Credit Risk and Efficiency: Comparative Study between Islamic and Conventional Banks during the Current Crises

Veiw figure View Table

When analyzing the preceding table, we notice that, for equations (1) and (4) where the LLRGL is used as proxy of the credit risk, the probabilities of the dummy variable of the Islamic banks (isbd) take values lower than 5% but these coefficients take negative values equal to (- 2.224195) and (- 1.390625), respectively. These results show that conventional banks’ credit risk is higher than that of their Islamic counterparts. Several studies, such as that of Pirner (2003) and Coyle (2000), state that credit risk has a high effect on the exposure to financial crises. As a consequence, credit risk can threaten the bank if not managed correctly.

For the variable size, which implies the bank size, and according to the estimated results of equations (1) and (4), the probabilities take positive values equal to (0. 043) and (0.039) respectively which are lower than 5% whereas their coefficients take negative values equal to (- 2.682952) and (- 2.189806). Therefore, the relationship between the size and the credit risk degree is negative, which seems coherent with the diversification possibility and the advantages of the economies of scale. According to these results, it appears that the positive impact of the size on the quality of the loans for Islamic banks is lower than that of traditional banks.

According to the methodology adopted by Morgan and Samolyk (2003), Stiroh (2004), Stiroh and Rumble (2006) and Mercieca et al. (2007), we measure diversification using the herfindhal Hirschman index (HHI), which is a concentration index. The higher the HHI index, the more the bank is largely concentrated and less diversified in a given segment. Based on the obtained results using equations (1) and (4), where LLRGL is introduced as a credit risk proxy, the HHI concentration index is significant and positive. These results show that, the more the concentration within the Islamic banks increases, the higher the credit risk will be.

This phenomenon can be explained by the fact that credit concentration can take the form of a risk higher than the average in relation to any economic or geographical sector, making the lending bank vulnerable to the difficulties of an industry or to a particular area. It is therefore important that banks systematically identify and evaluate sectoral or regional risk so that the management will be aware of the incurred risks and sets up a better balance, if necessary. It seems that the relationship between concentration and risk continues to be a subject of debate between researchers. There are several studies which claim that this concept is said to improve the profitability of the banking institutions and therefore reduce risk. However, some claim the opposite by providing evidence that concentration is the cause of a risk increase and a decline of banking performance.

The ETA negative sign in equations (1) and (4), means that a more important equity capital leads to a weaker risk. The conclusions of Konishi and Yasuda, (2004), show that requirements for equity capital reduce risk taking of the banking incentives. For Demsetzand Strahan (1997), banking capitalization positively and significantly affects the probability of bank defect. These authors find a negative correlation between overall risk and equity capital. According to our results, Islamic banks are more affected than conventional ones by the negative impact of the ETA on the credit risk.

It can also be noted that the variables D9, D10 and D11, which represent years 2007, 2008 and 2009, and the results in columns (1) and (4), take negative but significant signs. On the other hand, variable D12, which indicates the year 2010, seems to have a significant and positive effect. One can conclude that Islamic banks are less risky than the conventional ones during the three mentioned years. Year 2007 is the outbreak of the subprime crisis.

Several authors assessed the level of losses of conventional banks during the period of this crisis. Among these authors, we find Mark Zandi (2009), an economist of the agency of Moody swimming, who states that there are great losses which can reach 225 billion dollars. Another study carried out by Deustche Bank assessed these losses at 400 billion dollars.

Therefore, our results show that, in case the LLRGL presents the credit risk proxy, Islamic banks are not hit by the subprime crisis. This can be explained by the fact that the Islamic banks are far from developing mortgages as are prohibited by Islamic law (Sharia). For this reason, all researchers and experts said that the subprime crisis has not affected the Islamic financial institutions.

The difference between systems (1) and (2) lies in the fact that we added the variable Size X ISBD (the interaction between the size and the dummy variable of Islamic banks). The interaction term between the size and the dummy variable of Islamic banks is positive and significant in equation (4) where the LLRGL variable is the credit risk proxy. This indicates that the positive impact of the size on the credit risk is lower for conventional banks.

Our results in equation (6) show that the larger the size of Islamic banks the higher inefficiency will be.

Table 2. Estimation results of the ETA and LLRGL simultaneous equations

Download as

Veiw figure View Table

Both systems in the table above are based on Hughes and Moon (1995) and Hughes and Mester (1998). A two-equation system of both the (LLRGL) risk and the capital in which the level of inefficiency is controlled is estimated. In this configuration, the inefficiency lagged value is used as a preset variable.

The results show that Islamic banks have a lower credit risk and a higher capital than their conventional counterparts. According to these two equations, improvement in profitability and efficiency helps strengthen bank capital. Our results are similar to those generated by Fiordelisi and al (2010). By adding the interaction term between the size and the dummy variable of Islamic banks to the third and fourth equation, we notice thatIslamic banks profit less from the negative impact of the size on the credit risk than conventional banks.

Table 3. Estimation results of the simultaneous equations of PLGL (Problem Loans / Gross Loans), capital and inefficiency

Download as

Veiw figure View Table

In the first equation, and according to the results identified in the table above, where the PLGL(Problem Loans / Gross Loans) is used as a proxy, the coefficient on the dummy variable for Islamic banks (ISBD)is insignificant with a probability greater than 5%.

On the basis of the third equation, we can say that the level of inefficiency of Islamic banks is not significantly different from that of conventional banks.

The size measured by the logarithm of total assets is said to be a control variable as it can affect the capital level, the risk and the profitability of the bank through the economies of scale. According to our results, there is a negative relationship between the size and the credit risk level, in other words, large banks can easily enter capital markets and draw a greater diversification of their portfolio. They are supposed to have a lower capital and risk level than smaller banks.

In the second system, the term of interaction between the size and the dummy variable of the Islamic banks is added. It appears from estimation (4) that the dummy variable of Islamic banks (isbd) has a value lower than 5% whereas this coefficient has a value of (-0.3625112). This shows that Islamic banks have lower risk than their conventional counterparts.

We also see that, according to equation (6), the dummy variable of Islamic banks is equal to (0.020) and lower than 5% and their coefficient is equal to (- 0.0280105). One can thus say that the efficiency level of Islamic banks is higher than that of conventional ones.

Regarding the variables D9, D10, D11 and D12, which indicate the years 2007, 2008, 2009 and 2010, respectively, we concluded, according to results in the table above, that during the first three years, the credit risk of Islamic banks is lower than that of conventional banks.However, at the end of the year 2010, we noticed that there is not a great difference between the Islamic banks and their conventional homologous.

These results make us say that, during the subprime crisis period, Islamic banks were not affected since the mortgages are prohibited by the Islamic laws.

However, on the basis of the results of 2010, we can say that IB were not excluded from the credit risk which can occur in several types of contracts of Islamic banks, such as Salam, Mourabaha, Musharaka, and Mudaraba contracts.The credit risk rose during that year since 2010 is the fourth phase of the current financial crisis called “the sovereign debt crisis”. This crisis affected the world markets in the East as well as in the West, such as the Japanese stock exchange market and other markets in Europe and in the United States and many countries around the world. This crisis also affected the Islamic banks as these ones have economic ties with the global equity markets as well as the International Monetary Fund.

Table 4. Estimation results of the PLGL (Problem Loans / Gross Loans) and ETA simultaneous equations

Download as

Veiw figure View Table

According to the first equation, the credit risk level (PLGL) does not greatly differ from that of conventional banks. On the other hand, and according to the result estimated in equation (3), the dummy variable of Islamic banks (ISBD)is equal to (0.022) and lower than 5% and its coefficient is equal to (- 1.184548). We can thus conclude that the credit risk level (PLGL) within Islamic banks is lower than that in conventional banks.

The interaction between the size and the dummy variable of Islamic banks (ISBD) in the third equation prove to be significant and positive with a probability equal to (0.030) and a coefficient equal to (3.442718). This result shows that Islamic banks benefit less than the traditional banks from the negative impact of the size on the credit risks.

In the third equation, the HHI concentration degree is significant and positive with a coefficient equal to (14.34249) and a probability of (0.017). These results show that Islamic banks are less affected than conventional banks by the diversification impact on the rise of the credit risk. The relationship between inefficiency and ETA in equations (2) and (4) seems to be negative and significant. This implies that the improvement of banks’ profitability and efficiency increases and strengthens the bank capital. Our results confirm the conclusions of Van Roy (2003) who state that capital regulation can help increase the banks’ capitalization level. Moreover, the improvement of the capital was accompanied by a decline of the credit risk and by a rise of the interest margin.

Table 5. Estimation results of the simultaneous equations of LLPAGL (Loan Loss Provisions /Average Gross Loans), ETA and Inefficiency

Download as

Veiw figure View Table

The LLPAGL (Loan Loss Provisions /Average Gross Loans) is the credit risk proxy in both estimates (1) and (2) in the table above. According to the result estimation in equation (1), we can say that the dummy variable of Islamic banks (ISBD)is equal to (0.022) and lower than 5% and its coefficient is equal to (- 0.073586). These results show that Islamic banks are less affected by the credit risk than the conventional ones.

As for the four variables D9, D1, D11 and D12, which respectively indicate the years 2007, 2008, 2009 and 2010, we notice that, according to the results in equations (1) and (4), during the first two years, the credit risk of Islamic banks is lower than that of conventional banks. However, at the end of the third and the fourth year, it rose in Islamic banks.

In the second system, we added the variable Size X ISBD (it is the interaction between the size and the dummy variable of Islamic banks).

According to the results of the table above, in which LLPAGL is the credit risk proxy, we notice that the relationship between risk credit (LLPAGL) and the interaction term between the size and the dummy variable for Islamic banks is significant and positive with a probability of (0.005) and a coefficient equal to (0.1615581). This means that credit risk (LLPAGL) is higher for large-sized Islamic banks.

Table 6. Estimation results of the simultaneous equations LLPAGL and ETA

Download as

Veiw figure View Table