International Journal of Economics and Financial

Issues
ISSN: 2146-4138

available at http: www.econjournals.com

International Journal of Economics and Financial Issues, 2017, 7(1), 269-278.

Perceived Risk, Investment Performance and Intentions in

Emerging Stock Markets

Phung Thai Minh Trang1*, Nguyen Huu Tho2

1
Department of Accounting and Finance, Faculty of Economics and Commerce, Hoa Sen University, Ho Chi Minh City, Vietnam,
2
Department of Finance-Project, Faculty of Business Administration, University of Economics, Ho Chi Minh City, Vietnam.
*Email: [email protected]

ABSTRACT
The study built the new measurement scales of risk perception in investing in stock types trading on the emerging stock market, and then explored the
effects of perceived risk on investment performance and intentions of individual investors. The study employed mixed research methods including
in-depth interviews, a pilot study involving 50 investors, a survey distributed to 465 retail investors. Results showed that perceived risk had the
direct positive impact on both investment performance and intentions. Perceived risk also had the indirect influence on intentions to invest through
investment performance. For managerial decisions, investors are recommended to draw attention to the risks of investing in stocks labelled as “warned”,
“controlled”, and “halted trading”. The higher investors perceive the risks of these stock types, the greater they are satisfied with their investment
decisions, the recent rate of return achieved, and the more they intend to invest in the next time. Securities corporations have regularly organized
seminars, workshops or training courses about investing in the kinds of stocks, updated regulations of stock investment in time, and improved the
quality of listed companies to attract more investors to the stock market.
Keywords: Investment Performance, Intentions, Perceived Risk, Emerging Stock Markets
JEL Classifications: E44, G31

1. INTRODUCTION certainly affected by psychology. More importantly, “perceived

risk” is considered as powerful instruments in investment since
Vietnam stock exchange, founded in 2000, consists of the people seem to expect to prevent risk rather than maximize utility
Ho Chi Minh City (HCMC) stock exchange (HOSE) and the when making investing decisions (Mitchell, 1999).
Ha Noi stock market (HNX). The share market capitalization
reached VND 1,325,000 billion, approximately USD 62 billion, “Perceived risk” has recently been studied in a number of fields
equivalent to 34 per cent of the Vietnam gross domestic product such as online consumer product and service, e-banking, and
in 2015 (www.bnews.vn). Moreover, the total number of private stock markets. However, methods for assessing and measuring
investors’ accounts in the Vietnam stock market was 1552 million this perception are diverse. For instance, the domain of consumer
in 2015 (www.baomoi.com). This number makes up such a small goods focused on product risk, performance risk or financial
percentage of the overall Vietnamese population when Vietnam’s risk (Dai et al., 2014; Forsythe and Shi, 2003), conversely, the
current population is between 90 and 91 million. With improved internet banking industry related to social risk, time loss risk,
policies, the Vietnam stock market has increasingly brought opportunity cost risk, and information risk (Kassim & Ramayah,
opportunities for investments. In fact, making investing decisions 2015). Interestingly, in the financial field risk perception was
on intangible products is not easy and might base on psychology as measured by attitude towards risk such as risk taking and risk
Schwager (1993) quoted “trading is emotion; It is mass psychology, aversion (Kahneman and Tversky, 1979; Barberis and Huang,
greed and fear” (p. 49). Additionally, Baker and Nofsinger (2010) 2001; Mayfield et  al., 2008). Apparently, perceived risk of
also stated that we were all human beings, thus our behavior was investing in types of stocks trading on the stock market has yet

International Journal of Economics and Financial Issues | Vol 7 • Issue 1 • 2017 269
 Trang and Tho: Perceived Risk, Investment Performance and Intentions in Emerging Stock Markets

to be explored; particularly its direct and indirect impact on was learned that functional risk, physical risk and financial risk had
investment performance and intentions have not been studied in no significant impact on this attitude. In addition, attitude towards
Vietnam. Hence, three key research questions are raised: “What the use of electronic banking (e-banking) had a positive influence
new measurement scales of perceived risk are explored?”; “to on intention of using internet banking in the future.
what extent does “perceived risk” directly influence investment
performance and intentions?”; “to what extent does “perceived Cuong and Jian (2014) explored attitude toward the behavior
risk” indirectly influence investment intentions through investment such as overconfidence excessive optimism, herd behavior, and
performance?” In addition, The main objectives are Firstly to build psychology of risk positively affected behavior intention to
the new measurement scales of perceived risk of investing stock decision-making of individual investors on the Vietnam stock
types on the Vietnam stock market. Secondly, to test the study test market. This result was consistent with other scholars such as
the validity and reliability of the new scales, and then to explore Gervais et al. (2002), Johnsson et al. (2002).
the direct and indirect effects of perceived risk on investment
performance and intentions. Dai et al. (2014) explored that perception of product risk had the
strong impact on online purchase intentions for both digital and
2. LITERATURE REVIEW non-digital products. The findings were similar to Forsythe and
Shi (2003)’s. However, the financial risk affected online purchase
2.1. Consumer Behavior and Perceived Risk intentions for non-digital products only while Bhatnagar et al.
Consumer behavior is defined as “the process and activities (2000), Chang et al. (2005), and Forsythe et al. (2006) gave results
people engage in when searching for, selecting, purchasing, using, of this impact for both digital and non-digital ones. Likewise,
evaluating, and disposing of products and services so as to satisfy privacy risk had no impacts on buying intentions for both kinds
their needs and desires” (Belch and Belch, 2003. p. 105). Howard of products, whereas online safety had the strong influence strong
and Sheth (1969) initially developed the model of consumer influence in researches of Miyazaki and Fernandez (2001) and
behavior and eventually became the “theory of buyer behavior.” Noort et al. (2008). Different consequences could be caused by
The theory is widely used since it has given a sophisticated different respondents.
combination of the diverse social, psychological, and marketing
impacts on buyer choices (Foxall, 1990). Bauer (1960), initially Alleyne and Broome (2011) also studied future investors who
stated that perceived risk was involved only in subjective risk. were currently business students in an undergraduate institution.
Perceived risk is defined as “in the sense that any action of a More impressively, these authors added risk propensity as an
consumer will produce consequences which he cannot anticipate independent variable to three independent variables in TPB. Four
with anything approximating certainty, and some of which at least predictors consisted of attitude toward the behavior, subjective
are likely to be unpleasant” (Bauer, 1960). Perceived risk is also norms, perceived behavioral control, and risk propensity positively
considered as” the citizen’s subjective expectation of suffering a affected intentions to invest. This exploration matched results of
loss in pursuit of a desire outcome” (Warkentin et al., 2002. p. 160). East (1993) and Gopi, Ramayah (2007).

2.2. The Theory of Planned Behavior (TPB) and Oberlechner and Osler (2008) used primary data: a survey
Intentions distributed to 416 current market professionals in North America
The TPB derived from the reasoned action theory is “open to and found the significant relationship between behavioral finance,
the inclusion of additional predictors if it can be shown that they overconfidence and investment performance. Lin and Swanson
capture a significant proportion of the variance in intention or (2003) used secondary data to explore the significant relationship
behavior after the theory’s current variables have been taken between trading behavior and investment performance of foreign
into account” Ajzen (1991. p. 199). This theory proposes that an investor trading on the Taiwan stock market. These authors
individual’s attitude toward a behavior, subjective norms, and measured the outcome variable through three dimensions such as
perceived control can influence intentions (Ajzen, 1991). An raw returns, returns adjusted from risk or momentum. In addition,
attitude toward a behavior is considered as one’s assessment of Kim and Nofsinger (2003) also showed that possessing risky
the given behavior based on his/her beliefs; a subjective norm stocks, purchasing current winners, influenced performance of
relates to one’s perception; perceived control concern the perceived Japanese individual investors.
difficulty to carry out the behavior (Ajzen, 1991). This theory
provides a model which can predict one’s behavior via intentions 3. RESEARCH METHODS AND MODELS
defined as individual perception towards likelihood to conduct
behavior (Ajzen and Fishbein, 1980). The article employed mixed research methods: Qualitative and
quantitative with techniques comprising in-depth interviews and
2.3. Recent Research of Perceived Risk, Investment survey. Specifically, the study firstly applied in-depth interviews
Performance and Investment Intentions which lasted for six weeks to six brokerage managers who each
Kassim and Ramayah (2015) studied the relationship among have 10 years of experience in financial investment. The aim of
perceived risk factors and intentions to continue using internet these interviews is to explore the new measurement scales of
banking in Malaysia. These authors found that social risk, time loss perceived risk. These brokerage managers are working for top
risk, opportunity cost risk, and information risk, all significantly securities corporations in Vietnam such as HCMC Securities
affected the attitude towards the use of internet banking. Also, it Corporation, Rong Viet Securities Corporation, Bao Viet Securities

270 International Journal of Economics and Financial Issues | Vol 7 • Issue 1 • 2017
 Trang and Tho: Perceived Risk, Investment Performance and Intentions in Emerging Stock Markets

Corporation, and Thien an Viet Commerce and Investment Joint changes to the model if no other problems are associated with
Stock Company. Secondly, the authors used the following five- those two items. Standardized residual covariance is not more
point Likert scale survey questions: Perceived risk explored by than |2|. Modification indices (MI) of approximately 4.0 or greater
authors of this study, investment performance from Oberlechner suggest that the fit could be improved significantly by freeing the
and Osler (2008), and investment intentions from Dodds et  al. corresponding path to be estimated.
(1991) and Soderlund and Ohman (2003). Thirdly, the pilot study
was distributed to 50 investors with the aim of testing the reliability 3.3. Model Testing
of all items of variables. Finally, the questionnaires were sent to This study established the SEM for path analysis: Exogenous
480 private investors trading on the Vietnam stock exchange, and construct: Perceived risk labeled as PERISK and endogenous
465 valid returned ones. The survey spanned 8 months with strong constructs: Investment performance (INVEST) and intentions
support from securities corporations. (INTENT). A  path diagram of specific hypothesized structural
relationships and measurement specification are shown in Figure 1.
The study used SPSS and AMOS software for exploration factor
analysis (EFA) with principal axis factoring for extraction method Basing on overall perceived risk relevant to product specific risk
and Promax with Kaiser normalization. Confirmation factor and product class risk (Dowling and Staelin, 1994) and the results
analysis (CFA) and structural equation modeling (SEM) were from in-depth interviews, this study explored types of stocks
applied for exploring the effects of perceived risk on investment trading on the stock market. The study expects to have the positive
performance and intentions. The authors applied SEM model correlation between perceived risk and investment performance.
for the research since SEM is “a family of statistical models that This relationship might be written under the following equation:
seek to explain the relationships among multiple variables” (Hair
et al., 2014. p. 546). Additionally, SEM can examine a series of INVESTi = α + β1PERISKi + εi(1)
dependence relationships simultaneously as well as is useful to
know how to spot an exogenous and endogenous construct. Where, INVEST: Investment performance, PERISK: Perceived
risk, i: ith investor.
For testing the reliability of variables. Hair et al. (2014) suggests
that Cronbach’s alpha should be from 0.6; the Kaiser–Meyer– Perceived risk of individual investors is considered as one’s
Olkin and Bartlett’s test should be more than 0.6 and the accepted attitude toward a behavior in TPB (Ajzen, 1991). Many scholars
significant level (P) is not more than 0.05. Initial Eigen value explored the effects of perceived risk on intentions to invest
should be >1 and cumulative percentage is not <50%. FLs of the (Cuong and Jian, 2014; Alleyne and Broome, 2011; Gopi and
items on a factor are >0.3, the corrected item-total correlation index Ramayah, 2007; East, 1993). However, risk perception focused
is 0.3 and the rotation sums of squared loadings is suggested to on types of stocks seems to be new ideas. The study, therefore,
be more than 50%. After loading factors, the following tests of proposes a model of this impact shown under the equation below:
measurement theory validation with CFA are:
INTENTi = µ + ℓ1PERISKi + θi(2)
3.1. Unidimensionality
Multiple fit indices should be used to assess a model’s goodness- Where, INTENT: Investment intentions, PERISK: Perceived risk,
of-fit (GOF) and include: The value of Chi-square (χ2) # 0, and i: ith investor.
P ≤ 0.05; the associated df: χ2/df ≤ 0.5; one incremental fit index:
Comparative fit index ≥ 0.8; Tucker Lewis Index (TLI) ≥ 0.8; one Gopi and Ramayah (2007) mentioned high earnings/return as
GOF index: GFI ≥ 0.8; TLI ≥ 0.8; one badness-of-fit index: Root an item in “attitude toward the behavior” that had the positive
mean square error of approximation (RMSEA) ≤ 0.08. Details are influence on intentions to use online stock trading. This
presented in Appendix. consequence was likely to be similar to East (1993)’s and Alleyne
and Broome (2011)’s. In practice, if one is satisfied with his/her
3.2. Construct Validity earnings/return, it seems to be certain that he/she will invest in
Validity is defined as the extent to which research is accurate. stocks in the next time. Thus, the model relevant to this relationship
First, FLs: Standardized loading estimates should be 0.5 or higher; might be presented under the equation below:
second, convergent validity: Average variance extracted (AVE)
should be 0.5 or greater to suggest adequate convergent validity. INTENTi = ϭ + ⱱ1INVESTi + ∂i (3)
Construct reliability should be 0.6 or higher to indicate adequate
convergence or internal consistency. Third, discriminant validity: Where, INTENT: Investment intentions, INVEST: Investment
Compare the AVE values for any two constructs with the square performance, i: ith investor.
of the correlation estimate between these two constructs.
The TPB has been studied by many a scholar with a number of
Model diagnostics: Standardized residuals are less than |2.5| do domains such as consumer products, banking, finance, investment.
not suggest a problem, but greater than |4.0| suggest a potentially More notably, behavior intention has been considered as a
unacceptable degree of error that may call for the deletion of an dependent variable over the past decades (Ajzen, 1991; Dodds
offering item. Standardized residuals for any pair of item between et  al., 1991; East, 1993; Soderlund and Ohman, 2003), but
|2.5| and |4.0| deserve some attention, but may not suggest any “perceived risk” of this study is the new independent variable.

International Journal of Economics and Financial Issues | Vol 7 • Issue 1 • 2017 271
 Trang and Tho: Perceived Risk, Investment Performance and Intentions in Emerging Stock Markets

Figure 1: A path diagram of perceived risk, investment performance and intentions

As a result, the authors also suggest considering intentions as the The study needs to be at least 95% confident that all the issues have
outcome variable and testing the effects of perceived risk and emerged which are represented in 40% of the population because
investment performance on intentions to invest. The equation of brokerage managers living in HCMC. After computing, the
might be indicated as follows: number of interviewees required was 6.

INTENTi = ᴪ + γ1PERISKi + ᴣ1INVESTi + πi(4) 3.4.3. Sample size for surveys

The number of investors who actually traded on the Vietnam stock
Where, INVEST: Investment performance, INTENT: Investment market in 2015 was around one million five hundred (1.5 million)
intentions, PERISK: Perceived risk, i: ith investor. (www.baomoi.com). The study employed the formula of Krejcie
and Morgan (1970) to calculate the sample size as follows:
3.4. Scope of the Research and Sample Size
3.4.1. Scope of the research χ 2 NP(1 - P)
S=
Vietnam has 63 provinces, two stock exchanges including HNX d 2 (N - 1)+ χ 2 P(1 - P)
and HOSE, and three key cities HCMC, Ha Noi City, and Da
Nang City. HCMC is the largest City and represented for the Where,
South; Ha Noi City as the Capital is behalf on the North; and S = Required sample size;
Da Nang City as the central Vietnam. Questionnaires were sent χ2 = The table value of Chi-square for 1 degree of freedom at the
to individual investors living in cities of Vietnam, but in these desired confidence level (3.841);
main areas more than others since these cities are the heart of N = The population size;
Vietnam’s economy. P = The population proportion (assumed to be 0.50 since this
would provide the maximum sample size);
3.4.2. Sample size for interviews d = The degree of accuracy expressed as a proportion (0.05).
The question has repeatedly been raised that how many
interviews are needed in a qualitative research? In practice, there Basing on the formula above, the sample size required was 384.
is no sampling rule and “the specific number will depend on the The number of investors requested to ensure the reliability and
complexity of the research questions and of the interview topic validity of the study was consistent with suggestions of Hair et al.
guide, the diversity of the sample and the nature of the analysis” (2014) “minimum sample size – 300 for models with seven or
(Francis et al., 2010. p. 1234). A researcher will stop collecting fewer constructs” (p. 574). Sample size for the pilot test was 50
data if one reaches “saturation” defined as “the point at which private investors. Then, 480 questionnaires were sent to individual
no new relevant information is forthcoming even if more people investors and only 465 returned valid questionnaires.
are interviewed” (Galvin, 2015. p.  3). However, to be more
conceivable, the authors used the formula of small sampling of
qualitative data proposed by Galvin (2015) as follows:
4. THE FINDINGS
ln (1 - P) 4.1. Findings of the New Measurement Scales
n =
ln (1 - R ) The authors initially interviewed six brokerage directors working
for top securities corporations in Vietnam and explored the new
Where, n: The number of interviews, P: Level of confidence, measurement scales of perceived risk. Five types of stocks trading
R: Proportion of the population. on the Vietnam stock market include: (1) Stocks labeled as

272 International Journal of Economics and Financial Issues | Vol 7 • Issue 1 • 2017
 Trang and Tho: Perceived Risk, Investment Performance and Intentions in Emerging Stock Markets

“warned”; (2) stocks labeled as “controlled”; (3) stocks that have Table 1: Summary of respondents’ characteristics with
halted trading or suspended trading; (4) highly speculative stocks; highest percentage
(5) blue chip stocks, fund certificates, or VN30 indexed stocks. Characteristics Percentage
Five questions with 5-point Likert scales, from totally disagree Male 58.8
to totally agree, concerned with perceived risk are proposed as Age: 26‑35 28.4
follows: University degree 70.9
Income level: 6‑12 million 44.9
• RISK1: I feel it is risky to invest in stocks labeled as “warned”
Work experience: <5 years 49.1
• RISK2: I  feel it is risky to invest in stocks labeled as Seniority of stock investment: 1‑3 years 31.6
“controlled” Training courses: Yes 63.5
• RISK3: I feel it is risky to invest in stocks that have halted or The amount of investment: 100‑300 million VND 26.7
suspended trading
• RISK4: I feel it is risky to invest in highly speculative stocks
4.3.2. Results of FLs
• RISK5: I feel it is risky to invest in blue chip stocks, fund
With extraction method: Principal axis factoring and rotation
certificates or VN30 indexed stocks.
method: Promax with Kaiser Normalization, EFA results divided
all items into three groups. Namely, the first group labeled as
Next, defining individual constructs based on published literature
INTENT included INTEND3, INTEND4, INTEND2, INTEND1,
review and interviews with brokerage managers, the study
emphasized on three key constructs below: INTEND5, and INTEND6; the second group named PERISK
• Perceived risk (PERISK): The extent to which investors involved RISK1, RISK2, and RISK3; the final factor entitled
perceive risk of investing in types of stocks trading on the INVEST covered three items such as RETURN1, RETURN2,
stock market and RETURN3. These FLs met standardized FLs >0.7, variance-
• Investment performance (INVEST): The extent to which extracted measures exceed 70% (74.776%) (Table 3).
investors are satisfied with the rate of return of recent stock
investment compared to their expectation as well as with the 4.4. Assessing the Structural Model Validity
investment decisions 4.4.1. Overall fit
• Investment intentions (INTENT): The extent to which Most GOF indices were within a range that would be associated
investors intend/plan/want/would like to invest in stocks in with good fit. Firstly, the overall model, Chi  square (χ2), was
the next time. 99.566 with degrees of freedom. The P value associated with this
result was 0.000 (P < 0.05). Secondly, the value for RMSEA, an
Finally, the test of face validity was performed as follows: the absolute fit index, was 0.045 below the 0.08 guideline for a model
experts - interviewees proposed a set of multiple-item reflective with 12 measured variables and a sample size of 465. Using the
scales to assess each construct. The conceptual definitions also 90% confidence interval for this RMSEA, the study concluded
matched well with the item wordings. Moreover, a pretest was the true value of RMSEA was between 0.032 and 0.058. Next,
carried out where judges of items of three variables were suitable the standardized root mean square residual with a value of 0.0212
for the construct names. After establishing face validity, the was below the conservative cut-off value of 0.05. Lastly, the
study proceeded to finalize the scales by pretesting involving 50 normed (χ2) of 1.952 was considered very well. These diagnostics
investors, and then, a survey distributed to 465 investors. The suggested the model provides a good overall fit. Details were
results were presented below. presented in Table 4 and Appendix.

4.2. Respondent Description 4.4.2. Convergent validity

As Table 1 presented participant information, there were 465 Table 5 displayed standardized loadings or standardized regression
individual investors, trading on the Vietnam stock market who weights using AMOS terminology. The lowest loadings obtained
were male was 58.8% of the total sampling. Most of them, 51.2% 0.706, linking INVEST to item RETURN2. The AVE estimated
were from 26 to 35  years old. 70.9% of the total participants and the construct reliabilities were shown at the bottom of Table 5.
achieved university degree, 44.9% from 6 to 12 million per month. The AVE estimated range from 54.6% for INVEST to 85.6%
49.1% of them had <5 years of work experience and 31.6% owned for PERISK, in which all exceeded the 50 per cent. Construct
1-3 years of seniority of stock investment. Individual investors reliabilities ranged from 0.78 for INVEST to 0.95 for PERISK,
took training courses of investment stocks, taking 63.5% and and 0.96 for INTENT, which exceeded 0.7, and suggested adequate
26.7% investors used from 100 to 300 million VND for their stock reliability. Taken together, the evidence supported the convergent
investment (Table 1). validity of the measurement model, and the model fitted relatively
well. Hence, all the items were retained at this point and adequate
4.3. EFA Results evidence of convergent validity was provided.
4.3.1. Results of reliability test
The study applied 5-point Likert scales: 1 point for totally 4.4.3. Discriminant validity
disagree and 5 points for totally agree. The construct of PERISK Table 6 showed AVE of all variables was greater than the estimated
included 5 items; 3 items involving the construct of INVEST, and correlation. Namely, AVE of PERISK of 0.9 was greater than the
the construct of INTENT owned 3 components. Moreover, the estimated correlation between PERISK and INTENT (0.197), and
Cronbach’s alpha of three variables was >0.7 (Table 2). INVEST (0.16). Likewise, AVE of INTENT of 0.8 was greater

International Journal of Economics and Financial Issues | Vol 7 • Issue 1 • 2017 273
 Trang and Tho: Perceived Risk, Investment Performance and Intentions in Emerging Stock Markets

Table 2: Item description and Cronbach’s alpha of all variables

Item Scale type Description Construct Cronbach’s alpha
RISK1 1‑5 Likert I feel it is risky to invest in stocks labeled as “warned” PERISK
RISK2 Above I feel it is risky to invest in stocks labeled as “controlled” PERISK
RISK3 Above I feel it is risky to invest in stocks that have halted or suspended trading PERISK
RISK4 Above I feel it is risky to invest in highly speculative stocks PERISK
RISK5 Above I feel it is risky to invest in blue chip stocks, fund certificates or VN30 indexed PERISK
PERISK 0.747
RETURN1 Above The return rate of my recent stock investment meets my expectation INVEST
RETURN2 Above My rate of return is equal to or higher than my last rate of return INVEST
RETURN3 Above I feel satisfied with my investment decisions INVEST
INVEST 0.781
INTEND1 Above I intend to invest in shares in the next time INTENT
INTEND2 Above I plan to invest in shares in the next time INTENT
INTEND3 Above I want to invest in shares in the next time INTENT
INTEND4 Above There is a high probability I will invest shares in the next time INTENT
INTEND5 Above I am likely to invest in shares in the next time INTENT
INTEND6 Above I would like to invest in shares in the next time INTENT
INTENT 0.958

Table 3: Summary of FL of items Table 7, all standardized residual covariance had values under
Group 1 FL Group 2 FL Group 3 FL |2| which met the standards of good measurement practice and
INTEND3 0.939 RISK1 0.957 RETURN1 0.769 appeared to hold quite well.
INTEND4 0.895 RISK2 0.926 RETURN2 0.731
INTEND2 0.893 RISK3 0.892 RETURN3 0.714 4.4.5. MIs
INTEND1 0.885
As Table 8 showed, most of the values above 4.0 were associated
INTEND5 0.878
INTEND6 0.846 with the items in the construct and there were no large values for
the variables. It concluded that the model had a solid theoretical
FL: Factor loadings
foundation and questionnaire measured these key construct well.
Table 4: GOF measurers of the research model 4.5. Results of Testing Models
GFI The research model After testing the reliability and validity of the research model,
Chi‑square (χ2)
the study used SEM with AMOS software to find out results. All
Chi‑square 99.566
Degree of freedom 51 the assumptions were supported and demonstrated in Figure 2.
P 0.000
Absolute fit measures The first model, H1, was supported. The following equation might
GFI 0.965 be demonstrated as follows:
RMSEA 0.045
90% confidence interval of RMSEA 0.032‑0.058
RMR 0.019 INVESTi = 0.16PERISKi** (1’)
SRMR 0.0212
Normal Chi‑square 1.952 The equation (1’) showed that PERISK positively affected
Incremental fit indices INVEST. Specifically, the standardized coefficient for PERISK
NFI 0.979 was 0.16 (**P < 0.01) in which it can be mathematically
TLI 0.987
interpreted that for every additional 1% of PERISK, INVEST will
CFI 0.99
RFI 0.973 increase by 16%. In other words, the higher investors perceive
Parsimony fit indices risks of investing in stocks labeled as “warned,” “controlled” or
AGFI 0.946 stocks that have halted or suspended trading, the greater they
PNFI 0.757 are satisfied with their investment decisions or the recent rate of
GFI: Goodness‑of‑fit index, RMSEA: Root mean square error of return achieved. This consequence is consistent with previous
approximate, RMR: Root mean square residual, SRMR: Standardized root mean research conducted by Oberlechner and Osler (2008) and Lin
residual, NFI: Normed fit index, NNFI: Non‑normed fit index, TLI: Tucker‑Lewis index,
CFI: Comparative fit index, RFI: Relative fit index, AGFI: Adjusted goodness‑of‑fit and Swanson (2003).
index, PNFI: Parsimony normed fit index
The result also supported the second model. The equation may
than its correlation with INVEST (0.39). Therefore, this test be written as follows:
indicated that there were no problems with discriminant validity
for the research model. INTENTi = 0.14PERISKi** (2’)

4.4.4. Standardized residual covariance PERISK positively influenced INTENT with the standardized
Standardized residual covariance are analogous to Z scores and coefficient for PERISK at 0.14 (**P < 0.01) as the equation
most should have a value under |2| or |2.58| (Byrne, 2013). In (2’) presented above. It can be explained that an increase of

274 International Journal of Economics and Financial Issues | Vol 7 • Issue 1 • 2017
 Trang and Tho: Perceived Risk, Investment Performance and Intentions in Emerging Stock Markets

1 standard deviation in PERISK will increase by 14 standard will have more plans to invest in stocks in the future. This finding
deviations of INTENT. In effect, if investors enhance the level is similar to previous researches of scholars such as Kassim
of risk perception of investing in stocks labeled as “warned,” and Ramayah (2015), Cuong and Jian (2014), Dai et al. (2014),
“controlled” or stocks that have halted or suspended trading, they Alleyne and Broome (2011).

The finding indicated INVEST had a positive impact on INTENT

Table 5: Standardized FL, AVE, and reliability estimates
as the third model proposed. The following equation might be
INTENT PERISK INVEST
depicted as follows:
INTEND4 0.897
INTEND2 0.898
INTEND1 0.888 INTENTi = 0.37INVESTi***(3’)
INTEND5 0.878
INTEND6 0.866 More interestingly, INVEST had the strong impact on INTENT
INTEND3 0.914 with the standardized coefficient for INVEST at 0.37 (***P <
RISK2 0.952
RISK1 0.925
0.001). The equation (3’) can be interpreted as follows: if INVEST
RISK3 0.897 increases/decreases by 1 standard deviation, INTENT will go up/
RETURN3 0.792 down by 0.37 standard deviations. More specifically, an increase
RETURN1 0.716 of 1% of satisfaction level of investment decisions or of the
RETURN2 0.706 recent rate of return achieved will lead to an increase of 37% in
AVE 79.3% 85.6% 54.6% intentions to invest in stocks. Apparently, investment performance
CR 0.96 0.95 0.78
and intentions have a strong relationship (37%). This exploration
AVE: Average variance extracted, CR: Construct reliability, FL: Factor loadings
is similar to research of Gopi and Ramayah (2007), Alleyne and
Broome (2011), and East (1993).
Table 6: AVE and correlation estimates
PERISK INTENT INVEST Finally, the results also supported the fourth model related to the
PERISK 0.925 indirect effects of perceived risk (PERISK) on intentions to invest
INTENT 0.197 0.890 (INTENT) through investment performance (INVEST). More
INVEST 0.160 0.390 0.739 impressively, PERISK had direct and indirect impact on INTENT.
AVE: Average variance extracted Standardized coefficient total effects were shown in Table 9.

Table 7: Standardized residual covariance

RETU2 RETU1 RETU3 RISK3 RISK1 RISK2 INTE6 INTE5 INTE1 INTE2 INTE4 INTE3
RETU2 0
RETU1 0.158 0
RETU3 −0.07 −0.04 0
RISK3 −0 −0.41 0.766 0
RISK1 0.217 −0.69 0.593 −0.01 0
RISK2 −0.55 −0.75 0.507 0 0.006 0
INTE6 0.387 0.351 0.871 0.846 0.259 0.46 0
INTE5 −0.02 0.148 0.195 0.061 −0.32 −0.56 0.226 0
INTE1 −0.48 −0.11 0.411 0.647 0.253 −0.31 −0.57 −0.23 0
INTE2 0.042 0.112 0.056 0.708 0.39 0 −0.06 −0.2 0.599 0
INTE4 0.304 −1.02 0.508 1.231 0.29 0.636 0.417 0.297 −0.16 −0.41 0
INTE3 −0.84 −0.35 −0.42 −0.2 −1.04 −1.09 −0.07 −0.05 0.133 0.051 −0.02 0

Figure 2: Path significant coefficient of the research model

International Journal of Economics and Financial Issues | Vol 7 • Issue 1 • 2017 275
 Trang and Tho: Perceived Risk, Investment Performance and Intentions in Emerging Stock Markets

As shown in Table 9, total effects of PERISK on INTENT was More significantly, investors strictly pursue this allocation in order
0.1992. This number mathematically means that when PERISK and to avoid bankrupt if they face with failure.
INVEST simultaneously goes up/down by 1 standard deviation,
INTENT increases/decreases by 0.5692 standard deviations. The Furthermore, securities corporations should organize seminars,
equation (4’) might be proposed as follows: conferences, meetings or training courses in which investors have
a number of opportunities to discuss as well as learn precious
INTENTi = 0.1992PERISKi + 0.37 INVESTi(4’) lessons. Particularly, updating regulations of stock investment
and supporting investors for using financial leverage are one of
5. CONCLUDING REMARK the important missions of securities corporations. However, some
of them have not enough capital for this support, which caused
The study primarily met requirements of the reliability and validity difficulties for investors. The State Securities Commission of
tests and generalization to the entire individual investors trading Vietnam, therefore, needs to control and mitigate a quantity of
on the Vietnam stock market. Additionally, the study built the securities corporations and listed companies because of an increase
new measurement scales of perceived risk, and then using these of capacity as well as the services quality.
measurement scales to explore the effect of perceived risk on
investment performance, and on investment intentions. Especially, Finally, the limitation of the study is to focus on the relationship
perceived risk had direct and indirect impact on intentions to among perceived risk, investment performance, and investment
invest. Generally, most of the investors had high risk perception in intentions of individual investors. In fact, many behavioral bias
investing stock types, in which the higher investors perceive risks might affect investment intentions such as representativeness bias,
in investing, the greater they gratify their investment decisions or overconfidence, anchoring bias, gambler’s fallacy, availability
the more they intend to invest in stocks. bias, herding, over-under reaction, mental accounting, self-control,
regret aversion, etc. Among heuristics, gambler’s fallacy has been
Importantly, the study answered three research questions. Firstly, proposed to require further research since most of the Vietnamese
the new measurement scales (5 items) of perceived risk. Secondly, investors are young and less than 5 years of experience in stock
the level of impact of perceived risk on investment performance by investment. They still believe that if stocks have traded many times
0.16, and intentions by 0.14; thirdly, the indirect effect of perceived for this period, these stocks will be less traded next periods or if
risk on intentions through investment performance by 0.1992. a stock’s price has fallen in multiple sessions, the price will be
impossible to decline more. In addition, how perceived uncertainty
In addition, investors need to understand what cases stocks are of an investor is before making a decision: whether or not one is
placed under alert, halted trading or controlled. Normally, stocks afraid of making a mistake. This, perception should also be taken
violate the securities listing regulations at HOSE with Decision notice since it might affect investment intention and performance.
No.  10/QĐ-SGDHCM dated 10/01/2014 such as a decrease
of charter capital under VND 120 billion, negative profit after REFERENCES
tax or accumulated losses, stopped business or stopped main
activities from more than 3 months, often violating regulations of Ajzen, I. (1991), The theory of planned behavior. Organizational Behavior
information announcement, and the stock not trading within 6 or and Human Decision Processes, 50(2), 179-211.
9 months for being controlled. According to experts, investing high Ajzen, I., Fishbein, M. (1980), Theory of reasoned action/Theory of
risky stocks was still better than playing games or buying lucky planned behavior. Social Psychology, 2007, 67-98. Available from:
lottery tickets that their results were completely based on destiny. http://www.hsc.usf.edu/~kmbrown/TRA_TPB.htm.
However, before investing, investors should divide up their total Alleyne, P., Broome, T. (2011), Using the theory of planned behaviour
income into smaller parts, for instance, short-term, long-term and risk propensity to measure investment intentions among future
investors. Journal of Eastern Caribbean Studies, 36(1), 1-20.
investments, savings, daily expenses, and high risk investment.
Baker, H.K., Nofsinger, J.R., editors. (2010), Behavioral Finance:
Investors, Corporations, and Markets. Vol. 6. New York: John Wiley
Table 8: MIs of the variables & Sons.
Variables MI Barberis, N., Huang, M. (2001), Mental accounting, loss aversion, and
INTEND4←RETURN1 4.651 individual stock returns. The Journal of Finance, 56(4), 1247-1292.
INTEND4←RISK2 4.239 Bauer, R.A. (1960), Consumer Behavior as Risk-Taking, Dynamic
INTEND3←PERISK 6.721 Marketing for Changing World. Chicago: American Marketing
INTEND3←RISK1 7.419 Association.
INTEND3←RISK2 6.449 Belch, G.E., Belch, M.A. (2003), Advertising and Promotion: An
MI: Modification indices Integrated Marketing Communications Perspective. Boston: The
McGraw-Hill.
Table 9: Standardized coefficient total effects of INTENT Bhatnagar, A., Misra, S., Rao, H.R. (2000), On risk, convenience, and
Internet shopping behavior. Communications of the ACM, 43(11),
Variables INTENT
98-105.
Direct Indirect Total Chang, M.K., Cheung, W., Lai, V.S. (2005), Literature derived reference
PERISK 0.14 0.0592 0.1992 models for the adoption of online shopping. Information and
INVEST 0.37 ‑ 0.37 Management, 42(4), 543-559.
Total 0.51 0.0592 0.5692 Cuong, P.K., Jian, Z. (2014), Factors influencing individual investors

276 International Journal of Economics and Financial Issues | Vol 7 • Issue 1 • 2017
 Trang and Tho: Perceived Risk, Investment Performance and Intentions in Emerging Stock Markets

behavior: An empirical study of the Vietnamese stock market. Kahneman, D., Tversky, A. (1979), Prospect theory: An analysis of
American Journal of Business and Management, 3(2), 77-94. decision-making under risk. Econometrica, 47(2), 263-291.
Dai, B., Forsythe, S., Kwon, W.S. (2014), The impact of online shopping Kassim, N.M., Ramayah, T. (2015), Perceived risk factors influence on
experience on risk perceptions and online purchase intentions: Does intention to continue using Internet banking among Malaysians.
product category matter? Journal of Electronic Commerce Research, Global Business Review, 16(3), 393-414.
15(1), 13-24. Kim, K.A., Nofsinger, J.R. (2003), The behavior and performance of
Dodds, W.B., Monroe, K.B., Grewal, D. (1991), Effects of price, brand, individual investors in Japan. Pacific Basin Finance Journal, 11(1),
and store information on buyers product evaluations. Journal of 1-22.
Marketing Research, 28(3), 307-319. Krejcie, R.V., Morgan, D.W. (1970), Table for determining sample
Dowling, G.R., Staelin, R. (1994), A model of perceived risk and intended size from a given population. Educational and Psychological
risk-handling activity. Journal of Consumer Research, 21, 119-134. Measurement, 30, 607-610.
East, R. (1993), Investment decisions and the theory of planned behaviour. Lin, A.Y., Swanson, P.E. (2003), The behavior and performance of
Journal of Economic Psychology, 14(2), 337-375. foreign investors in emerging equity markets: Evidence from Taiwan.
Forsythe, S., Liu, C., Shannon, D., Gardner, L.C. (2006), Development of International Review of Finance, 4(3-4), 189-210.
a scale to measure the perceived benefits and risks of online shopping. Mayfield, C., Perdue, G., Wooten, K. (2008), Investment management
Journal of Interactive Marketing, 20(2), 55-75. and personality type. Financial Services Review, 17(3), 219-228.
Forsythe, S.M., Shi, B. (2003), Consumer patronage and risk perceptions
Mitchell, V.W. (1999), Consumer perceived risk: Conceptualisations and
in Internet shopping. Journal of Business Research, 56(11), 867-875.
models. European Journal of Marketing, 33(1/2), 163-195.
Foxall, G. (1990), Consumer Psychology in Behavioral Perspective.
Miyazaki, A.D., Fernandez, A. (2001), Consumer perceptions of privacy
Frederick, MD: Beard Books.
and security risks for online shopping. Journal of Consumer Affairs,
Francis, J.J., Johnston, M., Robertson, C., Glidewell, L., Entwistle, V.,
35(1), 27-44.
Eccles, M.P., Grimshaw, J.M. (2010), What is an adequate sample
Noort, G.V., Kerkhof, P., Fennis, B.M. (2008), The persuasiveness of
size? Operationalising data saturation for theory-based interview
online safety cues: The impact of prevention focus compatibility of
studies. Psychology and Health, 25(10), 1229-1245.
Web content on consumers risk perceptions, attitudes, and intentions.
Galvin, R. (2015), How many interviews are enough? Do qualitative
interviews in building energy consumption research produce reliable Journal of Interactive Marketing, 22(4), 58-72.
knowledge? Journal of Building Engineering, 1, 2-12. Oberlechner, T., Osler, C.L. (2008), Overconfidence in currency markets.
Gervais, S., Heaton, J.B., Odean, T. (2002), The positive role of Available from: http://www.ssrn.1108787.
overconfidence and optimism in investment policy. Rodney L. White Schwager, J.D. (1993), Market Wizards: Interviews with Top Traders.
Center for Financial Research-Working Papers. New York: Collins.
Gopi, M., Ramayah, T. (2007), Applicability of theory of planned Shadfar, S., Malekmohammadi, I. (2013), Application of structural
behavior in predicting intention to trade online: Some evidence from equation modeling (SEM) in restructuring state intervention
a developing country. International Journal of Emerging Markets, strategies toward paddy production development. International
2(4), 348-360. Journal of Academic Research in Business and Social Sciences,
Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E. (2014), Multivariate 3(12), 576-618.
Data Analysis. Vol. 7. UK: Pearson Education Limited. Soderlund, M., Ohman, N. (2003), Behavioral intentions in satisfaction
Howard, J.A., Sheth, J.N. (1969), The Theory of Buyer Behavior. Vol. 14. research revisited. Journal of Consumer Satisfaction, Dissatisfaction
New York: Wiley. and Complaining Behavior, 16, 53-66.
Johnsson, M., Lindblom, H., Platan, P. (2002), Behavioral Finance-and Warkentin, M., Gefen, D., Pavlou, P.A., Rose, G.M. (2002), Encouraging
the Change of Investor Behavior During and after the Speculative citizen adoption of e-government by building trust. Electronic
Bubble at the End of the 1990s. Markets, 12(3), 157-162.

International Journal of Economics and Financial Issues | Vol 7 • Issue 1 • 2017 277
 Trang and Tho: Perceived Risk, Investment Performance and Intentions in Emerging Stock Markets

APPENDIX

Appendix Table

GFI The research model Standard (Hair et al., 2014. p. 584)

Chi‑square (χ2)
Chi‑square 99.566 (P = 0.000) # 0, Significant P values expected
Degree of freedom 51
Absolute fit measures
GFI 0.965 Above 0.90
RMSEA 0.045 Values < 0.07 with CFI of 0.90 or higher
90% confidence interval of RMSEA 0.032‑0.058
RMR 0.019
SRMR 0.0212 0.08 or less (with CFI above 0.92)
Normal Chi‑square 1.952 <5
Incremental fit indices
NFI 0.979 Above 0.90
NNFI or the TLI 0.987 Above 0.90
CFI 0.99 Above 0.90
RFI 0.973 Above 0.90
Parsimony fit indices
AGFI 0.946 Above 0.90
PNFI 0.757 Above 0.747 (Shadfar and Malekmohammadi, 2013. p. 587)
GFI: Goodness‑of‑fit index, RMSEA: Root mean square error of approximate, RMR: Root mean square residual, SRMR: Standardized root mean residual, NFI: Normed fit index,
NNFI: Non‑normed fit index, TLI: Tucker‑Lewis index, CFI: Comparative fit index, RFI: Relative fit index, AGFI: Adjusted goodness‑of‑fit index, PNFI: Parsimony normed fit index

278 International Journal of Economics and Financial Issues | Vol 7 • Issue 1 • 2017