scispace - formally typeset
Open AccessJournal ArticleDOI

A fuzzy knowledge-based framework for risk assessment of residential real estate investments

Reads0
Chats0
TLDR
In this article, a conceptual reference model for risk assessment of residential real estate using fuzzy cognitive mapping is developed, which allows cause and effect relationships between determinants to be identified and better understood, which in turn allows for better informed investment decisions.
Abstract
Risk analysis of residential real estate investments requires careful analysis of certain variables (or determinants). Because real estate is a key sector for economic and social development, this risk analysis is seen as critical in supporting decision processes relating to buying or selling residential properties, partly due to the pressures caused by the current economic environment. This study aims to develop a conceptual reference model for risk assessment of residential real estate using fuzzy cognitive mapping. This fuzzy model allows cause-and-effect relationships between determinants to be identified and better understood, which in turn allows for better informed investment decisions. The results show that the use of cognitive maps reduces the number of omitted criteria and favors learning with regard to how the criteria relate to each other, holding great potential and versatility in structuring complex decision problems. Practical implications, strengths and weaknesses of our proposal a...

read more

Content maybe subject to copyright    Report

Copyright © 2016 Vilnius Gediminas Technical University (VGTU) Press
http://www.tandfonline.com/TTED
TECHNOLOGICAL AND ECONOMIC DEVELOPMENT OF ECONOMY
ISSN 20294913 / eISSN 2029-4921
2017 Volume 23(1): 140–156
doi:10.3846/20294913.2016.1212742
Corresponding author Fernando A. F. Ferreira
E-mails: fernando.alberto.ferreira@iscte.pt; fernando.ferreira@memphis.edu
A FUZZY KNOWLEDGEBASED FRAMEWORK FOR RISK
ASSESSMENT OF RESIDENTIAL REAL ESTATE INVESTMENTS
Mónica I. F. RIBEIRO
a
, Fernando A. F. FERREIRA
b
,
Marjan S. JALALI
c
, Ieva MEIDUTĖ-KAVALIAUSKIENĖ
d
a
School of Management and Technology, Polytechnic Institute of Santarém, Complexo Andaluz,
Apartado 295, 2001-904 Santarém, Portugal
b
ISCTE Business School, BRU-IUL, University Institute of Lisbon, Avenida das Forças Armadas,
1649-026 Lisbon, Portugal
b
Fogelman College of Business and Economics, University of Memphis, Memphis, TN 38152-3120, USA
c
ISCTE Business School, BRU-IUL, University Institute of Lisbon,
Avenida das Forças Armadas, 1649-026 Lisbon, Portugal
d
Faculty of Business Management, Vilnius Gediminas Technical University,
Saulėtekio al. 11, LT-10223 Vilnius, Lithuania
d
BRU-IUL, University Institute of Lisbon, Avenida das Forças Armadas,
1649-026 Lisbon, Portugal
Received 07 October 2014; accepted 03 November 2015
Abstract. Risk analysis of residential real estate investments requires careful analysis of certain vari-
ables (or determinants). Because real estate is a key sector for economic and social development, this
risk analysis is seen as critical in supporting decision processes relating to buying or selling residen-
tial properties, partly due to the pressures caused by the current economic environment. is study
aims to develop a conceptual reference model for risk assessment of residential real estate using
fuzzy cognitive mapping. is fuzzy model allows cause-and-eect relationships between determi-
nants to be identied and better understood, which in turn allows for better informed investment
decisions. e results show that the use of cognitive maps reduces the number of omitted criteria
and favors learning with regard to how the criteria relate to each other, holding great potential and
versatility in structuring complex decision problems. Practical implications, strengths and weak-
nesses of our proposal are discussed.
Keywords: decision making, risk analysis of real estate investments, residential real estate, fuzzy
cognitive maps.
JEL Classication: C44, C45, M10, R31.

Technological and Economic Development of Economy, 2017, 23(1): 140–156
141
Introduction
Risk analysis of real estate investments is simultaneously one of the most important and
undervalued areas of nance. e uncertainty of the current economic climate, however,
motivated initially by the subprime crisis and, more recently, by the sovereign debt crisis,
has been asserting itself as a determining factor in how the nancial sector analyzes the
housing segment. Changes in the real estate market have a signicant impact on other sec
-
tors of economic activity and, therefore, on the well-being of the society at large. Ebru and
Eban (2009), Kauko (2010), Rybak and Shapoval (2011) and Warren (2011), among oth
-
ers, argue that the real estate market, namely the housing segment, is crucial for economic
development. is idea is further reinforced by Syz etal. (2008) and Rybak and Shapoval
(2011), who point to the relevance of this market for national wealth. Yet despite the grow
-
ing importance that has been given to risk analysis of residential real estate investments,
this eld of research is still relatively unexplored, an issue that is reinforced by the eects
of the current global economic crisis. As a result, there is considerable scope for new meth
-
odological approaches to support decision-making processes and allow for better informed,
more transparent and robust investment decisions.
Starting from the premise that the use of fuzzy cognitive mapping techniques fosters an
understanding of how the determinants of investment risk relate to each other, this study
aims to contribute to the development of a decision-making framework for risk analysis of
residential real estate investments. Specically, by constructing a fuzzy cognitive map (FCM),
our framework aims to: (1) identify the determinants of the risk of investment in residential
real estate; (2) contribute to reduce the number of omitted criteria in the decision making
process; and (3) increase our understanding of how the determinants of risk analysis in the
context of this study relate to each other. In this sense, it is worth noting that, according to
Carlucci etal. (2013: 208), “FCM is a well-established articial intelligence technique, incor-
porating ideas from articial neural networks and fuzzy logic, which can be eectively applied
in the domain of management science”.
e remainder of this paper is structured as follows. e next section provides the liter-
ature review on risk assessment of residential real estate investments. e ensuing section
presents the methodological background, justifying the use of FCMs in the context of this
study. e following section describes the process followed for the construction of our FCM,
and discusses the major advantages and shortcomings of our methodological proposal. e
last section presents concluding remarks and some lines for future research.
1. Risk assessment of real estate investments and related work
Owning a house is the largest single investment most households will make; and at the
same time, in most cases, it is limited by severe budget restrictions. In such circumstances,
bank loans become the most common and, perhaps, easiest practice for house acquisition
(Ferreira etal. 2013a). From an economic development standpoint, it is worth noting that,
due to the current economic instability (which aects the real estate market, reducing the
purchasing power of households and, consequently, motivating falls in private consump
-

142
M. I. F. Ribeiro et al. A fuzzy knowledge-based framework for risk assessment of residential ...
tion, housing included), nancial institutions have become more demanding in approving
loan applications and, as a result, have reduced lending concessions by imposing higher
credit underwriting standards to compensate the risk they face (cf. Ferreira etal. 2013a).
In practice, the real estate market has presented a generalized oversupply over the last
two decades (cf. Catalão 2010) and, due to the current economic climate, it is particularly
directed at those who have capital available to invest. For those who do not, bank loans have
been the solution. us, it seems clear that nancial institutions help support investment in
this sector, and that without the support of banks it would not be possible for families to
invest in real estate. It is in this sense that bank loans stimulate the economy of a country; i.e.
they encourage property acquisition by householders, strengthening the construction sector
and, consequently, increasing employment, money circulation and GDP growth (for further
discussion, see Ferreira etal. 2013a).
Real estate investments are generally considered high risk. Not only do they typically in-
volve signicant amounts of money, but the investor is negotiating an asset that is expected to
be protable in the long term. is type of business is also risky for the nancial institutions
which fund the investment and, naturally, require a thorough evaluation of loan applications.
is, in turn, requires taking into account the house value, the amount required and the
duration of the lending contract. According to Tavares etal. (2009), real estate appraisal is
an activity that depends on many factors, and should be conducted by those who operate in
the eld and possess a broad range of knowledge about prices, construction costs, urbanism,
supply-demand behavior, as well as about market trends and uctuations.
Given the increased risk of default, risk analysis is thus paramount for the housing mar-
ket. It is worth noting, however, that emerging markets’ real estate performance is nowadays
heavily aected by lack of investor condence, risk perceptions, increasing cost of nance and
nally market fundamentals (Onofrei, Anghel 2012: 481). In this sense, risk evaluation of real
estate investments needs to be as complete as possible. As pointed out by Yancang and Juan-
juan (2009: 138), “the risk evaluation of the real estate is more and more important. But, how
to nd an eective method to determine the weight of every risk factor and how to deal with the
uncertainty of the evaluation are urgent questions. Lots of eorts have been done. But, we still
have a long way to go. is premise is further supported by Wenpo and Minli (2012: 1815),
who argue that “real estate investment is a high-risk […] activity, the key of real estate analysis
is the identication of their types of investment risk and the risk of dierent types of eective
prevention. is means that it is important to identify both the risks associated with invest-
ing in real estate but also to achieve eective solutions that allow appraisals to be improved.
Following this, and according to Doumpos and Zopounidis (2001: 98), “while several multi-
variate statistical and econometric analysis techniques (e.g. discriminant analysis, logit and pro-
bit analysis, the linear probability model, etc.) have been used to address this type of problems,
their methodological shortcomings have already led researchers towards the exploitation of new
operational approaches. Indeed, as noticed by Šušteršic etal. (2009: 4736), the classic para-
metric approaches (e.g. linear discriminant analysis, linear regression, logit, probit, tobit and
binary tree) are reported to have a lack of accuracy [in this eld]. In addition, the current ap-
proaches for residential real estate risk evaluation are usually limited by: (1) lack of necessary
data (Lopez, Saidenberg 2000); (2) lack of rationality in the way trade-os between criteria

Technological and Economic Development of Economy, 2017, 23(1): 140–156
143
are calculated (Ferreira etal. 2012); and (3) the need to make subjectivity explicit in the deci-
sion making process (Santos etal. 2002) (for further discussion, see also Wang etal. 2011).
In light of these limitations, there seems to be considerable scope to explore the applica-
bility of fuzzy cognitive mapping techniques in the context of risk analysis of residential real
estate investments. Although several fuzzy risk evaluation models for real estate investments
exist in the literature (e.g. Wenpo, Minli 2012), it is worth noting that cognitive mapping not
only enables a large number of determinants to be identied, but it increases transparency
in the sense that it is clear where the data is coming from (Ackermann, Eden 2001; Ferreira,
Jalali 2015). In addition, with FCMs, the relative importance of the criteria is calculated
according to the experts’ own perceptions of that importance and aer discussion and ne-
gotiation among the panel members, whereby subjectivity is not only made explicit, but it is
incorporated and turned into a strength of the process.
Because FCMs are grounded on the practical experience, technical skills and realism
brought by the decision makers, they can potentially be used by parties investing in real
estate. e proposal presented here is constructivist in nature (see Ferreira etal. 2014), and
oers a perspective of complementarity rather than substitution. e next section presents
the methodological background of our proposal and explores the applicability of the fuzzy
cognitive mapping approach in the context of this study.
2. Methodological background
Steiger, D. and Steiger, N. (2008: 313) defend that mental models are tacit, hypothetical
knowledge structures that integrate the ideas, practices, assumptions, beliefs, relationships,
insights, facts and misconceptions that together shape the way an individual views and in
-
teracts with reality”. Decision aids based on human cognition can thus be seen as an op-
portunity for problem structuring because, according to Keeney (1996), decision makers
usually think of decision situations as problems to be solved, not as opportunities to be taken
advantage of . Cognitive maps become useful in decision making processes, because they
can help identify opportunities of action, reduce errors and search for good solutions (cf.
Ferreira etal. 2012). In practice, these maps are tools for structuring complex problems;
and as such, they contribute to reduce the rate of omitted criteria, promote discussion, and
lead to increased learning among the actors involved in the decision making framework,
as a result of the exchange of ideas and experiences (cf. Tegarden, Sheetz 2003; Eden, Ack
-
ermann 2004; Jalali etal. 2016).
Carlucci etal. (2013) and Ferreira etal. (2016), among others, note that cognitive maps
have two main functions: (1) a descriptive function, i.e. they provide visual representations,
helping individuals to have a better perception of the problem at hand, thus facilitating its
resolution; and (2) a function of reection, in which the map is seen as a tool to support the
development of new ideas. In practical terms, a cognitive map consists of a network of ideas,
hierarchically structured and connected by arrows, whose direction indicates the cause-and-
eect relationship between criteria (cf. Eden 2004; Eden, Ackermann 2004). In addition,
the arrows can have positive (+) or negative (–) signs, depending on the type of cause-and-
eect relationship between the existing concepts (cf. Montibeller, Belton 2006; Ferreira etal.

144
M. I. F. Ribeiro et al. A fuzzy knowledge-based framework for risk assessment of residential ...
2012). In short, cognitive maps are presented as representations of the environment; provide
a snapshot of reality and allow the understanding of cause-and-eect relationships between
concepts or variables to be claried.
2.1. Fuzzy cognitive maps
e concept of FCM was introduced by Kosko (1986), who combined cognitive maps with
fuzzy logics. As pointed out by Carlucci etal. (2013: 212),Kosko enhanced the power of cog
-
nitive maps considering fuzzy values for the concepts of the cognitive map and fuzzy degrees of
interrelationships between concepts. Fuzzy logics was developed in the 1960s and has been
widely used to model social, economic and political problems (cf. Carvalho 2013). It is an
approach that holds great potential in dealing with investment decisions in residential real
estate, namely because it helps to understand and analyze the associated risk.
FCMs have two particular characteristics: (1) the cause-and-eect relationships W
ij
be-
tween concepts/criteria C
i
and C
j
follow a fuzzy logic; and (2) the system is dynamic, i.e. it
involves feedback links between criteria (Fig.1).
In addition to the graphical representation, FCMs have a mathematical basis. According
to Kok (2009), Mazlack (2009), Carlucci etal. (2013) and Ferreira and Jalali (2015), there is
a state vector n × 1, which includes the value of n concepts; and a n × n weight matrix W
(also known as adjacent matrix), that gets together all the weights W
ij
and the relationships
between the n criteria. Accordingly, the value of each concept is inuenced by the values
of its interconnected concepts and by its own previous value. is can be represented by
formulation (1), where A
i
(t+1)
is the activation level of concept Ci at time t+1; f stands for a
threshold activation function; A
i
(t)
represents the activation level of concept C
i
at time t; A
j
(t)
is the activation level of concept C
j
at time t; and W
ji
is the weight dened for the relation-
ship between both concepts:
( ) ( ) ( )
+
=

= +




1
1
.
n
t tt
ji
ii
j
ji
j
A fA A W
. (1)
Fig.1. Typical structure of an FCM
Source: Salmeron (2012: 3706).

Citations
More filters
Journal ArticleDOI

A Hybrid MCDM Technique for Risk Management in Construction Projects

TL;DR: The analytical network process (ANP) methodology in the D numbers domain is extended to handle three types of ambiguous information’s, viz. complete, uncertain, and incomplete, and assesses the weight of risk criteria.
Journal ArticleDOI

Combining cognitive mapping and MCDA for improving quality of life in urban areas

TL;DR: This study sought to overcome some of the more common limitations of current evaluation methods, by combining cognitive mapping and the measuring attractiveness by a categorical based evaluation technique (MACBETH), to develop a more informed, coherent and transparent evaluation system.
Journal ArticleDOI

Developing a green city assessment system using cognitive maps and the Choquet Integral

TL;DR: This study sought to develop a multiple-criteria model that facilitates the evaluation of green cities’ sustainability, based on cognitive mapping techniques and the Choquet integral (CI), and validated both by the panel members and a parliamentary representative of the Portuguese ecology party “Os Verdes” (The Greens), who confirmed that the evaluation system created distinguishes between cities according to how strongly they adhere to “green” principles.
Journal ArticleDOI

Exploring the determinants of digital entrepreneurship using fuzzy cognitive maps

TL;DR: A fuzzy cognitive map was developed to identify and analyze the determinants of digital entrepreneurship, resulting in a well-informed framework that was validated by the panel members and enabled an objective analysis of the dynamics behind digital entrepreneurship.
Journal ArticleDOI

A multiple criteria group decision-making approach for the assessment of small and medium-sized enterprise competitiveness

TL;DR: In this article, the authors developed a method for assessing SMEs' competitiveness based on a constructivist epistemology, which makes an integrated use of cognitive mapping and the measuring attractiveness by a categorical-based evaluation technique (MACBETH).
References
More filters
Journal ArticleDOI

Fuzzy cognitive maps

TL;DR: A fuzzy causal algebra for governing causal propagation on FCMs is developed and it allows knowledge bases to be grown by connecting different FCMs.
Book

Multiple Criteria Decision Analysis: An Integrated Approach

TL;DR: This book is not only a theoretical document, but also provides good coverage of practical issues and can be recommended to a broad audience, ranging from those in academic institutions to practitioners, as well as those who are interested in finding information on multiple criteria approaches, methods and techniques.
Book

Rational analysis for a problematic world revisited : problem structuring methods for complexity, uncertainty and conflict

TL;DR: In this paper, Rosenhead and Mingers present a new Paradigm of Analysis, called SODA, which is a new paradigm of analysis for analysis in the context of M&A.
Journal ArticleDOI

Ecological models based on people’s knowledge: a multi-step fuzzy cognitive mapping approach

TL;DR: In this paper, a multi-step fuzzy cognitive mapping approach is proposed to create ecological models with both expert and local people's knowledge, which can be made of almost any system or problem.