Determinants of Muslim Millennials' Intention Towards Savings Sukuk

. Savings Sukuk is one of the sharia investment products issued by the Indonesian Government. Investors from the millennials generation always dominate Savings Sukuk sales, although it is a new investment instrument in Indonesia. Therefore, we interest in examining the intention towards Savings Sukuk more deeply. In particular, this study analyzes the effect of religiosity, product knowledge, and pricing on Muslim millennials' intention toward Savings Sukuk in Indonesia. This study used a quantitative approach, and the data collected was analyzed by Partial Least Square – Structural Equation Modeling (PLS-SEM). Purposive sampling was used to get the required sample. Respondents in this study were Muslim millennials in Indonesia and were not limited by region. There are at least 138 respondents who participated in the survey. The results showed that product knowledge and pricing influenced Muslim millennials' intention toward Savings Sukuk. In contrast, religiosity does not affect Muslim millennials' intention toward Savings Sukuk in Indonesia.


INTRODUCTION
Sukuk which is also known as the Islamic bond is one of the Islamic investment instruments. Based on sharia standards number 17 of 2009, The Accounting and Auditing Organization of Islamic Financial Institution (AAOIFI) defines sukuk (plural of Sak) are certificates of equal value representing undivided shares in ownership of tangible assets, usufruct, and services or (in ownership of) the assets of particular projects or special investment activity. The Malaysian Government is the first to issue sukuk in 2002 with a US$ 600 million value, followed by the Solidarity Trust Sukuk's issuance with $ 400 million in September 2003 by the Islamic Development Bank (Ayub, 2009). Then, sukuk has become popular as a financing instrument for corporations and the Government.
In Indonesia's Islamic finance industry, PT Indosat issued the first sukuk for Rp175 billion in 2002 (Nasution et al., 2012). Sukuk with mudharabah (profit-and-loss sharing) scheme issued by PT Indosat received tremendous response from capital market players so that many other companies Savings Sukuk (Sukuk Tabungan) is a variant of retail sukuk issued by the Indonesian Government. The first Savings Sukuk (ST-001) was issued in 2016 and earned an investment return of IDR 2.6 trillion (Press Release of Indonesia Ministry of Finance, 2016). According to DJPPR Indonesia (2018), the Government targeted IDR 1.7 trillion for the second Savings Sukuk (ST-002). In the issuance of ST-002, the Government lowered the minimum order price from IDR 2 million to only IDR 1 million. As a result, the sales of ST-002 increased to IDR 4.9 trillion (Press Release of DJPPR Indonesia Ministry of Finance, 2018). The oversubscribed sales results show the high enthusiasm of the public to invest in Savings Sukuk. 2019 has been a productive year for the Government to issue 141 AL-MUZARA' AH Vol. 10  Savings Sukuk up to the sixth series. In 2020, the Government issued another seventh Savings Sukuk  at the end of the year.
One thing that catches our attention when discussing this instrument is because we see investors' dominant composition. Press Release of DJPPR Indonesia Ministry of Finance regarding the first to sixth Savings Sukuk acquisition shows that most investors are millennials (born in 1980 -2000).

Figure 2 Millennials participation rate in Savings Sukuk
In ST-001, millennials' participation was only 26%, but at the issuance of ST It follows the Government's goal to encourage millennial investors to become smart investors while caring about the environment because ST-007 is green sukuk sustainable financing. It can be concluded that the Savings Sukuk, among the many investment instruments, is one that millennials are interested in. At the same time, Savings Sukuk is a new investment instrument in Indonesia. Therefore, the authors feel the need to examine the intention towards Savings Sukuk more deeply.
Studies on intention in sukuk investment are not the first. We find some of them, such as research conducted by Warsame and Ireri (2016) and Ashidiqi and Arundina (2017), using the theory of planned behavior to measure investment decisions in sukuk instruments. Awn and Azam (2020) also use a theory of planned behavior to measure Libyan investors' attitudes toward investing in sukuk. Meanwhile, Yuliati (2011) found that the risk of investment and Islamic attributes of the product positively affect public intention to invest in sukuk. Firdaus et al. (2018) (Abduh and Hussin, 2018).
However, studies on this topic are still limited, especially in the case of Indonesia. Sukuk has become a critical Islamic financial investment instrument for advancing the Islamic financial industry in Indonesia, so it still needs to be explored deeper. This study will mainly review the Muslim millennials' intention of Savings Sukuk and its determinants. Peter and Olson (2017) stated buying intention is a prediction of purchase behavior in the future. Similar to that expressed by Ajzen and Fishbein in the Theory of Planned Behavior (TPB). The intention is the desire to do something in the future (Ajzen and Fishbein, 2011). TPB is a theory that is often used in measuring consumer intention. Likewise, to measure intention on Islamic financial products such as in the research of Warsame and Ireri (2016), Ashidiqi and Arundina (2017) Awn and Azam (2020). To measure Muslim millennials' intention of Savings Sukuk, we used the concept of intention expressed in the TPB.

LITERATURE REVIEW
Furthermore, as a sharia financial instrument, sukuk emphasizes the value of Islam in its use. Likely, investors who are interested to invest in Savings Sukuk are mainly Muslims. According to Huda et al. (2017), the sukuk market's basis is sharia-minded and conventional investors, while bonds are only for conventional investors. Sharia-minded investors are investors who prioritize Islamic financial products rather than conventional ones for sharia reasons. It because they decided something based on the rules of their religion. Therefore, religiosity can undoubtedly be a factor that encourages investors to invest in Savings Sukuk. In religion, some values are believed by its adherents and are always attached to its adherents. Values will become teachings that can be used as guidelines or basic teachings in living life. Therefore, every religious community will view things as influenced by their religious views. Including if Muslim consumers buy a product will consider buying a product if they are assured that it does not contradict their sacred beliefs (Sharma et al., 2017). Souiden and Rani (2015) used the term religiosity to indicate individuals' respect for their professing religion. According to Glock and Stark (1968) in a book written by Ancok and Suroso (2011), religiosity has five dimensions; belief (aqidah), practice ('amal), knowledge (ma'rifah), experience (ihsan), and consequence (natijah) (Sharma et al., 2017). Another measurement scale of religiosity was proposed by a team of researchers supported by the Fetzer Institute and the National Institute on Aging (NIA) Working Group (1999) (Souiden and Rani, 2015). They detail the 12 dimensions of religiosity; daily spiritual experiences, meaning, values, beliefs, forgiveness, private religious practices, religious/spiritual coping, religious support, religious/spiritual history, commitment, organizational religiousness, and religious preferences (Souiden and Rani, 2015). The dimensions of religiosity above can be used to assess a person's spiritual level. Sharma et al. (2017) explained how Muslims' religiosity dimensions support purchase intention of Islamic financial products in Bangladesh. They explained that devout consumers place greater importance than less religious consumers on Shariah-compliant financial products. More specifically, Newaz et al. (2016) stated that religiosity was influenced positively toward the intention to Islamic financial products for insurance and capital market products. Bananuka et al. (2020) indicate that religiosity is significantly associated with intention to adopting Islamic financing. There is a positive influence of religiosity on Muslim students' intention with and without Islamic economics educational backgrounds to save in Islamic banks (Nugraheni and Widyani, 2020). Several other studies have also found that religiosity positively influences intention (Fauzi et al. 2020;Kristiyadi and Hartiyah, 2016;Zahri and Hafasnuddin, 2016;Abduh and Hussin, 2018 Product knowledge becomes an essential part of this research because Islamic products have different unique characteristics. Islamic products and services must comply with sharia standards such as halal and haram regulations, which are legalized by fatwas. The fatwa issued refers to the main guidelines for Muslims, namely the Quran and Sunnah. So, if the investor knew about the unique characteristics of Islamic financial products, they will perceive this product differently and especially. Therefore, apart from religiosity, product knowledge can influence buying intention in Islamic financial products. Besides, product knowledge is essential for purchase intention and purchasing decisions. Because when consumers have a lot of knowledge about a particular product, it will be a reference in purchasing decisions (Nurhayati and Hendar, 2020), especially if consumers' knowledge is positive about the product. Awn and Azam (2020) said the main factors that can increase sukuk and Islamic financial assets' popularity were the information availability, issuance of sukuk, and source features for investors.
Product knowledge collects various kinds of information that consumers have about the product physically and abstractly (Nurhayati dan Hendar, 2020). It involves knowledge about product categories, brands, product attributes, prices, benefits of the product, and satisfaction. Hierarchically, consumers have product knowledge levels; product attributes knowledge, product consequences knowledge, and product value knowledge (Peter and Olson, 2017). Product attributes knowledge is the consumer's knowledge of product characteristics, including actual and abstract attributes. Product consequences knowledge is knowledge about the results that will be felt after the product is purchased. Product Consequences Knowledge is divided into two, functional consequences and psychosocial consequences. Meanwhile, product value knowledge involves emotional affection related to the needs and goals of life in depth.
So, this study's second factor in measuring millennials' Muslim intention on Savings Sukuk is product knowledge. Refers to Johan et al. (2020), there is a positive relationship between knowledge and intention to hold sharia-compliant credit cards for Muslim customers. Knowledge also became a significant variable in influencing intention to use Islamic banks among the Muslim community (Wahyuni, 2012). Meanwhile, Nugraheni and Widyani's research (2020) found a significant positive influence of knowledge on Muslim students' intention with Islamic economics educational backgrounds to save in Islamic banks. Meanwhile, Nurhayati and Hendar (2020) use variable product knowledge to test halal products' intention. The results show a positive relationship between halal product knowledge and halal product intention. Sriminarti and Nora (2018) also did the same, and the results showed that product knowledge has significant and positive value on purchase intention.
Lastly, we discuss about pricing. Pricing refers to the act of selecting the most appropriate price to sell the financing product and striking a balance between the bank and the customers' intention (Ebert and Griffin, 1998). As such, the price of sukuk refers to selecting the best practice of pricing to sell the financing product by taking into consideration an issue of sukuk profit generation and a customer's welfare. The consumer's perception of the product's value becomes the upper limit of the price (Kotler and Armstrong, 2018). The price limit is the fair limit of the price paid. Consumers will not buy if consumers consider the price of a product to be greater than the benefits obtained.
Kotler and Armstrong (2018) explain four price indicators: Affordability, the suitability of the price with the quality of the product, price competitiveness, and the suitability of the price of benefit. Islamic financial products have a different method of pricing. Islamic financial institutions use profit and loss sharing in charging for financial transaction services. Therefore, the pricing and price of Islamic financial products are different from conventional finance. That is why pricing will affect customer intention in Islamic financial products. Previous research found that pricing provided a significant influence on intention to use sharia microfinancing (Shabrina et al., 2018). Sharfia (2019) also found that price has a positive effect on intention. Meanwhile, Amin et al. (2010) argue that the price of financing was also the key determinant for financing acceptance, in this case for Islamic financial products Qardhul Hasan.

METHOD
This study uses a quantitative approach that aims to test the hypothesis. The data used comes from primary and secondary sources. Secondary data comes from various literature, such as research articles, books, and official websites. It is needed to summarize the theory to form a conceptual framework. Primary data is necessary to empirical test using Partial Least Square -Structural Equation Modeling (PLS-SEM). It was collected from respondents through an online survey. The sample in the study amounted to 138. In the PLS-SEM, the minimum sample size should be five to ten total indicators (Hair et al., 2017). Consider the total indicators that we used are 27 items (see Model Development). Therefore, the minimum sample size is 135 to 270. However, PLS-SEM allowed using less sample. Bootstrapping method used in PLS-SEM makes no assumptions about the shape of the variables' distribution or the sampling distribution of the statistics (Hair et al., 2017). So, it can be applied to small sample sizes with more confidence.
Purposive sampling uses to ensure the netted sample follows the research needs. The sample criteria are Muslim millennials Indonesian citizens. Millennials are identified as the generation born between 1980and 2000(DeBard, 2004. The questionnaire was used as a measurement tool with a five-point Likert scale ranging from "strongly disagree" to "strongly agree." Then, the measurement items in this study developed from theory and previous researches. Three The data collected were analyzed by Partial Least Square -Structural Equation Modeling (PLS-SEM) using smart PLS software 3.2.9. PLS-SEM is one type of structural equation modeling based on variance. PLS-SEM is evolving as a statistical modeling technique and is primarily used to develop theories in exploratory research (Hair et al., 2017). It is used because it can be to predict and explain target constructs. Hair et al. (2017) stated that PLS-SEM also works efficiently with small sample sizes and complex models because makes practically makes no assumptions about the underlying data. Another advantage over other SEM methods is that PLS-SEM has high efficiency in parameter estimation, manifested in the method's greater statistical power (Hair et al., 2017 The PLS-SEM method has two steps; the first step is to analyze the outer model, also called the measurement model. The outer model describes the relationships between the latent variables and their indicators. The second is an analysis of the inner model, also called the structural model. The inner model describes the latent variables' relationships (Hair et al., 2017;Garson, 2016).

Descriptive Analysis
This study analyzes the results of a survey filled out by 138 Muslim millennials respondents with the following characteristics:

Measurement Model
The first step of PLS-SEM analysis is to test the outer model or measurement model as a validity and reliability test. The validity test with PLS-SEM is divided into two: convergent validity and discriminant validity. The convergent validity measure correlates positively with alternative measures of the same construct by evaluating outer loading values and Average Variance Extracted (AVE). The outer loading values must be more than 0.70. It should consider indicators with outer loadings between 0.40 and 0.70 to remove the scale because it is not loaded in the construct (Hair et al., 2017;Garson, 2016). However, according to Hair et al. (2017), weaker outer loading does not need to remove if it does not affect the AVE score (> 0.50). The result of convergent validity is under the threshold and affects the AVE score of Religiosity (0.350) (see Table 2). Therefore, to continue the analysis with PLS-SEM, we removed measurement items with low outer loading values, namely R1, R2, R3, R4, and R5.

Figure 3 Final research model
After removing measurement items with low outer loading values and reanalyzing them, we got an AVE value above 0.5 for all variables (Table 3). The AVE values as follows: Intention: 0.845, Pricing: 0.546, Product Knowledge: 0.701, Religiosity: 0.522. Then the analysis was continued by evaluating the cross-loadings and Fornell-Larcker criterion for discriminant validity assessment. The role of thumb on the PLS-SEM, all indicators' outer loading on the associated construct, is must higher than any of its cross-loadings on other constructs (Hair et al., 2017;Garson, 2016).  The Fornell-Larcker criterion is the second approach to assessing discriminant validity. In the discriminant validity test using the Fornell-Larcker criterion, AVE's square root should be higher than its correlation with any other latent variable (Hair et al., 2017;Garson, 2016). On the Fornell-Larcker criterion (Table 5), AVE's square root appears in the diagonal cells, and correlations appear below it. As can be seen, all the square roots of AVE are higher than its correlation with the values of 0.919 (Intention), 0.739 (Pricing), 0.838 (Product knowledge), 0.723 (Religiosity). composite reliability values are shown in Table 3. All variables are well with values greater than 0.70. It explains that all statements of items that are valid and reliable have also met the criteria.

Structural Model
Firstly, we do the collinearity test to examine each predictor construct separately for each structural model subpart. According to Hair et al. (2017) and Garson (2016), VIF values above 0.2 and below 5.0 are needed to prove no multicollinearity in the model. And we get sufficient VIF values on the outer and inner models. So, we can move on to the inner model analysis. Inner model is done to see construct's accuracy and hypothesis.
Construct's accuracy or model fit is analyzed from the resulting R2 (R-square) value. The R2 values show inside the ellipses in figure 5. R2 or also called the coefficient of determinant is 0.476. The independent variable explains the dependent variable by 47.6%, and other variables outside the model explain the rest. Following the role of thumb, an R2 value of 0.476 can be considered moderate (Hair et al., 2017). It is mean the model has a medium predictive power to calculated the correlation between X variables and Y variable.
Then we check the F2 value (F-square) to find out the size of the effect. The F2 value of pricing of 0.80 means that pricing has a large impact on intention. In contrast, product knowledge has a small effect on the intention, with an F2 value of 0.148. Religiosity is known to not affect with an F2 value of 0.00. The last and most important of all the test series is testing the hypothesis. Hypothesis test using the path coefficient, t-value, and p-value. Garson (2016) noted that the path coefficient is between -1 to +1, and weights closest to absolute 1 reflect the positive most robust paths. On the contrary, weights relative to 0 reflect the weakest paths. The path weight of 0.015 shows religiosity has a positive effect on Intention but weak. Meanwhile, pricing and product knowledge have a medium positive effect on Intention.
Furthermore, t-statistics' value must be greater than the t-table, and the t-table value for the two-tailed model is 1.96 (Hair et al., 2017;Garson, 2016). Suppose the result meets the criteria >1.96. In that case, it means that there is a significant influence between the X variable on the Y variable. But if it does not meet the criteria or <1.96, it does not impact the Y variable on the X variable. While the conclusion of the hypothesis based on the results of p-value is that X variable affects Y variable if pvalue <0.05. As shown in Table 7, the decision to test the hypothesis of this study is as follows: 1. H1 is rejected because t-value for religiosity is 0.247 < 1.96 and p-value is 0.805 > 0.05, so religiosity has positive paths but does not have a significant effect on Muslim millennial intention toward Savings Sukuk. 2. H2 is accepted because t-value product knowledge 3.638 > 1.96 and p-Value 0.00 < 0.05, then product knowledge has a positive and significant impact on the Muslim millennial intention toward Savings Sukuk. 3. H3 is accepted because t-value pricing is 2.750 > 1.96 and p-value 0.006 < 0.05 then pricing has a positive and significant impact on Muslim millennial intention toward Savings Sukuk.

Discussion
It is interesting to conclude that religiosity does not significantly affect Muslim millennials' intention toward Savings Sukuk. However, sukuk is one of the Islamic investment instruments. That means Muslim religiosity is not the main factor that will support Muslim intention in Islamic products, especially for the millennial generation. Therefore, this finding contradicts Sharma et al. (2017) 's research which found that religiosity supports the purchase intention of Islamic financial products in Bangladesh. Also contrary to many other studies including Newaz et al., (2016), Bananuka et al., (2020), Nugraheni and Widyani (2020), Fauzi et al. (2020), Kristiyadi and Hartiyah, (2016), Zahri and Hafasnuddin (2016), Abduh and Hussin (2018). However, this conclusion is not the first because Johan et al. (2020) also revealed that religiosity does not support Muslims' intention to hold shariacompliant credit cards. Shabrina et al. (2018) also stated that a religious factor is not a factor that can affect the intention to use sharia micro-financing.
In this research, we concluded Muslim religiosity does not affect intention. It happens not because Muslim millennials in Indonesia have low religiosity. The survey results show respondents have a high average value of religiosity on a scale of 4.64 out of 5. Therefore, other factors can cause religiosity to have no significant effect on interest. We analyze that this can happen because the issuer of investment products has not been maximized in promoting its religiosity. Such as emphasizing riba-free and other sharia standards. So, the promotion of Savings Sukuk has not attracted shariaminded investors. Another reason, according to Souiden and Rani (2015) likewise, some Muslims might not believe in spiritual matters when dealing with business and financial transactions.
Meanwhile, we found that product knowledge has a positive effect on intention. Also, the results show that there is a strong medium relationship between product knowledge and intention. Therefore, we conclude that product knowledge can be a determinant of Muslim millennial intention toward Savings Sukuk. It also means that the greater the knowledge of millennial Muslims about Savings Sukuk, the greater their interest. Nurhayati and Hendar (2019) concluded that consumer knowledge about products would affect the information search behavior and influence the purchasing decisionmaking process, including purchase interest. This result is in line with Johan et al. (2020), Nugraheni and Widyani (2020), Kristiyadi and Hartiyah (2016), and Wahyuni (2012).
On the other hand, we also found that Muslim millennial's knowledge of Islamic investment products still lacked even though the hypothesis testing showed significant results. The questionnaire results show that the average respondent's knowledge about Savings Sukuk products is 3.83 out of 5. It should be a concern of issuers because if the public's understanding of Savings Sukuk products increases, it will also increase public interest.
Finally, H3 is also accepted. It's an indication that product pricing significantly influences Muslim millennials' intention on Savings Sukuk. The result means the welfare pricing of Savings Sukuk builds Muslim millennial intention towards Savings Sukuk. This empirical study proves that fixing a fair price is very important to make customer intention in products or services. In this research, the fair price referred to is stated by Kotler and Armstrong (2018) and is used as a measurement indicator, namely, affordability, the suitability of the price with quality of the product, price competitiveness, and the suitability of the product the price of benefit. This finding also strengthens the previous research conducted by Shabrina et al. (2018), Lajuni et al. (2017) and Amin et al. (2011).
Besides, we indicate that the millennial generation's intention to invest in Savings Sukuk increased with the second issuance when the offering price was lowered. According to Kotler and Armstrong (2018), price adjustment is part of a marketing strategy that influences buying interest. That is why we think that the beginning of the increasing purchasing power of the millennial generation for Savings Sukuk was due to a more affordable price.

CONCLUSION Conclusion
Finally, we got a conclusion based on the survey results and analysis that had been done. Conclusions refer to the research objectives, as well as the proposed models and hypotheses. The empirical study results conclude that product knowledge and pricing significantly affect Muslim millennials' intention toward Savings Sukuk in Indonesia. Meanwhile, religiosity does not work on Muslim millennials' intention toward Savings Sukuk in Indonesia.
Then, we found some additional findings. The first prospective Savings Sukuk investors, especially Muslim millennial still have limited knowledge about sukuk even though product knowledge affects their intention. The second Muslims millennial who have sharia-minded were not yet interested in sukuk products, especially Savings Sukuk.

Recommendation
Based on the findings of this study, we have recommendations for practitioners, regulators, and academics. We found that respondents still have limited knowledge about sukuk even though product knowledge affects their intention. Therefore, practitioners and regulators, especially sukuk issuers or the Government as issuers of Savings Sukuk should further increase consumer product knowledge about sukuk. It can be done through promotion and education such as seminars, discussion even though it holds virtually. We also recommend sukuk issuers carry out promotions and education through digital media such as social media to attract more potential millennial investors. Postings on social media can also increase consumer product knowledge about Savings Sukuk because we know that the millennials very familiar with social media.
We also found that sharia-minded investors were not yet interested in sukuk products, especially Savings Sukuk. It can happen because the issuer has not maximally promoted the sukuk from the sharia side. Hence, it is highly recommended for issuers to increase promotion from the sharia side to attract sharia-minded investors. Issuers can cooperate with the Council of Indonesian Ulama (Majelis Ulama Indonesia/MUI) or other competent institutions to promote savings sukuk by explicitly raising the shariah side. Of course, the institution is more trusted and gets the attention of the Muslim community.
The research topic about intention in sukuk is still interesting to continue because sukuk instrument continues to develop, especially in Indonesia. So, for further research recommendation, academics can add another independent variable related to the dependent variable, both qualitative and quantitative variables such as age and income. It is considered because, generally, investment intention is related to a person's income. Regarding the use of the religiosity variable, the data is also studied more deeply, for example, by changing indicators obtained from other theories. Because many studies show that religiosity's variable influences purchase intention, this is still a gap that can be investigated.