartment Valuation in Belo Horizonte: 2005-2015
We will illustrate the previous results measuring quarterly apartments prices rate of
appreciation for Belo Horizonte. Table 8 summarized the results for different methodologies.
RP-X RP-5 RP-1 AC DIL DIP DIT RTD4 RTD2 MIX-UP
RMSE 0,0723 0,0642 0,0628 0,0678 0,0636 0,0654 0,0642 0,0633 0,0667 0,1032
MAD 0,0629 0,0569 0,0551 0,0575 0,0524 0,0559 0,0541 0,0537 0,0564 0,0869
MIN 2,4779 3,6662 4,0808 3,6532 4,7136 2,4382 3,5696 3,3962 3,5155 -5,6054
MAX 31,0888 28,5716 29,0456 32,4961 32,0324 29,2254 30,6214 29,6188 30,0129 32,8800
RMSE 0,0723 0,0656 0,0649 0,0684 0,0649 0,0664 0,0655 0,0647 0,0681 0,0777
MAD 0,0558 0,0533 0,0530 0,0559 0,0506 0,0548 0,0527 0,0515 0,0542 0,0632
MIN 2,4779 3,6662 4,0808 3,6532 4,7136 2,4382 3,5696 3,3962 3,5155 -5,6054
MAX 31,0888 28,5716 29,0456 32,4961 32,0324 29,2254 30,6214 29,6188 30,0129 32,8800
RMSE 0,0773 0,0739 0,0757 0,0750 0,0715 0,0739 0,0726 0,0723 0,0751 0,0846
MAD 0,0668 0,0651 0,0656 0,0636 0,0594 0,0635 0,0614 0,0608 0,0638 0,0766
MIN 3,6008 2,6512 3,4586 3,3906 4,0176 2,5779 3,2952 3,2302 3,2725 2,8609
MAX 33,3710 29,0503 31,0202 30,4117 30,4740 29,0106 29,7403 30,0454 30,6093 32,7789
RMSE 0,0753 0,0750 0,0761 0,0774 0,0748 0,0769 0,0757 0,0747 0,0776 0,0896
MAD 0,0654 0,0651 0,0661 0,0653 0,0616 0,0655 0,0635 0,0622 0,0653 0,0800
MIN 3,9608 1,2501 1,5353 2,7994 2,8468 1,9609 2,4029 2,3601 2,1151 1,3200
MAX 29,1409 28,1510 28,6306 29,7209 29,7547 29,0226 29,3882 30,1850 30,1020 30,5135
RMSE 0,0326 0,0294 0,0291 0,0306 0,0289 0,0294 0,0290 0,0283 0,0294 0,0470
MAD 0,0277 0,0253 0,0243 0,0243 0,0233 0,0243 0,0237 0,0237 0,0242 0,0388
MIN -2,6650 -1,7031 -1,7031 -1,7786 -1,5315 -1,5806 -1,5560 -1,5670 -1,6609 -5,8666
MAX 11,9847 10,6583 10,6583 13,3085 11,8749 10,4981 11,1844 10,4400 11,0705 13,2288
Sources: IPEAD/UFMG; author's calculation
Quater-on-Quarter
Table 7 - Volatily of The House Price Indices in Belo Horizonte
Year-on-Year (Q1)
Year-on-Year (Q2)
Year-on-Year (Q3)
Year-on-Year (Q4)
Year Quarter RP-1 AC DIP RTD4 RTD2 MIX -UP
2005 Q2 1,59 1,22 1,32 1,51 1,48 1,60
2005 Q3 1,39 1,50 1,52 1,31 1,30 1,32
2005 Q4 0,59 0,85 0,92 0,88 0,97 -3,70
2006 Q1 7,86 7,86 7,65 7,73 7,75 8,52
2006 Q2 1,43 1,16 0,95 1,16 1,10 5,40
2006 Q3 1,84 2,04 2,03 1,98 1,90 1,53
2006 Q4 2,00 2,00 2,09 1,99 2,06 8,09
2007 Q1 8,33 8,25 8,25 8,30 8,24 -3,55
2007 Q2 5,14 5,23 4,95 5,08 5,07 12,59
2007 Q3 2,66 2,63 2,63 2,71 2,74 5,59
2007 Q4 4,33 4,27 4,37 4,24 4,15 1,49
2008 Q1 7,19 7,07 7,16 7,09 6,96 10,12
2008 Q2 6,64 6,65 6,44 6,28 6,56 8,50
2008 Q3 3,24 3,65 3,19 3,35 3,26 4,09
2008 Q4 4,34 2,71 3,33 3,57 3,12 2,53
2009 Q1 8,39 10,94 10,32 10,44 10,44 4,73
2009 Q2 5,66 5,59 5,80 5,57 5,49 7,02
2009 Q3 3,55 5,00 4,57 5,37 5,20 4,25
2009 Q4 7,31 5,47 5,70 5,97 6,15 11,70
2010 Q1 9,91 13,31 10,50 9,96 10,36 0,09
2010 Q2 3,68 3,87 4,53 5,45 5,62 4,89
2010 Q3 7,14 5,05 5,66 5,84 5,56 13,23
2010 Q4 2,46 4,13 2,29 2,38 3,85 5,38
2011 Q1 10,66 10,37 10,12 8,78 11,07 5,86
2011 Q2 6,45 4,35 5,27 4,87 4,76 6,13
2011 Q3 6,80 6,87 5,53 6,01 6,06 9,26
2011 Q4 2,25 2,63 2,69 2,38 2,34 0,02
2012 Q1 5,61 6,92 6,47 6,85 6,73 11,67
2012 Q2 1,19 2,40 1,88 1,70 2,04 -0,25
2012 Q3 2,75 2,21 1,68 2,66 2,49 1,97
2012 Q4 2,95 3,31 3,53 3,78 3,79 9,78
2013 Q1 7,30 7,24 6,97 6,67 6,96 -1,59
2013 Q2 2,44 2,46 1,63 2,32 2,01 0,02
2013 Q3 2,10 2,70 2,74 2,94 2,88 -1,91
2013 Q4 2,85 2,56 2,59 2,61 2,56 10,75
2014 Q1 5,39 5,14 5,16 5,38 5,61 6,94
2014 Q2 -0,92 -0,73 -1,23 -1,08 -0,95 -3,14
2014 Q3 -1,46 -1,12 -1,48 -1,14 -1,13 -5,87
2014 Q4 2,57 0,96 0,69 1,09 1,15 3,91
2015 Q1 3,93 4,60 4,55 4,59 4,50 -0,37
2015 Q2 -1,27 -1,78 -1,58 -1,57 -1,66 3,38
2015 Q3 -1,70 -0,32 -1,00 -0,81 -0,64 -3,90
2015 Q4 0,67 0,38 0,09 0,24 0,01 3,91
Sources: IPEAD/UFMG, author's calculation
Table 8 - Rate of appreciation (%) of apartment prices in Belo Horizonte: 2005-2015
From the first quarter of 2007 to the first quarter of the first quarter of 2014 there was
an intense apartment’s price appreciation in Belo Horizonte. This situation agrees with
Brazilian real estate outlook. This appreciation was contemporary with the expansion rate of
housing credit in Brazil. Some institutional improves like fiduciary alienation law’s
refinement in 2004 agreed with income growth and the decline interest rates helped the
housing credit’s growth (Aguiar, 2014). Cardoso and Leal (2009) highlighted the government
politics and the restructuring (more market concentration) of real estate development’s firm
role in the real estate market expansion.
Source: IPEAD/UFMG, author’s calculation
The figure 4 shows the quarterly variation of Belo Horizonte’s apartments prices. It’s
clear in figure the great volatility of the median index (MIX-UP), as we seen in the previous
section. From 2005 to 2011 there was a significant housing prices growth path, from then
there was a declined tendency. After 2014 quarter 2, the decline was more intense due to the
Brazil’s economic crises which began in this period.
Conclusion
Brazil does not have an official price index yet. In this paper we used a database from
Belo Horizonte, a big Brazilian city, to test some hedonic quality adjusted price index. The
indices constructed was the same using for European Statistical Institutes as described by
Hill et al (2018). Our results suggested that the hedonic quality adjusted indices exhibited a
good performance in volatility terms. However, it was detected some drift in double
-4,0
-2,0
0,0
2,0
4,0
6,0
8,0
10,0
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Figure 4 - Apartmente annual valuation rate (%) - Triennial moving
geometric average -Belo Horizonte: 2005-2015
RP-1
AC
DIP
RTD4
RTD2
MIX -UP
imputation Laspeyres and Paasche indices. The former with a strong upper-ward bias relative
to the other hedonic indices and the latter with a strong down-ward bias.
From our analyses, the double imputation Tornqvis (DIT) and the repricing with an
annual base’s update (RP-1) produced very similar magnitude’s indices. The same could be
said about the average characteristics (AC) and the 2-quarter rolling time dummy (RT2).
However, the index price lines of the latter stayed above the former. The 4-quarter rolling
time window, in its turn, exhibited an intermediate behavior, as compared with the previously
listed indices.
In contrast to Hil et al (2018) estimation for Sidney and Tokyo’s evidence, the
repricing method with no base update (RP-X) have an upper-ward bias relative to the other
hedonic indices. The five years update base repricing method, as well as Hill et al (2018)
empirical evidences, exhibited a down-ward bias. Finally, our models result suggested that
median indices are not the most appropriate to estimated housing price indices. The stratified
median index (MIX-UP) used was more volatile than the hedonic indices. This is because
this kind of index are imperfect in control for housing quality variation over time.
This paper emphasized the potentiality for constructed housing price indices in Brazil
using hedonic quality adjusted price methods and for administrative data. The Real Estate
Transfer Tax (RETT) emerges as hopeful database once this tax is collected in the whole
country. Further analysis could extend the hedonic price models to estimate price indices for
the Brazilian smaller city context and its less frequently housing sales reality. The smaller
number of observations imposes new challenges to estimated housing prices hedonic quality
adjusted indices. In addition, future analysis could extend the types of real estate units,
including houses and different types of commercial real estates.
References
AGUIAR, Marina Moreira de, SIMÕES, Rodrigo; GOLGHER, André Braz. Housing Market
Analysis using a hierarchical-spatial approach: the case of Belo Horizonte, Minas Gerais,
Brazil. Regional Studies, Regional Science, v.1, n.1, 2014, p. 116-137.
AGUIAR, Marina Moreira de. Crédito e mercado imobiliário no Brasil: fatores institucionais
e macroeconômicos (1964-2013). XVI SEMINÁRIO DE ECONOMIA MINEIRA,
Diamantina, 2014. Anais... Diamantina: CEDEPLAR/UFMG, 2014.
ALBUQUERQUE, Pedo; NADALIN, Vanessa; NETO, Vicente and MONTENEGRO,
Mariana. Construção de índices de preços de imóveis para o Distrito Federal por meio de
vendas repetidas e GWR. Nova Economia, v.28, n.1, p 181-212.
CARDOSO, Adauto Lúcio; LEAL, José Agostinho. Housing Markets in Brazil: recent
trends and governmental responses to the 2008 crisis. ISA INTERNATIONAL HOUSING
CONFERENCE, Glasgow, 2009. Anais… Glasgow: ISA, 2009.
COURT, Andrew. Hedonic price indexes with automotive examples. The Dynamic of
Automobile Demand: General Motors, New York, p.98-119, 1939.
DIEWERT, W. Erwin. The Paris OECD-IMF Workshop on Real Estate Price Indexes:
conclusions and future directions. In: DIEWERT, E. Erwin et al (Ed.). Price and
Productivity Measurement: volume 1 – housing. Trafford Press, 2009.
GONZÁLEZ, Marco. Variação qualitativa e índice de preços na análise do comportamento
recente dos aluguéis em Porto Alegre (1994-1997). Análise Econômica, v.15, n.28, 1997.
GOODMAN, Allen; THIBODEAU, Thomas. Age-related heteroskedasticity in hedonic
price equations. Journal of Housing Research, v. 6, n. 1, p. 25-42, 1995.
GRILICHES, Zvi. Hedonic Price indexes for automobiles: an econometric of quality change.
In: The Price Statistics of the Federal Government, General Series No. 73. Columbia
Univ. Press for the National Bureau of Economic Research, New York, p. 137-196, 1961.
GRILICHES, Zvi. Introduction: hedonic price indexes revisited. In: _____ (Ed.). Prices
indexes and quality change: studies in new methods of measurment. Cambridge: Harvard
University Press, 1971.
GRILICHES, ZVI. The Demand for fertilizer: an econometric reinterpretation of a technical
change. Journal of Farm Economics, v. 40, n. 3, p.591-606, 1958.
HILL, Robert and MELSER, Daniel. Hedonic imputation and the price index problem: an
application to housing. Economic Inquiry, v. 46, n. 4, p. 593-609, 2008.
HILL, Robert; SCHOLZ, Michael; SHIMIZU, Chihiro and STEURER, Mirian. An
evaluation of the methods countries to compute their official house prices indices.
HILL, Robert. Constructing price indexes across space and time: the case of the European
Union. American Economic Review, v.04, n. 5, p. 1379-1410, 2004.
HILL, Robert Price indexes for residential housing: a survey, evaluation and taxonomy.
Journal of Economic Surveys, v.27, n.1, p. 879-914, 2013;
NADALIN, Vanessa and FURTADO, Bernardo. Índices de preços para imóveis: uma
revisão. Boletim Regional, Urbano e Ambiental, n.6., p.21-32, 2011.
PAIXÃO, Luiz. Índice de preços hedônicos para apartamentos: Belo Horizonte, 1995-
2012. Tese (Doutorado em Economia). Programa de Pós-Graduação em Economia.
Universidade Federal do Rio de Janeiro, 2015.
PAIXÃO, Luiz. Índice de preços hedônicos para imóveis: uma análise para o município de
Belo Horizonte. Economia Aplicada, v.19, n.1, p.5-29, 2015a.
ROSEN, Sherwin. Hedonic price and implicit markets: products differentiation in pure
competition. Journal of Political Economy, v.82, n.1, p.35-55, 1974.
ROZENBAUM, Sérgio and MACEDO-SOARES, T. Diana. Proposta para a construção de
um índice de preços de imóveis a partir dos lançamentos imobiliários de condomínios
residenciais. Revista de Administração Pública, v.41, n.6, 2007.
ROZENBAUM, Sérgio. IMPA – Índice municipal de preço de apartamentos: proposta e
metodologia. Tese (Doutorado em Administração de Empresas). Programa de Pós-
Graduação em Administração de Empresas. Pontifícia Universidade Católica do Rio de
Janeiro, 2009.
SANTOS, Flávia and SALAZAR, Marlon. Índice de preços imobiliários para o Brasil:
estudos para discussão. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística (IBGE),
2011.
SIMÕES, Paulo Fernando Mahaz. Modelos longitudinais de efeitos mistos: aplicações ao
cálculo de índices de preços de imóveis e à análise da perda não técnica de eletricidade. Tese
(Doutorado em Engenharia Elétrica). Programa de Pós-Graduação em Engenharia Elétrica.
Pontifícia Universidade Católica do Rio de Janeiro, 2017.
TRIPLETT, Jack. Handbook on hedonic and quality adjustments in price indexes:
special application to information technology products. Paris: Organization for Economic
Co-operation and Development (OECDE), 2004. (Working Paper 2004/9)
WAUGH, Frederik. Quality factors influencing vegetable prices. Journal of Farm
Economics, v.10, n.2, p.185-196, 1928.
Appendix
Figure A1 – Belo Horizonte’s Regional and UP map
Source: TELEMAR
