Customer Purchase Intention on Sharia Mutual Fund Products: a Tpb Approach

This study aimed to determine the influence of variables in TPB (attitude toward behavior, perceived behavioral control, subjective norm), religiosity, knowledge, and risk perception to intention purchase of sharia mutual fund. The sample in this study was customer of Bank Sharia XYZ of 164 customers. The analysis applied in this study was Structural Equation Modeling-Partial Least Square. The result of this research showed that the independent variable had positive and significant influence to dependent variables attitude toward behavior and subjective norm. Religiosity and knowledge have no significant influence toward intention to purchase, while perceived behavioral control and risk perception had no significant influence with intention to purchases. The other result showed that religiosity had a positive significant influence to attitude toward behavior. The conclusion of this research shown that religiosity is the factor which influence attitude toward behavior, attitude toward behavior and subjective norm are the factors which influence intention purchase of sharia mutual fund.


Introduction
Currently, investment products have been developing in Indonesia. One of the investment that is being sought and promoted intensively is Mutual Fund. Indonesian Government Regulation about Capital Market number 8 year 1995 mentions that mutual fund is an investment vehicle that used to collect money from many investors for the purpose of investing in securities, mutual funds operated by professional investment PDQDJHUV ZKR DOORFDWH WKH IXQG ¶V LQYHVWPHQW. Then Custodian Bank will administrates asset of the investors.
Based on its management, mutual funds are divided into two types, namely conventional funds and sharia mutual funds. Mutual fund is a type of investment that is being intensified by the Financial Services Authority (FSA). Each year, the asset of mutual fund is constantly growing, both for conventional mutual fund and sharia mutual fund, but contribution of sharia mutual fund lower than conventional mutual fund. In 2017, sharia mutual fund is represented by less than 4 percent of investors' Asset Under Management (AUM). It shows that AUM of sharia mutual fund is still low. This low intention to invest in Sharia Mutual Fund among the investor might be the reason of the low percentage of Sharia Mutual Fund AUM.
Bank Sharia XYZ sees the potential to improve its fee based income by selling Sharia Mutual Fund in its outlets. To increase the selling of mutual fund in Bank Sharia XYZ, a study should be conducted to analyze customers' intention to purchase in Sharia Mutual Fund. This research analyzed consumers' interest in purchasing Sharia Mutual Fund using TPB (Theory of Planned Behavior) as an approach and expected to provide a solution related to the problem.
TPB is an attitude model that predicts intention or consumers' intention to carry out a behaviour or an action (Sumarwan, 2014). Firmandhani (2016) conducted a study on factors that could influence micro community's intention in purchasing sharia micro insurance product; this research used TPB and added knowledge and religiosity as variables. Firmandhani (2016) stated that religiosity has a positive and significant effect on micro FRPPXQLW\ ¶V LQWHQWLRQ LQ SXUFKDVLQJ VKDULD LQVXUDQFH 5HVHDUFK RI $EX <RXVVHI et al.
Another study on intention was conducted by Lestari et al. (2017) analyzing factors that determined one's intention in having a credit card using theory of planned behaviour as approach. Whereas Amelia (2017) conducted a study that analyzed TPB variables (attitudes, subjective norms, and behavioral controls) as well as financial knowledge, risk tolerance, and retirement financial planning. Research of Lim, Soutar and Lee (2013) found that product knowledge has a significant influence on interest. Research of Fabrigar et al. (2006) explained that knowledge has an influence on individual attitude, this is EDVHG RQ LQGLYLGXDO ¶V DWWLWXGH WR FRQVXPH RU XVH WKH SURGXFW LI WKH\ Xnderstand and know the product. Haekal (2016) stated that risk perception has simultaneous effect on buying intention. Ashoer (2016) explained that risk perception can influence the formation of intention.
TPB views that a person's behavior can be predicted by his intention, which is predicted through attitude toward behavior, subjective norms, and perceived control behavior (Ajzen, 2015). In addition, there are three other variables that were also examined, namely religiosity, knowledge, and risk perception variables to examines their influence on the intention to purchase Sharia Mutual Funds.
The purpose of this study is to analyze the impact of customer knowledge and religiosity on attitudes toward behavior of purchasing sharia mutual funds. Analyzing the influence of attitudes toward behavior, subjective norms, and consumer perceived behavioral control in purchasing sharia mutual funds, analyzing the influence of knowledge, religiosity, and consumer risk perceptions in purchasing sharia mutual funds, formulate the managerial implications related to purchasing sharia mutual funds.

Methods
This research is a quantitative descriptive study using the survey method. Data collection in this study was carried out through interview techniques supported by questionnaires. The technique used was the purposive sampling technique. The criterion for the sample used in the study was customers of Bank Sharia XYZ with a minimum monthly income of IDR 10 million. This study was conducted for 3 months, from June 2018 to August 2018. Data collection was divided into two primary data was the result of the questionnaires distributed, and secondary data from the FSA, Central Bureau of Statistics, and other literature studies. Hair et al. (2010) explained that sampling can be based on the number of indicators used with the provision of 5-10 times the indicator used. Based on, the number of respondents in this study was 164 respondents.
This study used Structural Equation Modelling-Partial Least Square analysis, in which the primary data obtained from the respondents of the study were processed using PLS, a type of variance-based SEM created to overcome problems caused by covariancebased SEM. PLS can work efficiently with both small sample sizes and complex models (Wijanto, 2008).The variables in this study were based on the theory of planned behavior with the addition of several other variables. TPB is an attitude model that predicts consumers' interest in carrying out a behavior or action. The TPB model explains that the main factor that influences a person's behavior is his intention or tendency to take action (Sumarwan, 2014). The main focus on this theory is on the intention of individuals to carry out certain behaviors. TPB states that human behavior is first influenced by intention (behavior intentions). Intention will be influenced by three main factors, namely attitudes toward behavior, subjective norms, and perceived behavioral control (Ajzen, 2015). Other variables used in this study included knowledge, religiosity, and risk perception. The scale used for variable measurement was the Likert scale with a scale of 1 to 5. The research hypotheses are as follows: 1. Religiosity has a significant effect toward attitude toward behavior 2. Knowledge has a significant effect toward attitude toward behavior 3. Religiosity has a significant effect toward intention to purchase significantly 4. Attitude toward behavior has a significant effect toward intention to purchase 5. Knowledge has a significant effect toward intention to purchase 6. Subjective norms has a significant effect toward intention to purchase 7. Perceived behavioral has a significant effect toward intention to purchase 8. Risk perception has a significant effect toward intention to purchase

Respondents' Characteristics
The total number of the respondents of this research is 164. There are a number of criteria that can be used as base information on the respondents. Table 1 presents data of the respondent. Table 1 shows important information is obtained; the majority of respondents were male with a percentage of 53 percent. Female respondents were 47 percent of the respondents. Respondents who were 25 to more than 45 years old is 84 percent, i.e., most of the respondents were in their productive age. The majority of respondents were 86.21 percent married. Based on respondentV ¶ HGXFDWLRQDO EDFNJURXQG 75 percent of the respondents hold a bachelor degree, 22 percent master/doctoral degree, 2 percent diploma, ad 1 percent high school graduated. The majority of the respondents hold a bachelor degree. This means that these respondents are categorized as well educated. Revenue is an important one that can provide information related to funds that can be used as part of investment activities. It is known that the majority of respondents have income between IDR 10 million and IDR 20 million which represent 74 percent. In addition, with the same proportion, respectively 13 percent has an income between IDR 20 Million -IDR 30 Million and also more than IDR 30 Million. The majority of respondents are in the mass affluent segment (1 year income > IDR 100 million), which means that the respondents can already set aside part of their income to invest. Based on the data, it can be known that 46.8 percent of the respondents are Javanese. Related to employees' occupation, the majority are private sector employees represented by 90.15 percent, which means that the majority of respondents receive regular income every month.

Modeling Purchase Intention on Sharia Mutual Fund Products
Structural Equational Modeling-Partial Least Square (SEM-PLS) is a tool to analyze the relation between variables. In this research SEM-PLS model was created consisted of seven variables and thirty two indicators. Independent variables in this research were intention (I) consisting of three indicators, while dependent variables in this research were attitude toward behavior (S) consisting of four indicators, subjective norms (N) consisting of six indicators, perceived behavior control (P) consisting of four indicators, knowledge (MAL) consisting of one indicator, religiosity (R) consisting of seven indicators, and risk perception (PR) consisting of seven indicators.

Validity Test (Convergent Validity)
Validity test was carried out using convergent validity. This was done to verify whether the indicators used were appropriation representing each variable. Five variables are determined as not valid. This is due to the output value which does not exceed the prescribed limits, as for P3, P4, PR2, PR3 and PR4. Thus, these five indicators should be taken out from PLS model to be studied. A retest to ensure that no variable is invalid in the model was conducted therefore the remaining 27 indicators can be used for further test, represented in the Table 2

Reliability Test
A reliability test was conducted for each variable in the model of intention to purchase sharia mutual fund. It was intended to test the reliability of each variable used in the research model. Three parameters can be used as base to determine the reliability in PLS test. The result is presented in Table 3. Table 3 shows that by using three parameters, namely cronbach alpha (CA > 0.7), Rho A (rho_A>0.6) a composite reliability (cr > 0.7) all variables have been declared reliable. Thus, they are reliable to describe a research model. On the other hand, reliability can describe the ability to provide the same results at all times.

R Square Research Model
R Square is the proportion of the variance in the dependent variable that is predictable from the independent variable. From seven variables, there were two variables that were considered to represent the model. In addition, R Square was used to describe a model designed by combining several variables that can describe the actual conditions. The output is presented in Table 4. Based on the Table 4, it is known that the value of R Square attitudes toward behavior is 0.175 or 17.5 percent which shows that the model is only able to be described 17.5 percent from variables in the model as 82.5 percent by the variables outside the model. On the other hand, the intention model is described 47.6 percent from variables in the model and 52.4 percent from the variables outside the model. While Figure 1 shows SEM-PLS model yielded from the analysis.

Figure 1. SEM-PLS Model Loading Factor of Variables
From the SEM-PLS, an illustration can be obtained to describe the correlation between variables to variables that can explain the hypothesis built in the research. The output can be seen in Table 5. Table 5 shows that the SEM PLS model has correlation values and new significance showing the best description to answer the research formulation. The result of the first hypothesis testing showed the influence of religiosity to attitudes toward behavior gave a positive (estimate of 0.415) and significant (t-value> 1.96) influence. This indicates that religiosity plays a strong impact in forming a person's attitude to invest in sharia mutual funds. Sharia mutual fund has religious values that in general it has value added and in particular it has its point of interest. The research of Youssef et al. (2015) found out that religiosity affected customers' attitude toward Islamic Banking in Egypt. On this basis, Bank Sharia XYZ needs to improve its approach through religion intended to influence attitudes first.
The result of the second hypothesis testing showed the influence between Knowledge and Attitude toward Behavior; the influence is not significant (1.768 < 1.96), thus hypothesis is rejected. Fabrigar et al. (2006) stated that knowledge is one needs to have to behave; if he has good knowledge, his attitude will be supporting, on the contrary, if his knowledge is not sufficient, it will result in the contrary. The result of the third hypothesis testing showed the influence of Religiosity to Intention is not significant since t value is 0.455 (t-value significant at > 1.96); thus the hypothesis is rejected. The result of this research is in line with the research of Amin et al. (2011) stating that religiosity has negative influence and not significant with the intention to use Islamic Personal Financing. Religiosity is one's principal values implemented each time with a series of ritual religious activities in his daily life. Each person has different religious values. The research of Shabrina (2018) also showed that religiosity has an insignificant influence to intention to use. The result of this research is not in line with that of Firmandhani (2016) stating that religiosity has significant impact on micro community's intention in purchasing sharia insurance.
The result of the forth hypothesis testing showed that the influence of Attitude toward behavior to intention to purchase is positive proven by the value estimate amount to 0.318. In addition, the influence of attitude toward behavior to intention to purchase proven to be significant (t-value 3.454 > 1.96). Thus, the hypothesis is accepted. Seba et al. (2012) explained that attitude can influence one's intention; good attitude will take one to show good intention. Islamoglu et al. (2015) described that investors' attitude can influence intention to invest, but this intention should be influenced prior to intention to give good impression.
The result of the fifth hypothesis testing shows that the influence of knowledge to intention to purchase is not significant, where t-value value is below 1.96; thus, the hypothesis is rejected. The result of this research is in line with the study of Sujiwo (2015) on Moslem respondent who had conventional insurance, knowledge does not have significant impact on the intention and Cuccinelli et al. (2016) stated that financial literacy level does not have significant impact on the intention to purchase medium/high risk financial investment. The results of hypothesis testing contradict Njuguna (2017) which described that the relationship between investors' knowledge greatly influenced intention to invest, good knowledge led to the formation of intention. The statement is in line with Hartoyo et al. (2013) stating that intention to make retirement financial plan increases with students' subjective knowledge. Beik (2011) stated that public education and socialization is a fundamental matter considering very low public awareness on sharia capital market products. Based on this finding, companies should promote more their product to the public to increase knowledge, not only related to the difference between sharia and conventional mutual fund, but also to mutual fund itself (NAV values, type of mutual fund, etc.) The result of the sixth hypothesis testing shows that the influence between subjective norm and intention to purchase is positive (estimate 0.382) and significant (tvalue 5.077 > 1.96); thus the hypothesis is accepted. This shows that norm has an important role in intention to invest in sharia mutual fund. It shows that investors are willing to invest after receiving recommendation whether from a friend, family and his surrounding; on the other hand, the impact is greater when a friend, family or his surrounding also invest in sharia mutual fund. The result is in line with the research of Lestari et al. (2017) stating that subjective norm affected on the intention to possess a credit card. Alam (2011) andNugroho et al. (2018) also showed that subjective norm had significant influence on one's intention.
The result of the seventh hypothesis testing shows that the influence between perceived behavioral control and intention is not significant since t-value value does not exceed 1.96; thus the hypothesis is rejected. In this research, customers had already known an overview on sharia mutual fund product, but it did not greatly influence the intention to purchase. The result of the eighth hypothesis shows that the influence between risk perception and intention is not significant since t-value is below 1.96; thus, the hypothesis is rejected. This is in line with the finding of Rosyidah and Lestari (2013) stating that risk perception provide negative impacts; investors tend to withdraw if the risk is too big. While Trang dan Tho (2017) stated that risk perception had small influence on the intention to invest, investors know already the risk to be faced and the amount of return that can be obtained.

Managerial Implication
The implication that was obtained from this research is Bank Sharia XYZ needs to sell Sharia Mutual Fund products that are believed to give good return and are issued by a good and trusted Investment Manager. This is to create an impression and recommendation from friends, family and environment. Bank Sharia XYZ needs to create a program such as member get member program, considering investor has the intention to purchase in Sharia Mutual Fund if he receives recommendation from friends, family and environment.
Knowledge on sharia mutual fund in this research is relatively high, but in this research, the impact of knowledge on the intention to purchase sharia mutual fund is not significant. Therefore, periodic evaluation on Sharia Mutual Fund that can be sold by Bank Sharia XYZ should be carried out. The profile and the track record of Investment Manager is very important to determine whether a product to be sold is conform or not to investors' expectation. This is also in line with attitude toward behavior variable that affects significantly intention to purchase where the strongest indicator is Sharia Mutual Fund is worth to be purchased because it gives good return.
In addition to basic product knowledge, Bank Syariah XYZ is also needs to socialize other knowledge in more deeply; for example for arguments stating the availability of Sharia Mutual Funds, and the obligation to invest in institutions that are in accordance with Islamic sharia. Whereas for employees, the management of Bank Syariah XYZ shall ensure that marketing officers does not make any mistake in explaining sharia aspects distinguishing sharia from conventional. Mistakes in delivering can have fatal consequences for customers; customers may feel that there is no difference between conventional and sharia Increasing competency is very important to improve customers' trustworthiness toward sharia financial industry.
The other implication is Bank Sharia XYZ needs to focus on customers active in religious ritual, since it is the real attitude of religiousness. On the other hand, companies shall underline the community on investors and future customers related on the use of halal investment product. Religiosity gives strong influence on attitude which in turn provides strong influence in intention. This can be a recommendation for companies to improve intention to purchase Sharia Mutual Fund.

Conclusion
Based on the results of the study, knowledge has an insignificant influence to attitude toward behavior. On the other hand, religiosity has a significant influence to attitude toward behavior. Attitude toward behavior and subjective norm have a significant influence to the intention to purchase sharia mutual fund. On the other hand, the influence of religiosity, knowledge, and risk perception to intention to purchase sharia mutual fund is insignificant.

Recommendation
This research provides important recommendations that can be conducted by several parties, such as to increase the sales of Sharia Mutual Fund, Bank Syariah XYZ can focus on subjective norms and attitude toward behaviour, and to ensure the impact of knowledge on intention to purchase sharia mutual fund, further research should be conducted using knowledge in differentiating community that has received education on Sharia Mutual Fund Product and the one who has not. In addition, further research is needed by using other variable (i.e., Investment Managers).