Keywords

Sukuk provides an alternative source of funding, especially for large-scale projects and investments. Here, differences in Sukuk financing modes in comparison to conventional instruments are considered regarding the extent to which risk levels yield significant similarities. The investors of Sukuk possess undivided beneficial ownership of the underlying assets, which entitles the certificate holders to shares in revenues generated from them Afshar (2013). Whereas, conventional bonds represent a contractual debt obligation in which the issuer is obliged to pay interest to the debt holder at an appointed date.

The first convertible Sukuk was issued in Dubai in 2007, in the same year, Malaysia's Khazanah National issued exchangeable Sukuk with an option to exchange them for the existing shares of the originator's subsidiaries. These issues attracted high interest both from investors as well as potential issuers of Sukuk, due to a narrative of risk reduction alternatives. Financial experts discussed the possibility of further innovation in Sukuk, such as contingent convertible Sukuk and reserve convertible Sukuk. Nevertheless, most of Shari’ah scholars have restricted these kinds of innovations due to their similarity with derivatives and excessive uncertainty (Al-Sayed, 2013).

Since the zero-coupon Sukuk is not tradeable in an Islamic secondary market, Sukuk investors hedge risk exposures applying embedded options by converting the Istisna Sukuk into real assets or usufructs. Investors are not forced to wait for the Sukuk's maturity before converting the Sukuk into cash. Whereas, investors avoid exposing the following market risks, such as, re-investment, credit, and interest-rate risks, particularly, where market volatility exists.

This study aims at empirically analysing the risk levels of sukuk portfolio compared to conventional bonds, using a sample data that is extracted from daily closing prices of Sukuk and bonds that are traded in Nasdaq Dubai’s security market. In doing so, this study employs Value at Risk (VaR) method to analyse and compare the risk level of Sukuk and conventional bonds portfolio.

This paper contributes to the literature on comparative analysis of Sukuk and conventional bonds by elaborating on their risk elements and challenges of Sukuk and bonds in the light of risk management. In the light of available large literature, this paper provides the insight to riskiness of Sukuk for potential investors as an alternative financial tool and also aims to bridge the gap in Sukuk risk analysis compared to conventional bonds.

The rest of the paper is organised as follows: the Sect. 17.2 presents a literature review to contextualise the subject matter, followed by the previous studies about the implementation of VaR models in portfolio risk measurements. The following section highlights the method of the research and data analysis, followed by VaR model application to Sukuk and conventional bonds and discussion on the empirical results. Finally, conclusion presents the summary of the findings and emphasizes recommendations for future research on the same characteristic studies.

1 Data and Method

Quantitative research methodology is applied, creating a statistical model analysing data through the Value at Risk method. A case study on United Arab Emirates’ Sukuk and conventional bonds issuances listed on Nasdaq Dubai's financial market is analysed. Secondary data is obtained from listed Sukuk and conventional bonds issuance in Nasdaq Dubai's financial market. The data details daily closing prices of Sukuk and conventional bonds for a period of one-year from 18th January 2016 to 17th January 2017. Relevant data used for analysis, the Sukuk and conventional bonds issuance satisfies the following two criteria: firstly, data excludes any Sukuk and bonds issuances traded in the market for less than one year. Secondly, it only considers a maturity period of 3–10 years from the listed Sukuk and conventional bonds issuances. Sale based Sukuk were purposly excluded from the analysis, as they have very short maturity, and therefore they does not align for the comparitive analysis with the bonds. The sample data of 30 (15 Sukuk and 15 bonds) Sukuk and conventional bonds are selected randomly from a listed 70 Sukuk and conventional bonds that have satisfied those criteria. For example, there were only 15 samples of traditional bonds which satisfied the above-mentioned criteria. Therefore, we selected equal samples of 15 from each site. The lists of Sukuk and conventional bonds and their descriptions are illustrated in Tables 17.1 and 17.2.

Table 17.1 List of Sukuk instruments
Full size table
Table 17.2 List of conventional bonds
Full size table

As demonstrated in the above Table 17.1, the range of Sukuk issue size differs between the amount of USD 178.1 million to USD 1.5 billion, while the total combined amount of Sukuk issuance is USD 12.38 billion. The Ijarah finance model was the most popular mode of financing a total of USD 6.23 billion, presenting a total percentage of 50.3% when compared to other financing modes. Secondly, the largest Sukuk issuances was the Mudarabah model offering a value of 28.3%, a total amount of USD 3.5 billion, and Wakalah asset financing was a percentage of 21.4%.

As shown in Table 17.2 above, the range of conventional bonds issue size varies between USD 124.9 million to USD 20 billion, a total of USD 43.1 billion altogether. Whereas, Bonds and Notes amount to a value of 48.8% equating to a total of USD 21 billion, while the second largest issued conventional bonds are Medium Terms Note at 38.6%. Euro Medium Notes was 10.3% and Perpetual Security has a value of 2.3%.

Sukuk and conventional bonds have a limited range of maturity from 3 to 10 years, except Emirates NBD Tier 1 Ltd- PRP, which is a perpetuity security with no limited time of maturity. Since there is limited available information on firms issuing both Sukuk and conventional bonds, attempts were made to find the Sukuk and bonds issuances from the same country to give an adequate comparison while eliminating state and industry risks.

1.1 Data Analysis

As previously mentioned, data analysis applied is quantitative in nature and involves several research methods. The study uses the Value at Risk (VaR) method as a main statistical tool by employing the Historical simulation and Monte Carlo simulation models to analyse risk levels of Sukuk and conventional bonds. An F-test is used to evaluate any differences between Sukuk and conventional bonds. Firstly, the log-normal daily returns are calculated from the daily closing prices of Sukuk and conventional bonds, and descriptive analysis through Excel calculating the expected average returns, variance, standard deviation, and skewness is applied. This descriptive statistic measures the central tendency and spread of data which allows a summary of data analysis. Historical simulation is then applied to examine the risk level of Sukuk and conventional bonds for the holding period of one year.

Secondly, the Monte Carlo simulation (MCS) model is applied to simulate the expected future returns distribution randomly to analyse possible losses of portfolio for Sukuk and conventional bonds. However, due to the literature recommendations and limited information on Sukuk and conventional bonds in the UAE’s financial market, the MCS is a more effective method to check the VaR risk level of Sukuk and conventional bonds since it allows the repeated simulations of daily returns of portfolios, using a random process of NORMSINV (RAND) function of Excel. The study employs two VaR estimates of MCS; firstly, the 10 days holding period of the VaR estimates with 99% confidence level, as per recommendations of the Basel Committee. Another reason behind the 10 days VaR estimates is because Sukuk are not very liquid products and require enough time to be traded on security markets. Secondly, the study generates (5000 repetitions) possible paths for the VaR estimates using randomly generated numbers to analyse the risk level of Sukuk and conventional bonds over the holding period.

1.2 Value at Risk Method

The Historical simulation method uses real data and reflects actual behaviour of the sample data. It provides a more reliable level of VaR estimation if the real distributions of sample data are ‘at tails’. In addition, it is easier to estimate unlike other methods of VaR models (Mentel, 2013). Drawbacks to the model include: it assumes the historically simulated distribution can represent the future distributions; it possesses higher variation compared to other models; and needs considerable number of data to estimate the quantiles of the empirical distribution (Hassan, 2006).

Whereas, the Monte Carlo simulation (MCS) model simulates the risk factors randomly rather than being analytically obtained, such as, in the Delta-Normal model. The MCS has several advantages making it the most comprehensive model to measure market risks, if accurately implemented including: it is flexible to incorporate time variations in the returns or the volatility of the returns and can be implemented in the presence of fat tails; it can also be applied to any type of portfolios. On the other hand, the MCS details disadvantages such as: it is more complex to compute, and system implementation is costly; and it relies on specific stochastic processes for the risk factors, which can lead to inaccurate results when the stochastic process used are wrong (Hassan, 2006).

1.2.1 VaR Calculations

In this section, the study derives general formulas of VaR models, including both Historical and Monte Carlo simulations. The basic concept of calculating VaR along with its probability and density function is shown in Table 17.3, which demonstrates the maximum loss of portfolios over a given time span where the chosen confidence level of VaR lies on a horizontal axis in a particular time. If the returns are normally distributed, the VaR confidence level of 95%’s Z-value is 1.645; and VaR at 99% confidence level, Z-value is 2.326. For example, the maximum loss of portfolio over a given time horizon is an amount of y-axis with a probability of 95% (see Boxed section) as shown in Fig. 17.1 (Ahmad et al., 2015, pp. 43–44).

Fig. 17.1
figure 1

VaR confidence level. Sources Ahmed et al. (2016, p. 44)

Full size image
Table 17.3 Descriptive characteristics of Sukuk daily returns
Full size table

In calculating historical simulation, the model uses the calculated log-normal daily returns of Sukuk and conventional bonds over the holding period of 250 trading days and the daily returns are calculated as follows:

$${\text{Daily returns}} = {\text{LN }}\left( {{\text{P}}_{\text{t}} {\text{/P}}_{\text{t - 1}} } \right)$$
(17.1)

where P1 is the price for today on Sukuk and conventional bonds and P0 is the price of Sukuk and conventional bonds for day before, and 0, 1…t, is the holding period of portfolios. Here, to calculate the VaR using the Historical simulation model with 99% of confidence level, the Excel function of Percentile on daily returns are applied as follows:

$${\text{VaR}} = {\text{PERCENTILE}}\left( {\left( {{\text{r}}_{1\ldots {\text{t}}} } \right),{\text{Z}}_{\upalpha}} \right)$$
(17.2)
$${\text{VaR = PERCENTILE}}\left( {{\text{RETURNS RANGE, 1\% }}} \right)$$
(17.3)

where r1…t is the periodic daily returns of Sukuk and conventional bonds in “t” for holding period over the time, Zα is the VaR confidence level of 99%.

In the case of Monte Carlo simulation, the study uses a geometric Brownian motion process to describe future returns on Sukuk and conventional bonds. The method uses two component parameters; drift which is a constant directional movement and random shock that presents the market volatility as shown below:

$${\text{Future daily returns}} = drift + \sigma * Z$$
(17.4)
$${\text{Future daily returns}} = (\mu - \left( {{\upsigma }^{2} {/2}} \right){{ * t + \sigma * Z}}$$
(17.5)

The drift is calculated from the mean (µ) minus half of the variance over time as shown in Eq. 17.5, σ is the volatility, and Z is the random variable from standard normal distribution N (0.1) using Excel function of NORMSINV (RAND), which will generate a simulated value of normal random variables having the parameters of mean (μ) and standard deviation (σ). Hence, the future expected daily returns are calculated by employing Monte Carlo simulation as follows:

$${\text{Future daily returns}} = NORMINV(RAND(), \mu , \sigma )$$
(17.6)

In this study, future daily returns are simulated using the above formula and the simulated daily returns use as 10 days for a holding period with a 99% confidence level of VaR estimates to analyse the risk level of Sukuk and conventional bonds portfolios. Also, to increase the accuracy of calculations 5000 simulations of Sukuk and conventional bonds daily returns will be run. According to Dowd (2005), ‘Monte Carlo simulation procedures, accuracy will vary with square root of number of trial’, so the more simulation, the better the result outcomes. In addition, the study presents the expected daily returns of Sukuk and conventional bonds in line chart and frequency, Histogram, to plot the simulated future returns to examine the normality distribution of the expected returns both Sukuk and conventional bonds.

F-Test Analysis

F-Test is used to test if the variance of two populations are equal. The test can be a two-sided test or one-tailed test. Whereas, the two-tailed version tests against the alternative hypothesis that the variances are not equal. On the other hand, the one-tailed version only tests in one direction that is the variance from the first population is either greater than or less than (but not both) the second population variance. Also, the F-test uses the Chi square test as combined method to find the p-value. Therefore, the F-test hypothesis is defined for a two-tailed test as follows NIST (2013):

$$\begin{aligned} {\text{H}}_{0}\, =\, & {\upsigma }_{1}^{2} = \sigma_{2}^{2} \\ {\text{H}}_{1}\, =\, & {\upsigma }_{1}^{2} \, \ne {\upsigma }_{2}^{2} \\ {\text{F}} \,= & \,{\text{s}}_{1}^{2} {{/s}}_{2}^{2} \\ \end{aligned}$$
(17.7)

where \(\sigma_{1}^{2}\) and \({\text{s}}_{2}^{2}\) are the sample variances. The more this ratio deviates from 1, the stronger the evidence for unequal population variances. The study will apply the two-tailed test, which tests against the alternative that the variance is not equal with a 99% confidence level. Therefore, this method will test if there is difference between Sukuk and conventional bonds risk level.

2 Analysis

The study presents VaR application to Sukuk and conventional bonds and discusses the empirical results that are obtained from the Historical and Monte Carlo simulation models and then presents the assessment and discussion of the results.

2.1 Descriptive Statistics of Sukuk and Conventional Bonds

As illustrated in Table 17.3 and Fig. 17.4, they report the descriptive statistics for the sample data of Sukuk and conventional bonds, using the Excel data tool which calculates daily returns’ mean of 250 daily returns, variance, standard deviation, and skewness of the returns distribution for Sukuk and conventional bonds.

As shown in the Table 17.3, most of the Sukuk issuances have a positive mean 0.012%, except for DP World Sukuk Ltd IDB Trust Services Ltd, and Hong Kong Sukuk Ltd which shows a negative of expected daily returns. Also, Dar Al-Arkam Sukuk and JAFZ Sukuk Ltd dispense the highest volatility of 0.899 and 0.882% in which their daily returns are more volatile than the others. In addition, DIB Sukuk Ltd has a higher volatility of 0.722% too.

Furthermore, the above Fig. 17.2 indicates that the expected daily returns are not normally distributed around the mean but inclined more towards negative values in the left tail. Whereas, the return distribution of Sukuk are skewed and falls in the left tail. Also, there are some extreme daily return outliers that effects overall daily return distributions. More precisely, the expected returns of Sukuk are more left skewed as presented in Fig. 17.2. This means that daily returns of Sukuk are in negative outcome which clusters more in left tail.

Fig. 17.2
figure 2

Distribution of average daily returns for Sukuk

Full size image

As shown in Table 17.4, most of the conventional bonds also have a positive mean, except for Dubai Holding C.o. Ltd, Emirates NBD PJSC 2015, and ICD Funding Ltd that shows a negative mean of daily returns as well as ICD Funding Ltd which provides a higher standard deviation of 0.760%, this shows that the expected of ICD Funding has a higher spread of returns when compared to other bonds. In general, conventional bonds daily returns are less volatile compared to Sukuk except ICD Funding Ltd and MAF Global Securities 2014.

Table 17.4 Descriptive characteristics of conventional bonds daily returns
Full size table

On the other hand, the overall mean of returns on Sukuk is 0.012% is slightly higher than conventional bonds’ mean of 0.010% were as the descriptive statistics result points out that Sukuk instruments are more volatile and riskier than conventional bonds.

The sample skewness of conventional bonds shows that daily returns are not distributed equally around the mean and there are some outliers of daily returns as illustrated the above Fig. 17.3. Also, the chart shows that the average daily returns of conventional bonds are skewed to the left tail. However, the expected daily returns are more skewed to the left tail but it shows that they are slightly less skewed compared to Sukuk daily average returns as presented in the histogram charts of Figs. 17.2 and 17.3. This means that daily returns of Sukuk are in negative outcome which clusters more in left tail.

Fig. 17.3
figure 3

Distribution of average daily returns for conventional bonds

Full size image

2.2 Historical Simulation Results

Historical simulation is implemented by using log-normal daily returns of Sukuk and bonds issuances taking 250 trading days in a year. The VaR with confidence of 99% is simply calculated as 1% percentile of hypothetical loss or gain of probability density function, but mainly we concern the left tail of the bell curve values which are the worst loss in a time.

Table 17.4 illustrates the value of VaR on Sukuk and conventional bonds with the VaR confidence of 99% level over a period of a given time. Whereas, the riskiest Sukuk are Dar al-Arkam Sukuk with the VaR of (−3.54%) and Indonesia III-144A with VaR of (−2.04), while the lowest VaR is the DP World Sukuk (−0.20%). On the other hand, the average VaR on Sukuk is (−0.90%), while the conventional bonds average VaR is (−0.48). This shows that Sukuk are twice as risky as conventional bonds detailed in the above table. In addition, the riskiest bonds are from MAF Global Securities-14 and Dubai Holding Co. VaR values of (−1.12% and −0.75%), while the lowest VaR on bonds is Emirates NBD PJSC 2014 for (−0.19) (Table 17.5).

Table 17.5 Historical simulation for Sukuk and conventional bonds at 99% of VaR confidence
Full size table

As illustrated the above in Fig. 17.4, Sukuk has a higher risk level of daily returns than conventional bonds which shows that bonds are less risky than a Sukuk. Whereas, the average VaR on Sukuk is almost twice as high as the average VaR on conventional bonds as shown in Fig. 17.4. Furthermore, the bonds daily returns are more stable than the Sukuk over the period. Therefore, this shows that Sukuk daily returns are more volatile and riskier than conventional bonds.

Fig. 17.4
figure 4

VaR of Sukuk and Bonds using confidence level of 99%

Full size image

2.3 Monte Carlo Simulation Method

Here, the risk level of Sukuk and bonds using Monte Carlo simulation (MCS) model with the VaR 99% confidence level are described. To compute VaR, the study applies two VaR estimations: firstly, the study estimates 10 business days of holding period for randomizing daily returns, as per recommendations of the Basel Committee. In addition, Sukuk are not very liquid products and require time to be traded. Secondly, the study estimates randomly the VaR by simulating daily returns into (5000 repetitions) where the model applies to two parameters: drift which is a constant directional movement and random shock for volatility as presented in Eq. 17.3. Table 17.6 presents the value of VaR on Sukuk and conventional bonds using Monte Carlo simulation with the VaR of 99% confidence level. Hong Kong Sukuk Ltd and Dar Al Arkam Sukuk provide the highest VaR risk levels of (−0.165% and −0.162) in 10 days holding period and the highest VaR values of (−2.13%) for Dar Al Arkam and (−1.99%) for Hong Kong Sukuk Ltd. Whereas, the lowest VaR values are provided by Indonesia III-144A in both the 10 days and 5000 runs of VaR estimates.

Table 17.6 Monte Carlo simulation for Sukuk and bonds in 99% of VaR confidence
Full size table

In the case of bonds, MAF Global Securities-14 and ICD Funding Ltd have the highest VaR Values of (−1.18% and −0.86%) in 10 days, as well as, the highest VaR estimates of (−1.73% and −1.42%) in the VaR estimates of 5000 repetitions. While the lowest VaR on bonds is Emirates NBD PJSC 2014 for (−0.11%). In summarising, the VaR values of Sukuk are higher than conventional bonds VaR values in both the day VaR and 5000-iterations as shown the above Table 17.6. Therefore, the model shows that the expected average for the daily returns loss of value cannot exceed (−0.68%) for Sukuk and (−0.34%) for bonds with 99% confidence level in 10 days VaR estimates. Whereas, the expected average of daily returns loss of value cannot exceed more than (−1.02%) for Sukuk and (−0.54%) with 99% confidence level in the second VaR estimates of 5000 iterations.

Figure 17.5 reveals clearly that VaR risk level of Sukuk are greater than conventional bonds in both the 10-day holding period and the VaR estimates of (5000 iterations) as shown in the above figure. This illustrates that expected daily returns of Sukuk are almost double those of conventional bonds in both the Historical and Monte Carlo simulation models. Therefore, Monte Carlo simulation strongly supports the results of Historical simulation, which means that Sukuk has more risk than conventional bonds and we can conclude that Sukuk risk levels are strongly higher than conventional bonds which indicates that Sukuk instruments are different and riskier than conventional bonds.

Fig. 17.5
figure 5

VaR of Sukuk and bonds in 10 days holding period and VaR estimates of (5000 repetitions) with 99% confidence level

Full size image

In addition, a line chart and frequency histogram are used to simulate the expected daily returns of Sukuk and conventional bonds as shown in Fig. 17.6, by selecting the highest and the lowest VaR values of Sukuk and conventional bonds. In the case of Dar Al Arkam Sukuk, the line chart in 10 days VaR estimates shows that the expected daily returns of the Dar Arkam are volatile in the 10 days’ period, indicating more risk. Whereas, expected returns are more stable and normally distributed after performing 5000 simulations as shown in Fig. 17.7. For the lowest Sukuk VaR values, the expected daily returns of Indonesian III Sukuk are less volatile and normally distributed, illustrated in Fig. 17.8 (see Appendix 2).

Fig. 17.6
figure 6

Sukuk distribution in line chart and histogram

Full size image
Fig. 17.7
figure 7

Conventional bonds distribution in line chart and histogram

Full size image

With conventional bonds, the expected daily returns of ICD Funding provides a changing trend in the10 days’ holding period applying the Monte Carlo simulation, which means that its daily returns are not stable compared to others. On the other hand, its expected returns appear stable and normally distributed in 5000 runs’ VaR estimates as presented in Fig. 17.7, while Emirates NBD BJSC-14 are more stable and normally distributed after performing 5000 simulations (see Appendix 3).

2.4 Comparing VaR Results of Historical and Monte Carlo Simulation

Comparison is, now, offered of both methods’ results of VaR calculations in Historical and Monte Carlo simulations (MCS) as reported in Figs. 17.6 and 17.7. The worst expected loss of values in a day for Sukuk and bonds’ portfolios are calculated by multiplying the worst daily returns of VaR estimates with a hypothetical portfolio value of $100,000. For example, historical simulation method shows a VaR of $900 for a portfolio of Sukuk issuance and a VaR of $480 for bonds’ portfolio. This implies that one can expect the maximum daily loss on the market value of the Sukuk portfolio will not be larger than $900 and for bonds’ portfolios will not be greater than $480 of its value at 99% of the time. It also means that there is a 1% chance the loss could be larger than those values (Table 17.7).

Table 17.7 Maximum loss of Sukuk and conventional bonds’ portfolios
Full size table

In the case of the MCS method, results show a VaR of $680 for a portfolio of Sukuk and a VaR of $340 for bonds’ portfolios in VaR estimates of 10 days. Whereas, a VaR of $1,020 for a portfolio of Sukuk and a VaR of $540 for a portfolio of bonds in VaR estimates of (5000 iterations). Therefore, this suggests that the worst daily loss of Sukuk portfolio will not exceed $680 and $340 for bonds with the 99% confidence level in 10 days VaR estimates. In the second model of (5000 iteration) MCS, the maximum loss in a day will not exceed $1,020 for a Sukuk portfolio and $540 for bonds, which means that there is a 1% chance the loss could be larger than these VaR estimates.

On the other hand, the historical simulation shows that the expected daily loss of Sukuk portfolios are greater than a portfolio of bonds with a value of 47%. Similarly, the results of Monte Carlo simulation show that the percentage change of both methods are close to the historical simulation method which highlights that the findings are robust when compared to each other. For example, the expected daily losses of Sukuk portfolios are higher than those of conventional bonds,’ illustrating 50% in a 10-day holding period. Whereas, the second method shows that Sukuk have higher losses of 47% compared to conventional bonds. However, the slight differences apparent to both methods can be attributed to the extent to which returns on Sukuk and bonds diverge from normal presumptions, whereby, deviation from normality can lead to a different estimate of VaR when a normality assumption is dropped both in Historical simulation and the MCS methods. As illustrated in the above Table 17.6, Sukuk portfolios have a higher VaR risk level of expected daily losses than the conventional bonds. This highlight Sukuk as riskier than conventional bonds.

However, differences between Sukuk and conventional bonds risk level were tested using the F-test model by employing Minitab software. The results show that p-value is less than 0.01 with a 99% confidence level as presented in Fig. 17.8, which confirms that there are significance differences between the risk level of Sukuk and conventional bonds (see Appendix 4). Therefore, this results in a null hypothesis to be rejected, and, hence, accept the opposite. Therefore, the statement that Sukuk and bonds should have similar levels of risk were incorrect as revealed through the above figures, hence, a rejection of the null hypothesis, which illustrates no difference between Sukuk and conventional bonds, and the results support the alternative hypothesis that stated Sukuk are riskier than conventional bonds.

Fig. 17.8
figure 8

F-Test results for Sukuk and conventional bonds

Full size image

Furthermore, the study shows that Sukuk inherits higher market and credit risks than conventional bonds due to some restrictions that Shar’iah principles impose on Sukuk risk management. Also, other risks related to Sukuk are legal and regulatory risks specific to Sukuk may effect the Sukuk returns, and operational risks, which donot effect conventional bonds. Therefore, these risk factors place a major disadvantage on Sukuk competiveness in the global market as an alternative investment product.

3 Conclusion

Sukuk is an important instruments of an Islamic financial system, and normally, allow mobilisation of resources. This helps Islamic financial institutions to match their assets and liabilities. Whereas, the Sukuk market has diverged from initial Sukuk issuances by financial institutions to issuers ranging from infrastructure development, aircraft financing, socially responsible investing, Takaful sector, capital enhancement purposes, and so forth. Also, Sukuk encourages small investors to participate in Islamic financing and earn profits at the same time.

The uniqueness of Sukuk from its conventional counterparts has been a debatable issue among scholars and within relevant literature as an alternative to investment products. This study introduced a portfolio of Sukuk and conventional bonds to evaluate the risk level of Sukuk and conventional bonds. The results show that Sukuk portfolios have a higher VaR risk level of the expected daily losses than the traditional bonds in both Historical simulation and Monte Carlo simulation. Therefore, the empirical evidence reveals that Sukuk instruments are different to, and riskier than, conventional bonds. However, the study shows that Sukuk inherits higher market and credit risk than traditional bonds due to some restrictions that Shari’ah principles impose on Sukuk risk management. Also, the results confirm that these risks may be associated with other risk factors of Sukuk, such as, legal, regulatory, and operational risks, not affecting conventional bonds. Therefore, these specific risks may place significant disadvantage on the competitiveness of Sukuk in global markets as an alternative investment product.

There is little research regarding Sukuk compared to conventional products in literature. Therefore, this study fills the gap by examining risk levels of Sukuk instruments, when compared to conventional bonds. The findings reflect those of Cakir and Raei (2007) and Hassan (2006).

Finally, this study evaluates risks of Sukuk when compared to conventional bonds by investigating it as an alternative investment instrument. Although, the findings of this study are based on a reasonable sample of Sukuk and conventional bonds’ portfolios, further studies including corporate and sovereign Sukuk and conventional bonds issuers, such as Qatar, Bahrain, Kuwait, and Saudi Arabia will produce better evaluations of Sukuk effects on investment portfolios.

Interesting aspects for future research include an examination and comparison of different industry sectors of Sukuk and conventional bonds in the Gulf states. Here, other methods of Sukuk risk evaluation may be applied using suitable software, since recent risk evaluation studies on Sukuk and bonds rely mainly on VaR Method analysis.