Volume 6 No 1 (2016)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2016.92  |   http://emaj.pitt.edu 

 
 
 

 

 

 

 

Investment Portfolios in an Emerging Economy: What Drives Portfolio’s 

Diversification? 
 
Pedro Luiz Albertin Bono Milan 
EAESP-FGV, GVCef: Finance Research Center | bonomilan@gmail.com  
 
William Eid Jr. 
EAESP-FGV, GVCef: Finance Research Center | william.eid@fgv.br 
 

 

 

Abstract 
 
This study sheds light on the investment portfolio’s decisions through behavioral insights. The study intends to identify 

personal characteristics that drive the level of diversification and lead investors to allocate resources in risky assets in an 

emergent economy, deepening the discussion about investment decisions and bringing some behavioral insights to the 

debate. The study has a unique and heterogeneous database of individual financial allocations from Brazil, one of the 

largest emergent economies. The characteristics of Brazilian investors play an important role in investment decisions, high 

educated and married investors tend to display diversified portfolios. To invest in risky assets, male investors have a 43% 

greater likelihood of investing in risky assets than females, highlighting the discussion on gender and investment decisions.  

Moreover, married investors tend to exhibit conservative portfolios. We observed that traditional investors are under-

diversified, allocating primarily in traditional and safety assets. The results suggest that the investment decisions can be 

subject to psychological biases defined in behavioral finance theory. 

 
 
 
 
 
 
 
Keywords: Investment Portfolios, Diversification, Risky Assets, Behavioral Finance. 
 

 

New articles in this journal are licensed under a Creative Commons Attribution 3.0 United States License. 

 

This journal is published by the University Library System of the University of Pittsburgh as part  
of its D-Scribe Digital Publishing Program, and is cosponsored by the University of Pittsburgh Press. 

  

Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120   |  http://emaj.pitt.edu |    
 
 
 

mailto:bonomilan@gmail.com
http://www.library.pitt.edu/
http://www.pitt.edu/
http://www.library.pitt.edu/articles/digpubtype/index.html
http://www.upress.pitt.edu/upressIndex.aspx
http://creativecommons.org/licenses/by/3.0/us/


Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120 |   http://emaj.pitt.edu 

 
 
 
 

 
Investment Portfolios in an Emerging Economy: What Drives Portfolio’s Diversification? 
 

Page |1| Emerging Markets Journal 
 
 

Investment Portfolios in an 
Emerging Economy: What 
Drives Portfolio’s 

Diversification? 

 
Pedro Luiz Albertin Bono Milan 
William Eid Jr. 
 
1. Introduction 
 

The growth of the financial market in recent 
years has led to greater availability of financial products 
and services in order to meet a new range of customers 
and investors. Among the new products, those intended 
for financial allocation require special attention because 
they offer new diversification possibilities and financial 
strategies for investors. According to Modern Theory of 
Finance, which has been built since the 1970s, when 
several investment opportunities are available in the 
financial market, investors diversify their investments 
mitigating risks and maximizing rates of return 
(Markowitz, 1952). 

  
Modern Theory of Finance analyzes the 

investors’ decisions based on assumptions, among them, 

the market efficiency and individual rationality, where 
investor decisions are perfectly rational based on the 
correct analysis of available information (Fama, 1970; 
Shiller, 1999). However, it is noteworthy that the vast 
majority of investors do not have diversified portfolios, 
and many invest without regard to the risk-return 
relationship. Noting the dissonance between the practice 
and the Modern Theory of Finance, some studies have 
begun to question investors’ rationality in presenting 

psychological factors affecting financial decisions and, 
furthermore, the emotional factors leading investor 
decisions. 

  
A pioneering study from 1979 states that 

financial decisions based on emotions or psychological 
influences can lead investors to allocate their financial 
resources in higher-risk investments and under-
diversified portfolios (Kahneman and Tversky, 1979). 
The discussion of the rationality of investors in the 
finance literature leads to the concept of Behavioral 
Finance, which analyzes investors’ decisions under 

psychology and finance theories highlighting the factors 
that drive financial decisions. 

The Brazilian financial market has a crescent 
number of investors and several investment possibilities, 
offering conservative investments ranging from savings 
and fixed income funds to risky investments via capital 
markets and structured products. Through the prism of 
the Modern Theory of Finance, Brazilian investors 
should allocate their financial resources in a rational and 
diversified way. On the other hand, according to 
Behavioral Finance Theory, Brazilian investors can make 
decisions under psychological bias when choosing under-
diversified or risky portfolios. 

 
Few empirical studies analyze Brazilians’ investment 

decisions and the intrinsic characteristics that lead to 
portfolio diversification and the investment in risky 
assets. Considering the size and sophistication of the 
financial market in Brazil, it is important to analyze what 
drives investment decisions to provide new subsidies to 
improve the financial education process of Brazilians and 
provide new tools for market participants to develop 
better products for investors. 
 

In this context, the main objective of our study 
is to analyze how Brazilian investors are investing their 
financial resources among several possibilities available 
in the market and to seek the factors that can influence 
the level of portfolio diversification and the financial 
allocation in risky assets. The paper is organized as 
follows: The first presents the motivation. The second 
part debates the underlying literature of financial theories 
and behavioral influences. The third presents the data and 
the methodology applied. The fourth discusses the 
results. The fifth part of the study is devoted to the 
conclusion. 
 
2. Underlying Theories 
 

The Modern Theory of Finance states that 
investors act rationally, seeking to minimize risks and 
maximize returns through the process of the 
diversification of their investments. The theory departs 
from a neoclassical microeconomic approach whose 
central paradigm is the rationality of economic agents 
(Yoshinaga et al., 2008). In this sense, individuals who 
operate in the financial markets must have the ability to 
process the information available and to make rational 
decisions consistent with the concept of Expected Utility 
(Von Neumann, 1947). 

  
However, over the course of thirty years, many 

studies have presented evidence of investors allocating 
financial resources in under-diversified portfolios and 
unbalanced portfolios concentrated within an average of 



Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120 |   http://emaj.pitt.edu 

 

 

Pedro Luiz Albertin, Bono Milan, William Eid Jr. 
Emerging Markets Journal | P a g e  |2 

two or three assets (Blume and Friend, 1978) (Barber and 
Odean 2000) (Polkovnichenko, 2005; Goetzmann and 
Kumar, 2008). Thus, a new theoretical approach to 
finance goes considers the individual's behavior in the 
investment decision. In this context, the Behavioral 
Finance Theory has the aim of improving the modern 
theory of finance through psychological and behavioral 
approaches (Lintner, 1998) (Kimura, Basso and Krauter, 
2006). Moreover, cultural differences shape the 
psychological aspects and play an important role in 
making economic and financial decisions. Levinson and 
Peng (2007) suggest that a systematic difference affects 
financial and economic decisions, whether rational or 
irrational, often assumed to be universal. The authors 
conducted an empirical study in the United States and 
China to examine how cultural background affects 
economic decision-making and test the framework, 
information morality and outside groups influence the 
judgments of financial value and the property across 
cultures. 

 
Important studies that have incorporated 

Behavioral Finance Theory present investor decisions 
driven by psychological needs, following Maslow's 
Hierarchy of Needs. The Hierarchy of Needs is a concept 
of motivational behavior, one of the most important 
motivation theories in psychology studies. In this 
concept, human actions and decisions are explained by 
five levels of human needs, ranging from basic 
physiological needs to the highest level of needs: self-
actualization (Maslow, 1943), as presented in Figure 1. 
At each level of the Hierarchy of Needs, the human being 
receives internal and external psychological influences 
that affect their decisions, adding the set of needs that 
man seeks to meet throughout life.  
 

 
Figure 1. Maslow's Hierarchy of Needs 
Source: Hierarchy of Needs, Maslow (1943). Prepared by 
the authors. 
 

In the context of investments arises the concept 
of Behavioral Portfolio Theory, which is visually 
presented in Figure 2, connecting Maslow's Hierarchy of 
Needs with traditional financial theories (Shefrin and 
Statman, 2000), (De Brouwer, 2009).   
 

 
Figure 2. Behavioral Portfolio Theory 
Source: Prepared by the authors. 
 

The connection between the theory of 
behavioral finance and the traditional theory of finance 
results in three approaches that investors can use in the 
investment decision-making process (Shefrin e Statman, 
2000). The first approach is the Safety-first Portfolio 
Theory, whereby investors essentially receive the 
influence of the second level of Maslow's hierarchy. In 
the safety-first approach, the investor aims to minimize 
the likelihood of financial distress Pr(W<s), where the 
investor aims to maintain its level of wealth (W) above 
the subsistence level (s), i.e., above the first level 
Maslow's Hierarchy of Needs. Investors concentrate their 
investments in under-diversified portfolios allocated 
primarily on traditional and low-risk assets. 
  

The second is the PS/A (Theory, Security, 
Potential and Aspiration Theory). The second theory 
follows the concepts of Safety-first theory, adding the 
intrinsic potential of the investor. The potential is the 
individual aspiration to achieve the highest level of 
Maslow's Hierarchy of Needs. Security is the major 
guide of financial decisions, but the investor receives an 
input from the need to self-actualize. This potential and 
aspiration lead the investor to create a target point of 
wealth, a desire for higher levels of wealth. The central 
difference in this second approach is in the interpretation 
of the variables. Specifically, the subsistence level (s) 
replaced by a more general aspiration level (A). Based on 
this approach, the investor can allocate portions of its 
investments in risky assets by seeking higher returns. 
Even though an investor has a potential aspiration for 
wealth, it keeps a major fraction of its wealth in 
traditional and low-risk assets. 

 
The third and final approach, BPT-AS (Theory, 

Security, Potential and Aspiration Theory + Prospect 
Theory), complements the two previous approaches to 
Prospect Theory (Kahneman and Tversky, 1979). The 
Investor’s motivation is the security instincts, but invest 

in different assets by seeking higher rates of return and 
higher levels of wealth. The difference in this approach is 
the addition of all cognitive biases in financial decision-
making. Cognitive biases in investment interfere in 
mental processes, generating decision biases that may 
violate the axioms of rationality and leading investors to 
allocate their resources in investments that generate 
lower rates of return or assume higher levels of risk. 
Among the biases are: loss aversion, anchoring, 



Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120 |   http://emaj.pitt.edu 

 
 

 
Investment Portfolios in an Emerging Economy: What Drives Portfolio’s Diversification? 
 

Page |3| Emerging Markets Journal 

availability, herd behavior, local trends and 
overconfidence (Kahneman and Tversky, 1979), (De 
Bondt e Thaler, 1994). 
 

Loss aversion may be present only when the 
investment has improved, imposing a desire to preserve 
the achieved return. In investments facing losses, 
investors tend to take greater risks because they believe 
they can recover past losses by relocating in riskier assets 
(Köbberling and Walker, 2005). Psychological biases can 

affect self-control, misleading their comprehension about 
their actions.  
 

Anchoring is a behavioral bias that biases 
decisions toward initial values, i.e., a limitation in the 
decision-making process, leaving the investor tied to a 
past target return, the result of a payoff from a past time 
(Jacowitz and Kahneman, 1995). Furthermore, the 
individual may invest based on available information, 
called availability bias, whereas investors do not examine 
other alternatives or additional information to allocate 
financial resources. Individuals evaluate the chance that 
some event will happen from the comparable events that 
have occurred previously (Barberis and Thaler, 2003).  
 

Herd Behavior drives investors to replicate the 
decisions made by other investors. The belief is that other 
investors invest with more information; thus, when an 
individual makes a decision based on what the other 
investor did, he undervalues the information available in 
the market (Bikhchandani and Sharma, 2000). Local 
knowledge can also be present as an investment decision 
bias if the investor chooses to allocate its resources only 
in markets in which it already has a level of information 
and greater familiarity (Coval and Moskowitz, 1999). 
 

The overconfidence bias could lead investors to 
believe that they have a comparative advantage in the 
analysis and valuation of assets relative to the market, 
holding losing positions (Plous, 1993). Few empirical 
studies seek to understand the behavior of Brazilian 
investors in light of Behavioral Finance. The discussion 
in the literature is intriguing and stimulates research to 
understand whether Brazilian investors allocate their 
financial resources following the model of rationality or 
whether they decide by psychological factors. 

 
In the United States, several studies have 

analyzed the portfolio diversification and investments in 
risky assets of individual investors. The theories of 
diversification do not appear to describe investors’ 

decisions properly, given that U.S. investors, analyzed 
based on the Survey of Consumer Finances 1983, have 
under-diversified portfolios (Kelly, 1995). As a possible 
explanation for investor behavior, the overconfidence 
bias can affect investment decisions and lead investors to 
less-diversified portfolios and a greater exposure to risky 

assets. Kelly (1985) tests the effect of various investors’ 

characteristics in the level of portfolio diversification, 
among them: age, education level, occupation, and risk 
profiles. However, none of the personal characteristics is 
significant to improve the understanding of investors’ 

decisions. 
 
A complementary study using data from the 

Survey of Consumer Finances between 1992 and 2001 
seeks to find the determinants of equity investment in the 
U.S. market (Shum and Faig, 2006). Shum and Faig 
(2006) include investors’ characteristics to understand 

their financial allocation. Authors note a great 
heterogeneity in the financial allocation, identifying that 
the decision to hold equities in investors’ portfolios 

positively correlated with an investor’s total investment, 

age, risk attitude and reasons for saving, such as 
retirement. Investors aged up to 50 years old tend to have 
more equities in their portfolios. As investors surpass the 
age of 50, they reverse this tendency, leading to a 
reduction in the number of equities. In addition, the total 
investments and savings have positive effects on equities 
held in a portfolio. 

 
A study conducted in 2008 analyzes American 

investor behavior and observes poorly diversified 
investment portfolios (Goetzmann and Kumar, 2008). 
The authors note that a low level of diversification is 
associated with young investors with lower income, less 
education and less sophistication. From the perspective 
of Behavioral Finance, Goetzmann and Kumar (2008) 
state that investment decisions are consistent with 
overconfidence bias, trend behavior and local bias. 

 
Following, Grable et. al. (2004) attempt to 

explain why individuals exhibit behaviors that do not 
maximize the expected utility. The study explores how 
the projection behavioral bias can shape financial risk 
tolerance attitudes. The results suggest that changes of 
gender, income, and stock market prices, analyzed by 
market indices, help to explain the risk attitudes. 
Tolerance appears to be risk elastic and attitude changing 
over time. International studies provide a strong base of 
empirical support for our testing and allow us to compare 
the investment portfolios of Brazilians, their level of 
diversification and investment in risky assets. 
 
3. Data and Methods 

 
We use a novel and heterogeneous database 

composed of 500 Brazilian investors and 2,224 
investment observations to attend our empirical 
investigation. Our database provides sufficient details to 
deepen the discussion of the investment decisions of 
Brazilians. We construct our database based on the 
Brazilian Supreme Electoral Court (TSE-Tribunal 
Superior Eleitoral) website. The TSE website maintains 



Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120 |   http://emaj.pitt.edu 

 

 

Pedro Luiz Albertin, Bono Milan, William Eid Jr. 
Emerging Markets Journal | P a g e  |4 

the investment portfolios of candidates for the political 
positions of mayors and councilors in the year 2012 for 
all Brazilian states. 

 
Investors are Brazilians with heterogeneous 

occupations not based on political occupation. By opting 
for the candidacy for a political position in Brazil, the 
candidate must release all of its assets and investments in 
the Supreme Electoral Tribunal website. While we 
understand that there may be a selection bias in the study 
sample, to our knowledge, it is the only database publicly 
available. Moreover, we believe that the impact is 
minimum by the sample size, the heterogeneity in the 
occupations of these people and the differences in their 
investment portfolios, serving as a good proxy for the 
investment portfolios of Brazilians. 

 
Our database reports detailed allocations in 

different assets with at least 17,000 Reais and at most of 
11 million Reais of investments, and we account for 
personal characteristics such as age, gender, marital 
status, education level, residence in the capital city or 
countryside and total investment. The total investment of 
each investor is in natural logarithm to capture the 
variation between investment portfolios without 
distorting the results. 

 
The average age of investors, presented in 

Table 1, is 53 years old, with a minimum of 30 years and 
a maximum of 87 years old. These investors allocate 
their resources in four different types of assets on 
average, with a minimum investment of 1 and a 
maximum of 13 different assets. Observing the number 
of investments, we can determine that the average 
Brazilian investor is under-diversified. Among all 
investors, 435 are men and 75 are women, 344 are 
married, 391 have higher education, and 56% of them 
reside in capital cities. 

 
Table 1: Summary Statistics of investors’ 

characteristics 

 
Age 

(years) 
Total Investment 

(R$) 
Number of 
investments 

Mean 53 731,669.55 4.45 
Median 54 462,067.64 4 

Std. Dev. 11 988,523.13 1.76 
Min 30 17,577.52 1 
Max 87 11,186,074.14 13 

N 500 500 500 

 
N % 

 
Man 425 85% 

 
Woman 75 15% 

 
Married 344 69% 

 
Not Married 156 31% 

 
Higher Education 391 78% 

 
No Higher Educ. 109 22% 

 
Residence in 

Capital 
282 56% 

 
Residence in 
Countryside 

218 44% 
 

Source: TSE – Tribunal Superior Eleitoral 2012. 
Prepared by the authors. 

Table 2 shows the allocation of Brazilian 
investors in different assets. Investors allocate most of 
their investments in residential real estate, followed by 
vehicles and urban land. Table 2 shows the under-
diversified portfolios and a conservative allocation with 
approximately 70% allocated to traditional and low-risk 
assets. It is important to highlight the resources allocated 
in vehicles, durables that traditionally do not generate 
positive financial returns. Financial investments represent 
approximately 15% of the total polarized in fixed income 
assets, reinforcing the evidence of conservative investor 
profiles.  
 

Table 2: Asset Allocation (% Participation) 

 
 
Source: TSE – Tribunal Superior Eleitoral 2012. 
Prepared by the authors. 
 
Figure 3 graphically presents the main investments made 
by investors. 

 
Figure 3. Asset Allocation (% Participation) 

 
Source: TSE – Tribunal Superior Eleitoral 2012. 
Prepared by the authors. 

The black vertical bars represent non-financial 
investments; the gray vertical bars represent financial 
investments. Stratifying the sample by total investments, 
we observe that there is a thin shift in asset allocation. As 
total investments increase, allocations in residential real 
estate and vehicles decline, and allocations in rural 
properties and entrepreneurial and financial allocations 
increase. However, investor portfolios remain under-
diversified.  



Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120 |   http://emaj.pitt.edu 

 
 

 
Investment Portfolios in an Emerging Economy: What Drives Portfolio’s Diversification? 
 

Page |5| Emerging Markets Journal 

 
Table 3: Allocation of the main investments by total 

investment (in R$) 

 
Source: TSE – Tribunal Superior Eleitoral 2012. 
Prepared by the authors. 
 
Figure 4 graphically presents the changes in asset 
allocations as total investments increase. 
 

 
 

Figure 4. Changes in asset allocations as total 
investments increase 

 
Source: TSE – Tribunal Superior Eleitoral 2012. 
Prepared by the authors. 
 
3.1. Diversification Metrics 

 
To estimate the level of portfolio 

diversification, we follow the diversification metric 
presented by Blume and Friend, 1978. The metric 
analyzes investor allocations compared to a market 
portfolio. Blume and Friend, 1978 analyze the stock 
portfolios of U.S. investors using the capital market 
index as a benchmark to measure the diversification 
index. Whereas Brazilian investors allocate little 
financial resources in stocks and the absence of a market 
benchmark for these investors, we adjust the metric 
creating a market portfolio based on the mean and 
median allocations of all investors in our database. 

 
Our adaptation analyzes the deviation of the 

individual allocation of each asset compared to the mean 
and median allocations of each asset among all investors 

in the sample. Thus, assuming that the mean and median 
portfolios are a good representation of the portfolio 
allocation of Brazilians, the mean and median of all 
investments in each asset are the benchmark in our study. 
We understand that our adaptation can indicate a 
specification error because there is not a benchmark 
publicly available; however, examining how individual 
investors differ from the mean and median allocation of 
all investors can be a good starting point to understand 
the factors leading Brazilian investors to diversify their 
portfolios. According to the metric, the lower the index 
measured by Equations 1 and 2, the greater the portfolio 
diversification. 
 
  Thus, the level of portfolio diversification 
follows Equation (1) and Equation (2): 
 
 
 
 
 
 
 
 
 
 
 
Where: 
 
DImeani is the diversification index using the mean of 
investor i 
DImediani is the diversification index using the median 
of investor i 
N is the number of securities held by the investor 
wi is the portfolio weight assigned to investment type i in 
the investor portfolio 
wbenchmeani is the mean weight assigned to an 
investment type i in the benchmark portfolio 
wbenchmedianI is the median weight assigned to an 
investment type i in the benchmark portfolio 
 
3.2. Metrics of Investment in Risky Assets 

 
Whereas Brazilian investors have concentrated 

portfolios in low-risk assets, we seek to understand what 
drives investors to take a divergent allocation. We use 
higher-risk investments and p personal characteristics to 
pursue factors that can lead investors to allocate financial 
resources in different assets. Stocks, equity mutual funds 
and entrepreneurial investments are risky investments in 
our study.  
 
We measure the level of allocation in risky assets via two 
procedures: 
 

The first procedure considers the weighted sum 
of risky assets contained in each individual portfolio. The 

                          
 

   

 

(1) 

         

                     
 

   

 

(2) 



Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120 |   http://emaj.pitt.edu 

 

 

Pedro Luiz Albertin, Bono Milan, William Eid Jr. 
Emerging Markets Journal | P a g e  |6 

formula allows us to observe the relative percentage of 
total investments allocated in risky assets. 
 
Equation (3) presents the formula for measuring the level 
of risk allocation: 
 

 
 
Where, 
 
Iriski is the risk-taking index of investor i 
Wriski is the sum of all risk investments of investor i 
Totali is the sum of all investments of investor i  
 

The second procedure to measure the allocation 
in risky assets follows Shum and Faig (2006), adding a 
dichotomous variable that takes the value of 1 if 
investors invest in risky assets or 0 if the investor has no 
risky assets in its investment portfolio.   

 
Finally, to find the factors that influence 

investors to adopt a more diversified portfolio, we 
developed an OLS regression model using the 
diversification index measured by Equation (1) and 
Equation (2) as explained variables. As explanatory 
variables, we use the investors’ characteristics: age (Ω), 

gender (λ), level of education (δ), residence (θ) (capital 

city or countryside), marital status (ξ) and total 

investment (Ψ). The objective of this regression model is 

to capture the intrinsic characteristics that lead to 
diversified allocations. The regressions relied on the 
application of a robust White Matrix.  
 
Equation (1) as an explained variable follow the OLS 
Equation 4: 
 
                                     

         
 
 
Equation (2) as an explained variable follow the OLS 
Equation 5: 
 
                                  

               
 
 

Similar to equations 4 and 5, to find the factors 
that influence investors to allocate resources in risky 
assets, we developed two regression models to verify the 
impact of age (Ω), gender (λ), level of education (δ), 

residence (θ) (capital city or countryside), marital status 

(ξ) and total investment (Ψ) on the decision to allocate 

financial resources in risky assets. The first regression is 
an OLS regression and has the explanatory variable 
measured by the first risk metric procedure, Equation 3. 
The second regression, conducted under the second risk 
metric procedure, has dichotomous explained variables; 
therefore, we adopted a Probit model. The regressions 
relied on the application of a robust White Matrix. 
 
Equation (3) as an explained variable follow the OLS 
Equation (6): 
 
                                    

          
 
 
Procedure (2) as a dichotomous explained variable 
follow the Probit Equation (7): 
 
                                     

              
 
 
4. Results 

 
Our results provide new insights about how 

Brazilian investors allocate their financial resources. The 
results indicate that investors have under-diversified 
portfolios concentrated mainly in traditional and low-risk 
assets. The average rate of diversification estimated by 
Equation 1 is 0.2153. The diversification index using the 
benchmark portfolio composed by the median 
allocations, Equation (2) presents an average 
diversification rate of approximately 0.5023. Despite the 
difference between the two models, we observe that the 
difference extends to all individual portfolios in 
proportional magnitude, preserving the comparative 
power between the models. Table 4 shows diversification 
rates in quartiles, and quartile 1 (Q1) indicates investors 
with higher levels of diversification. The diversification 
index becomes smaller as we move from the second 
quartile to the fourth quartile.  
 

There is a significant difference between 
investors with high diversification index, first quartile, 
and investors located in the fourth quartile for both 
metrics diversification. However, the differences must to 
be statistically significant to give consistency to the 
inferences. To test for significant differences between the 
investors with greater diversification rates and investors 
with smaller diversification rates, we conducted a t-test 
of means between the Q1 and Q4 groups. 
 
 
 
 
 

Risk 

Investment

Index

if risky assets ≠ 0            variable value = 1

if risky assets = 0            variable value = 0

      

   
      

      
 

 

   

 (3) 

(4) 

(5) 

(6) 

(7) 



Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120 |   http://emaj.pitt.edu 

 
 

 
Investment Portfolios in an Emerging Economy: What Drives Portfolio’s Diversification? 
 

Page |7| Emerging Markets Journal 

Table 4: Descriptive statistics of diversification rates 
  DImean Q1 Q2 Q3 Q4 

Mean 0.2153 0.0560 0.1103 0.2037 0.4911 
Median 0.1536 0.0582 0.1070 0.2019 0.4385 
Std. Dev. 0.1926 0.0157 0.0204 0.0315 0.1849 
Min 0.0192 0.0807 0.0807 0.1537 0.2719 
Max 0.9124 0.1534 0.1534 0.2709 0.9124 
N 500 125 125 125 125 

        DImedian Q1 Q2 Q3 Q4 

Mean 0.5023 0.2934 0.4193 0.5270 0.7697 
Median 0.4673 0.2992 0.4172 0.5227 0.7796 
Std. Dev. 0.1865 0.0531 0.0286 0.0370 0.1062 
Min 0.1584 0.1584 0.3695 0.4673 0.6113 
Max 10.000 0.3694 0.4672 0.6094 10.000 
N 500 125 125 125 125 

 
Source: Prepared by the authors 
 

The t-test compares the rates of portfolio 
diversification in the first quartile, which has higher 
diversification rates, with the lower diversification rates 
located in the fourth quartile. The test revealed a 
statistically significant difference at 1% with t-stat 
<0.000 for both models, ensuring that investors 
distributed in the first quartile had significantly different 
portfolios than investors located in the fourth quartile. 
This t-test validates the inferences for each group in the 
regression models.  
 

The regression models intend to identify 
aspects that can lead investors to increase their 
diversification levels and characteristics that can drive 
investors to increase investments in risky assets in their 
portfolios.  Table 5 presents the results of the four 
regression models. Columns (1) and (2) identify the 
characteristics of investors that can lead to a more 
diversified portfolio. Columns (3) and (4) present the 
characteristics that lead investors to invest greater 
financial resources in risky assets. 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Table 5: Results of the four regression models 
 

 
 
Source: Prepared by the authors 
 
Notes for Table 6: Regression Table for models of 
diversification and investment in risky assets. (1)(2) OLS 
regression model with a diversification index measured 
by Equation 1 and Equation 2 as an explained variable 
and the variables Age (Ω), gender (λ), education level 

(δ), residence (θ) (capital city or countryside), marital 

status (ξ) and total investment (Ψ) as explanatory 

variables. (3)(4) Regression models to verify the impact 
of Age (Ω), gender (λ), education level (δ), residence (θ) 

(capital city or countryside), marital status (ξ) and total 

investment (Ψ) in risky assets. Regression (3) is an OLS 

regression has explained the variable measured by 
Equation 3. Regression (4) is a Probit model that 
analyzes the impact of the explanatory variables on the 
level of investment in risky assets. The regressions relied 
on the application of a robust White Matrix 
 

The results displayed in Table 5 indicate the 
effects of investor characteristics on the level of portfolio 
diversification as well as the propensity to invest in risky 
assets. The characteristics of Brazilian investors play an 
important role in investment decisions, and the results 
support the findings in international studies. Higher 
education influences investors to higher levels of 
portfolio diversification. The result has statistical 
significance and follows evidence found by Goetzmann 
and Kumar (2008) for American investors. Marital status 
also positively affects the level of portfolio 
diversification. Married investors tend to have more 
diversified portfolios than single investors do. 

 
Goetzmann and Kumar (2008) observe more 

diversified portfolios in the United States as total 
investment increases, contrary to our finding for Brazil. 
In Brazil, higher financial resources do not necessarily 
lead to more diversified portfolios. In the presence of 
new financial resources, the Brazilian investors maintain 



Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120 |   http://emaj.pitt.edu 

 

 

Pedro Luiz Albertin, Bono Milan, William Eid Jr. 
Emerging Markets Journal | P a g e  |8 

their resources in the same types of investments. If the 
investor allocates its resources in residential real estate, 
land and savings, he will not necessarily allocate its 
resources in other assets, and in the presence of new 
financial resources, an investor will allocate more 
resources in the same types of assets. 

 
Comparing the results concerning investment 

in risky assets, we observe that a greater number of 
personal characteristics affect the investment decisions of 
Brazilian investors. The equations in models 6 and 7 seek 
the characteristics of investors that lead them to invest in 
risky assets. For both models, shown in columns (3) and 
(4) of Table 6, age may indicate higher risk exposure. 
The evidence accompanying the results found in the 
study of Shum and Faig (2006) in the United States 
shows that younger investors tend to invest more in risky 
assets. 

 
The model also shows that male investors have 

a 43% greater likelihood of investing in risky assets than 
female investors, highlighting the discussion on gender 
and investment decisions. Married investors tend to 
exhibit less risk and maintain conservative portfolios. As 
expected, as total investments increase, the greater the 
investment in risky assets will be, once more in line with 
the results of Shum and Faig (2006). 

 
Following international studies, our results 

suggest that Brazilian investors in our dataset cannot 
follow the complete model of rationality assumed by the 
Modern Finance Theory. Evidence suggests that their 
investment decisions are subject the psychological 
interference defined in Behavioral Finance theory, 
leading to poorly diversified portfolios. In addition, 
behavioral biases can lead the investor to allocate 
resources in risky assets regardless of the risk factors. 

 
Analyzing the study by Shefrin and Statman 

(2000), our results suggest that Brazilian investors make 
financial decisions that are consistent with the BPT-AS 
Theory approach (Theory, Security, Potential and 
Aspiration Theory + Prospect Theory). Investors can 
move essentially by security instinct, but a potential 
aspiration to higher levels of wealth can lead the investor 
to invest a portion of its assets in risky assets, seeking 
higher rates of return. 

 
The results that present Brazilian investors as 

less diversified with a high concentration of traditional 
and low-risk assets in their portfolios and a fraction of 
their resources in risky assets support the possibility that 
these investors may be subject to psychological biases in 
financial decisions. Personal characteristics that lead 
investors to take greater risk suggest individual 
overconfidence, a behavioral bias that leads investors to 
believe that their ability to process information is better 
than that of other investors, leading to greater exposure to 

risk. Age can play a strong role to overconfidence, given 
that younger investors tend to take greater risks. 
Goetzmann and Kumar (2008) and Kelly (1985) indicate 
overconfidence bias as an important factor that affects 
financial decisions. 

 
Married and high-educated investors are more 

conservative and invest less in risky assets. The results 
reinforce the theory of safety first. In addition, under-
diversified portfolios that are concentrated in traditional 
assets may follow the herd effect, a psychological bias 
that leads investor to invest only in well-known assets 
and local assets.  

 
Although our results are in line with the results 

of international studies that examine investors in the 
United States, there is a significant difference between 
Brazilian and American investors in terms of the 
allocation of resources. Most American investments are 
concentrated in financial assets, while most Brazilians 
investments are concentrated in non-financial assets. The 
Federal Reserve Bulletin 2012 reveals that American 
investors mainly focused on financial investments.  
 
Table 6 and Figure 5 present investor allocation in the 
United States. 

 
Table 6: Asset Allocation held by investors in the 

United States (% of Portfolio) 

  
Source: Federal Reserve Bulletin 2012. Prepared by the 
authors. 
 

 
Figure 5. Percentage Allocation of the major 

investments held by US investors 
 
Source: Federal Reserve Bulletin 2012. Prepared by the 
authors.  
 



Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120 |   http://emaj.pitt.edu 

 
 

 
Investment Portfolios in an Emerging Economy: What Drives Portfolio’s Diversification? 
 

Page |9| Emerging Markets Journal 

The black vertical bars represent non-financial 
investments and the gray vertical bars represent financial 
investments. 
 
5. Conclusion 

 
This empirical study examines the level of 

portfolio diversification of Brazilian investors, focused 
on aspects that can influence portfolio diversification and 
the intrinsic characteristics that can lead investors to 
allocate financial resources in risky assets, furthering the 
discussion on the investment decisions within an 
emerging economic context. Moreover, the study sheds 
light on the possible cognitive influences built by 
behavioral economic theory.  

 
From a new and heterogeneous database 

composed of 500 Brazilian investors and 2,224 
observations of investments, it became clear that 
Brazilian investors allocate their resources in under-
diversified portfolios concentrated in only four different 
types of investments on average. In addition, investors 
allocate approximately 92% of their resources in low-risk 
assets. Brazilians investments concentrate investments in 
fixed-income assets, reinforcing the conservative investor 
profile. 

 
The results suggest that the characteristics of 

investors studied can play an important role in Brazilian 
investment decisions. The results indicate that investors 
endowed with higher education tend to have a higher rate 
of portfolio diversification. In addition, married investors 
tend to have more diversified portfolios, distributing their 
resources in different assets. 

 
The investment decisions of Brazilians lead to 

low portfolio diversification allocated primarily in low-
risk assets, and they suggest the limitations of investor 
rationality theory advocated by the Modern Theory of 
Finance. Economic, sociological and psychological 
concepts may play an important role in explaining the 
economic decisions in real life. Thus, investment 
decisions may be subject to behavioral influence as 
defined in the Behavioral Finance literature.  
Among the characteristics that lead investors to allocate 
resources in risky assets are age, gender, marital status 
and education level. 
 

Although this study achieves the proposed 
objectives, there are caveats that need to be identified, 
such as the possible bias generated by the investment 
database and the benchmarks adopted. These issues do 
not invalidate the evidence obtained during this study, 
but they serve as a warning to the possibility of new 
evidence by the improvement in the dataset. Future 
studies should apply or develop models that best explain 
the relationship between investment decisions and the 

effects of psychological biases defined by Behavioral 
Finance Theory. New studies can contribute significantly 
to understand the purpose of this study, and we 
recommend a replication of the study with the inclusion 
of a larger number of investor portfolios. 
 
REFERENCES 
 
Arzac, E. R., & Bawa, V. S. (1977). Portfolio 

choice and equilibrium in capital 
markets with safety-first investors. 
Journal of Financial Economics, 4(3), 
277-288. 

Barber, B. M., & Odean, T.. Trading is 
hazardous to your wealth: The common 
stock investment performance of 
individual investors. The Journal of 
Finance, 55(2), 773-806.  

Barberis, N., & Thaler, R. (2003). A survey of 
behavioral finance. Handbook of the 
Economics of Finance, 1, 1053-1128.  

Bikhchandani, S., & Sharma, S. (2000). Herd 
behavior in financial markets. IMF 
Staff papers, 279-310.  

Blume, M., & Friend, I. (1978). The changing 
role of the individual investor: Wiley 
New York. 

Coval, J. D., & Moskowitz, T. J. (1999). Home 
bias at home: Local equity preference 
in domestic portfolios. The Journal of 
Finance, 54(6), 2045-2073.  

De Bondt, W. F., & Thaler, R. H. (1994). 
Financial decision-making in markets 
and firms: A behavioral perspective: 
National Bureau of Economic 
Research. 

De Brouwer, P. J. (2009). Maslowian portfolio 
theory: an alternative formulation of 
the behavioural portfolio theory. 
Journal of Asset Management, 9(6), 
359-365.  

Fama, E. F. (1970). Efficient capital markets: 
A review of theory and empirical 
work*. The journal of Finance, 25(2), 
383-417.  

Goetzmann, W. N., & Kumar, A. (2008). 
Equity portfolio diversification*. 
Review of Finance, 12(3), 433-463.  



Volume 7 No 1 (2017)   |   ISSN 2158-8708 (online)   |   DOI 10.5195/emaj.2017.120 |   http://emaj.pitt.edu 

 

 

Pedro Luiz Albertin, Bono Milan, William Eid Jr. 
Emerging Markets Journal | P a g e  |10 

Grable, J., Lytton, R., & O'Neill, B. (2004). 
Projection bias and financial risk 

tolerance. The Journal of Behavioral 

Finance, 5(3), 142-147 

Jacowitz, K. E., & Kahneman, D. (1995). 
Measures of anchoring in estimation 
tasks. Personality and Social 
Psychology Bulletin, 21, 1161-1166.  

Kahneman, D., & Tversky, A. (1979). Prospect 
theory: An analysis of decision under 
risk. Econometrica: Journal of the 
Econometric Society, 263-291.  

Kelly, M. (1995). All their eggs in one basket: 
Portfolio diversification of US 
households. Journal of Economic 
Behavior & Organization, 27(1), 87-96.  

Kimura, H., Basso, L. F. C., & Krauter, E. 
(2006). Paradoxos em finanças: teoria 

moderna versus finanças 

comportamentais. Revista de 
Administração de Empresas, 46(1), 41-
58.   

Köbberling, V., & Wakker, P. P. (2005). An 

index of loss aversion. Journal of 
Economic Theory, 122(1), 119-131.  

Levinson, J. D., & Peng, K. (2007). Valuing 
cultural differences in behavioral 

economics. ICFAI journal of 

behavioral finance, 4, 32-47. 

Lintner, G. (1998). Behavioral finance: Why 
investors make bad decisions.The 
Planner, 13(1)7-8. 

Markowitz, H. (1952). Portfolio selection*. 
The journal of finance, 7(1), 77-91.  

Maslow, A. H. (1943). A theory of human 
motivation. Psychological review, 
50(4), 370.  

Plous, S. (1993). The psychology of judgment 
and decision making: Mcgraw-Hill 
Book Company. 

Polkovnichenko, V. (2005). Household 
portfolio diversification: A case for 
rank-dependent preferences. Review of 
Financial Studies, 18(4), 1467-1502.  

 

Shefrin, H., & Statman, M. (2000). Behavioral 
portfolio theory. Journal of financial 
and quantitative analysis, 35(02), 127-
151.  

Shiller, R. J. (1999). Human behavior and the 
efficiency of the financial system. 
Handbook of macroeconomics, 1, 
1305-1340.  

Shum, P., & Faig, M. (2006). What explains 
household stock holdings? Journal of 
Banking & Finance, 30(9), 2579-2597.  

Von Neumann, J. (1947). Morgenstern, 
0.(1944) Theory of Games and 
Economic Behavior. Princeton: 
Princeton UP.  

Yoshinaga, C. E., de Oliveira, R. F., da 
Silveira, A. D. M., & Barros, L. A. B. 
d. C. (2008). Finanças 

Comportamentais: uma introdução. 

REGE Revista de Gestão, 15(3), 25-35.