DETERMINANT OF REPUTATION RISK MANAGEMENT IN ISLAMIC BANKS IN INDONESIA

_


INTRODUCTION
The company's reputation illustrates the stakeholders' perceptions of a company. (Deloitte, 2014) explained in his report that reputation risk ranks first as a risk factor that is a concern of management. This condition is reinforced by the increasingly integrated information through social media (Lee, Hutton, & Shu, 2015). The inability of companies to filter directly on social media makes it more difficult for companies to manage reputation risk. Therefore, cases related to social media require a different approach, so companies must pay close attention to produce effective solutions (Ott & Theunissen, 2015).
Financial institutions, especially banks, are industries that are closely related to the trust of customers or consumers (Fiordelisi, Soana, & Schwizer, 2014). Without the trust of consumers, banks cannot run their business. Banking is responsible for managing the assets of its customers. Therefore, the ease of carrying out transactions on their assets is a major demand for customers. This causes banking to depend on advances in information technology. The banking industry is aware of the close relationship between information technology and the reputation risk of a company (Economist Intelligent Unit, 2012).
(Basel Committe on Banking Supervision, 2009) define reputation risk as a risk that arises because of negative perceptions of customers, business partners, shareholders, or regulators that can affect the ability of banks to conduct business or create new business opportunities and the ability to obtain funding sources. In Indonesia, the regulator defines reputation risk as a risk due to a decrease in the trust from stakeholders, where the decline was caused by negative perceptions of the bank (Otoritas Jasa Keuangan, 2016). Rules related to reputation risk management are included in the regulations regarding risk management that must be applied by banks.
The potential for reputational risk to occur generally begins with the occurrence of operational losses. When losses incurred due to reputation risk outweigh losses caused by operational losses, that is when reputation risk occurs (Heidinger & Gatzert, 2018). Some studies show losses caused by reputation risk, in general, is greater than losses incurred by operational risk (Cummins, Lewis, & Wei, 2004;Fiordelisi et al., 2014).
In the case of Indonesia, several incidents that were expected to cause a decline in the reputation of the bank have occurred several times. (Suhardjanto & Dewi, 2011) mention several cases that reflect the poor management of the bank's business. These cases included the case of Bank Summa in 1992, the double financial reports carried out by Lippo Bank in 2002, the BNI bank fictional L / C cases in 2002, the closure of Global Bank in 2004 and the 2008 Century Bank case. In some cases the deterioration in the condition of the bank was accepted by the community, followed by a decline in public confidence. In some instances, the impact caused is large enough to cause the bank to be liquidated or change ownership. (Oorschot, 2009) states that one of the main conditions that cause a high risk of banks is that banks are risk-taking entities. Banks tend to take risky actions to run their businesses or improve their business. Therefore, the ability to recognize all the risks that might be faced must be carried out by the banking industry.
Based on the explanation of the conditions above, this study raises the following problems (1) How is the implementation of reputation risk management in Islamic banking in Indonesia? and (2) What factors are decisive da lam implementation of reputation risk management in Islamic banking in Indonesia? The purpose of this research is To find out the extent to which Islamic banks are aware of the reputation risks they must face and Factors that determine the implementation of reputation risk management.

LITERATURE REVIEW
In general, research related to the reputation risk of existing companies focuses on the reputation of the company, especially regarding its definition and measurement (Clardy, 2012). In addition, research on reputation risk also addresses reputation and its relation to financial performance (Gatzert, 2015).
Empirical research on reputation risk as the market focused on the reaction that followed the case of the occurrence of operational losses. The results of the research, in general, show that the negative impact shown by the market reaction is greater than the operational losses that occur (Cummins et al., 2004;Fiordelisi et al., 2014;Gillet, Hübner, & Plunus, 2010). Therefore, it indicates a reputation loss that has a significant impact. (Fiordelisi, Soana, & Schwizer, 2012) then investigate what factors determine reputation risk for the bank.
In other literature discuss how to manage reputation risk. Some studies focus on improving reputation after a crisis, including by conducting communication strategies but only a few of them take a proactive approach to managing reputation risk.  conduct research by incorporating reputation risk in Enterprise Risk Management (ERM) holistically. Meanwhile, (Gatzert, Schmit, & Kolb, 2016) examined insurance solutions for reputation risk as a measurement of risk management. While (Mukherjee, Zambon, & Lucius, 2009), analyze reputation risk disclosure in 20 European banks and calculate the frequency of related words.
Overall, the results of previous studies indicate it is important for companies to monitor the level of risk of company reputation because the reputation and negative events that follow after the damage to a company's reputation can worsen the company's performance (Gatzert, 2015). Therefore, managing the reputation of the company's risk must create value for the company. However, despite its considerable influence on the condition of a company, there is little empirical research that addresses the determinants and values of managing reputation risk in a company. Even though in a larger context, namely Enterprise Risk Management, discussions on this topic are often done (Beasley, Pagach, & Warr, 2007;Gatzert, 2015;Gordon, Loeb, & Tseng, 2009). Because the Enterprise Risk Management which is good to be able to capture the entire category of risk, including reputation risk, reputation risk as a continuation and development of ERM should be a positive signal for our shareholders and it should be a concern for the management.
This study aims to investigate the factors that determine how much awareness of Islamic banking is related to the reputation risk it faces. This awareness is seen from whether the Islamic bank has implemented the risk management reputation of the company. Furthermore, from the factors that have been determined analyzed whether these factors influence reputation risk management.
The first researcher to look at the impact of overall operational losses was Cummins et al., (2004). Using data sourced from OpVar, a database developed by OpVantage, a subsidiary of FitchRisk. Where in the OpVar collect all data operating losses experienced by all companies in the United States from the late 1970s until the research is done. Research at first only saw the impact of operational losses on the company's stock price. Have not seen the impact caused by the implementation of reputation losses. (Cummins et al., 2004) conducted a study using the event studies method to see the impact of operational losses on the stock prices of banking and insurance companies. The research includes data on all operational losses with a value of losses above USD 10 Million. The duration of the study was between 1978 and 2003 with the number of banks being the object of research as many as 403 banks and 89 insurance companies. The results of the study indicate a statistically significant relationship from the announcement of the existence of operational losses to the decline in stock prices. On average, the market response to the announcement is greater for insurance companies compared to banks. Furthermore, losses due to a decrease in share prices exceed the losses incurred due to operational losses. This condition indicates that the operational losses feared by investors can have negative implications for cash flows in the future. The greater Tobin's Q, the greater the impact. That is, companies with greater opportunities to grow will experience greater losses. The next study that looked at the relationship between operational losses and stock prices was carried out by (Perry & De Fontnouvelle, 2005). In this study, (Perry & De Fontnouvelle, 2005) first defines reputation loss as a loss that occurs when a decrease in the market value of a company is greater than the trigger, namely operational loss. In addition, they also did some updates. The first update is that the analysis of operational losses is made more profound. One way is to divide operational losses into seven parts in accordance with the distribution of operational risk based on Basel II. The seven risks are internal fraud; external fraud; safety of work environment and practices related to labor; relationships with clients, products produced, and business practices carried out; loss due to damage to assets in the form of physical; any disruption to the business or failure of the system; Exclusion, completion and process management. (Perry & De Fontnouvelle, 2005)  FIRST. The second update made by both is to include the rights of shareholders as further analysis. The measure of shareholder rights follows the G index of corporate governance based on measurements made by Gompers. The results of the study show that reputation losses occur following the occurrence of operational losses. A decrease in the market value of the company is greater than the operating loss that occurs. The next findings are companies with weak shareholder rights, there is no difference between the impact of operational losses due to internal fraud or the impact of operational losses due to non-internal fraud. In companies with strong shareholder rights, the impact of internal fraud is more than one to one. Thus, reputation risk is stronger in companies with strong shareholder rights. Gillet et al., (2010) is the next researcher who looks at the relationship between operational losses and stock prices. Departing from the two previous studies mentioned above, he did some updates on his research. The first update is to calculate the difference or difference between losses on market value and losses announced by the company. With this notation, it is easy for the author to isolate the real impact of the announcement of operational losses on the rate of return on shares.
The second update made by (Gillet et al., 2010) is to divide the announcement date not just one but into three parts. The first date is called the press date or when the news is first about operational losses entering the media coverage. The second date of the recognized date is the date when the number of losses incurred is calculated by the company. The third date is a settlement date, which is the date stated by the OpVantage database system. Furthermore, the results found from the research conducted by Gillet et al. is the following. In a global analysis, CAR (Cumulative Abnormal Return ) at settlement date is different from CAR at press date and recognition date. At settlement date, CAR is positive and not significant. Whereas in settlement date, CAR is negative and not significant. While at the press date, CAR proved to be significantly negative. This significant negative occurred before the announcement. This condition shows the existence of market overreaction in semistrong informational inefficiency and the existence of insider trading on strong form inefficiency. For cases in the United States, the same conditions also occur. In the EU there is little difference in the two previous conditions. In the European Union, the impact of purely mechanical operational losses is greater. Basically, the losses that occur on average are not much different, but because the market value of companies in the European Union is greater, the impact caused by operational losses is greater. The other main difference is the return around recognition date, where CAR is not different or has zero value before the event.

Data
This research aims to look at the application of reputation risk management and the factors that determine the level of reputation risk management in banking companies. As a population is all banking companies in Indonesia, while this study uses a sample of Islamic banking companies. In addition, the year period used is 2017. The selection of samples in Islamic banking, because the sharia industry is one of the fastest growing industries. Therefore, Prudential functions, including implementation of reputation risk management, it should be applied. This is so that the momentum of the development of the Islamic industry can be maintained properly. While the use of 2017 is because the year is the most recent year in which the annual report is published. Thus, it is expected that the most recent phenomena that occur will be known.
This study uses secondary data. Several data sources were used in this study. The first data source is the annual report issued by each Islamic bank. The second source is other reports issued by the Islamic bank, for example, GCG Reports, Sustainability Report, and other reports. The third data source is from the website of the Financial Services Authority (OJK) and Bank Indonesia (BI)

Variables
The variables used in this study refer to the variables used by (Heidinger & Gatzert, 2018). The explanation of the variable used is as follows:

Reputation Risk Management
To test how far a company applies reputation risk management in its company, this study uses a special category in classifying the level of implementation of reputation risk management. The value given is 1 if the company meets one of the categories. While the value of 0 is given if the company does not meet at all the categories: 1. Has its own risk category. Reputation risk has its own risk category naming. Whether it is combined in risk management or classified in other categories but has a separate definition. Thus, the company is already aware of the reputational risks they face. Naming

Company Size (Size)
Large companies usually face more complex risks. Reputational risk is generally seen as a "risk from risk", large companies should implement a reputation risk management program because the larger the size, the companies will face greater and more complex risks (Beasley et al., 2007). Furthermore, large companies generally have larger stakeholders and have greater public interest, thereby reinforcing the existence of reputational risk. Empirical studies show that firm size has a greater impact on reputation loss in events that cause reputation damage (Fiordelisi et al., 2014). In the end, larger companies also have more resources to implement risk management programs (Beasley et al., 2007). Thus, the hypothesis used in this study is that company size has a positive effect on reputation risk management. The size of the company in this study is defined as the natural logarithm of the book value of assets.
H 1 : Company size has a positive effect on reputation risk management.

Capital Adequacy Ratio (CAR)
The direction of the relationship between CAR and reputation risk is partly ambiguous.
On the one hand, companies with good risk management, including management of good reputation risk, will have considerable access to financing at a relatively low cost and therefore have considerable debt. On the other hand, several studies show companies with large CARs will statistically experience greater reputation losses (Sturm, 2013). This can cause a different reaction. A different reaction is that companies with CAR that will implement a reputable risk management program or companies that have a concern for reputation risk will reduce their debt so that they are not too heavily affected by reputation risk. Various empirical studies regarding the determinants of enterprise risk management also show several times that the relationship between CAR and risk management is not clear (Pagach & Warr, 2010). In accordance with other studies that examine the determinants of reputation risk management, this study uses the ratio of the book value of debt to book value of assets as a measure of debt.
H 2 : There is a relationship between CAR and reputation risk management Profitability (Fiordelisi et al., 2012) show that companies that have greater profits will usually suffer more from loss of reputation, which emphasizes the need for reputation risk management in these companies. Because companies with higher profit levels will usually also face costs related to reputation risk management programs that are also high (Lechner & Gatzert, 2018)), it is assumed that there is a positive relationship between RoA (as a proxy for profitability) and reputation risk management. Measurement of profitability uses the ratio of net income divided by the book value of total assets. H 3 : Profitability has a positive effect on reputation risk management.

Concern for Reputation
The frequency of the mention of the word reputation and reputation risk in the content analysis of the annual report will be used as a proxy for the company's concern for reputation risk. Companies that report or discuss risks related to reputation on the one hand and have concern for the reputation of the company, on the other hand, are generally aware of the importance of managing the risk of their company's reputation. Thus, it is expected that there is a positive relationship between reputation risk and reputation risk management.
H 4 : Concern for reputation has a positive effect on reputation risk management.

Concern for Risk
The greater the company's concern for risk, the more the company will be aware of the complex risks faced. The greater awareness will make them more able to mitigate the overall risks they may face. As a proxy for risk awareness, the measurement used is how much risk is mentioned in the annual report prepared by the company. The relationship that is expected to occur is a positive relationship.
H 5 : Concern for risk has a positive effect on reputation risk management.

Model
Based on the explanation of the variables above, the model used in this study is RESULT AND DISCUSSION -Heading 1 (TNR, 12pt, Bold, Align Left, Uppercase) Data collected from all Islamic banks that have complete financial reports between 2015 and 2017 on the website, www.ojk.go.id. If they do not have a complete financial report, the sample will be excluded from the sample selection.

Descriptive Statistic
Below is the descriptive statistic from the data. Whereas in profit it is found that the average profit or profit on Islamic banks in Indonesia is 2, 4% with the median or the value of the profit is 0.5%. Whereas in the CAR variable, the average CAR in Islamic banks in Indonesia is 20, 3% with a median of 17.2%.
Observing this value, most Islamic banks in Indonesia operate quite prudently or carefully so that they keep their CAR values far above what is required by the OJK, which is 12%.
Whereas related to ROA, the value is not too high, still around 2%. While overall, commercial banks in Indonesia have ROA of up to 5%. From the results of the regression conducted, the three variables, namely size, profitability and CAR do not significantly influence management decisions to pay more attention to and consider the company's reputation. Each variable has the following coefficients. Size is 0, 0000, while Profitability is -1.66 and CAR is 0.04. What this means is that if there is a size increase of 0, 0000, the possibility of management caring about the risk rises by 0.0000. While on profitability there is a negative relationship. Which means that the increase in profits actually causes management to ignore the reputation factor. While the CAR is positive 0.04 means that the company if it has a CAR increase of 1 unit then the increase in concern is equal to 0.04.

CONCLUSION
This study tries to see how the relationship between management concern for reputation risk with profit variables, company size and CAR. The study was conducted because the authors were interested in observing the extent to which reputation factors were of concern to many parties. Because it must be recognized, reputation is a risk that until now cannot be quantified.
From the results of the tests carried out, the three variables did not affect the reputation management of the banks. Therefore, it is necessary to conduct further searches related to variables that can explain management decisions in having more concern regarding reputation. Suggestions from this study are further research using variables other than the three variables above. Because the use of the above variables is proven not to significantly influence management decisions to have risk awareness. Example: