The effect of good corporate governance, firm size, leverage and profitability on accounting conservatism level in banking industry

This study aims to examine the effect of good corporate governance, firm size, leverage, and profitability on accounting conservatism level. The population used in this study is all banking industry companies, listed on the Indonesia Stock Exchange (BEI) 20082015, which present good corporate governance reports completely. The sampling technique used in this research is purposive sampling method. There are 18 companies qualified as the research sample. Data analysis method used is multiple linier regression analysis using SPSS 23 for Windows program, with the significance level used is 0.05. The results show that good corporate governance and firm size have no significant effect on accounting conservatism level, while leverage and profitability have significant effect on accounting conservatism level.


INTRODUCTION
Financial statement is a form of management accountability that describes the performance and ability of a company in managing its resources. The company is said to be successful if it is able to survive and has a good performance. In preparing the financial statement, the Financial Accounting Standards (SAK) have provided the company the freedom to choose the method or estimation used in accordance with the conditions of the company itself so that it can be more flexible in adjusting to the economic conditions experienced by the company. Future economic conditions are full of uncertainties and risks so that companies are expected to be more cautious (conservative) in choosing the accounting method to be used. Research on conservatism has been done a lot but the results are still inconsistent. Therefore, this research is expected to provide information and development of the concept of conservatism in the future.
The definition of conservatism, based on the glossary in the FASB Statement of Concept No. 2, is a prudent reaction in the face of uncertainty to ensure that the uncertainty and risk inherent in the business situation have been adequately considered well (Fani and Kusmuriyanto 2015). Conservatism in this study is assessed using the ratio of marketto-book value, which is a comparison between the company's stock market value and the book value, where the higher the market-to-book value ratio (more than 1), the more conservative the company in presenting its financial statements.
An increase in stock prices of some banking industries may result in an increase in market-tobook value ratio. The better the performance of a company, the higher the price of the company's stock, and this is evidenced by the increase in corporate profits. A company with better performance will attract the attention of investors so that the stock market price of the company is increasing. The rising of the stock market price will increase the ratio of market-to-book value. This means that the increase in this ratio indicates that the company is increasingly conservative in presenting its financial statements. There are several factors in determining the manager's decision to be more careful (conservative) in presenting his financial statements. In this study, the factors used to determine the level of accounting conservatism in the company is good corporate governance (GCG), firm size, leverage, and profitability.
The indication of good corporate governance (GCG) implementation is that everything related to the company is openly expressed in accordance with GCG principles, that is, transparent, reasonable and accountable. This can also be interpreted that in preparing financial statements, the management should be more careful (conservative) so that the information disclosed will always be in accordance with the conditions of the company. Another factor that affects accounting conservatism level is firm size. A large company has more complex problems and risks than the smaller one. So, large company will be subjected to high political costs. And the company's political costs can be reduced using accounting conservatism (Desak and I Gusti 2013). In addition, leverage is also able to affect the level of accounting conservatism. The company with high leverage level will be less conservative in presenting its financial statements, because the company tends to increase its earnings so that the financial conditions look good in the eye of creditors. Profitability is also one of the factors that affect the accounting conservatism level. A company with a high profitability will be more conservative in presenting its financial statements. This is because conservatism is used by management to manage earnings to look stable.

THEORETICAL FRAMEWORK AND HYPO-THESIS Agency Theory
According to Jensen and Meckling (1976), agency theory is a relationship between the agent and the principal, based on a set of contracts, where the agent, as a manager, is authorized by the principal, as the owner of the fund (investor), to run the company and make decisions. Through the delegation of this task, the relationship between agent and principal will result in information asymmetry, where the agent knows more information about the company than the principal.
The agency theory has a key element in which the principal and the agent have different preferences or goals. The principal is motivated to prosper himself through the profitability of the company, while the agent is motivated to maximize his economic and psychological needs, among others, in terms of obtaining compensation and loans (Fani and Kusmuriyanto 2015). These differences may lead to the conflict of interest between the agent and the principal.
Based on the description of the agency theory above, it can be said that the conservatism in presenting the financial statements applied by the management will maintain the relationship between the parties related to the company because conservatism is able to reduce the agency conflict which one of them is caused by the presence of information asymmetry among the parties who make transactions with the company.

Signaling Theory
Signaling theory explains the company's drive to provide financial statement information to external parties (Rheny and Denies 2012). According to Najmudin (2011: 308), signaling theory is an action taken by the company management to provide signal or clue to investors on how they assess the prospects of the company. The relationship between signaling theory and this research is that this theory gives a signal of manager's confidence on the future prospects of the company that can be assessed from the extent to which the manager applies the principles of conservatism by selecting accounting methods applied in the preparation of financial statements so that the asset or profit presented in the financial statements is not too large (overstate), and the investors will not expect too large financial compensation from the company.

Stewardship Theory
Stewardship theory is a theory built on human philosophical assumptions about human nature that human is essentially trustworthy, capable of acting responsibly, and having integrity and honesty with others (Thomas 2006). This means that stewardship theory believes that management is able to act in accordance with the interests of the public or stakeholders by creating good corporate governance so that the company's performance will be improved by monitoring the performance of management.
The relationship of stewardship theory with this research is the management is seen as a trustworthy, responsible and integrity party and capable of implementing good corporate governance as well as possible in accordance with the interests of the stakeholders and society so that the financial statements made by the management can provide quality information for the stakeholders.

The Relationship between Good Corporate Governance and Accounting Conservatism
A company that has self-assessment result on good GCG implementation in accordance with Circular Letter of Bank Indonesia No. 9/12/DPNP/2007 of less than 1.5 will improve its corporate image so that the consumer loyalty to the company will also be higher. Increased consumer loyalty will affect the increase in revenue and profitability of the company so that when corporate profitability increases, the company will have an appeal in the eyes of investors. An increase in attractiveness of investor will increase the market price of shares owned by the company so that the market-to-book value ratio, which is the ratio between stock market prices and book value, will increase. Market-tobook value ratio is a proxy for measuring accounting conservatism. The increasing ratio of market-tobook value indicates that the company is becoming more conservative. This statement is supported by the results of previous research conducted by Dariush et al. (2013) and Fani and Kusmuriyanto (2015) that good corporate governance has an effect on accounting conservatism. H1: Good Corporate Governance has an effect on Accounting Conservatism.

The relationship between Firm Size and Accounting Conservatism
The larger the size of a company, the greater the burden of welfare transfers (political costs) incurred by the company. This is because large company will be subject to higher tax rates and gets more demands from workers. The growing political costs indicate that the company has a high profitability. The higher the profitability of the company, the bigger the profit earned by the company. Increased corporate profits will increase the attractiveness of the public, especially investors to the company. Increased attractiveness of the public, especially investors, will increase the stock market price of the company. The increased stock market price will increase the value of the market-to-book value ratio, which is the proxy for measuring accounting conservatism. The increased ratio of market-to-book value indicates that the company is more conservative. Research conducted by Ni Wayan and Ni Made (2015) shows the result that firm size has an effect on accounting conservatism. H2: Company Size has an effect on Accounting Conservatism.

The Relationship between Leverage and Accounting Conservatism
Companies with high leverage levels will have a high level of financial risk to creditors and shareholders. This high risk can make the company's attractiveness decline in the eyes of potential investors. The decline in the attractiveness of society and investors to the company will lower the market price of the company's stock. The decline in stock price will lower the market-to-book value ratio which is a proxy of conservatism measures. The decline in market to-book-value ratio indicates that the company is becoming less conservative. Research conducted by Ni Wayan and Ni Made (2105) states that leverage has an effect on accounting conservatism. H3: Leverage has an effect on Accounting Conservatism.

The Relationship between Profitability and Accounting Conservatism
The higher the profitability of a company, the greater the profit earned by the company. The increased company profit will increase the attractiveness of the public, especially investors to the company. The increased attractiveness of the public, especially investors, will increase the company's stock market price. The increased stock market price will increase the value of market-to-book value ratio, which is the proxy of measuring accounting conservatism. So, the increased market-to-book value ratio indicates that the company is more conservative in presenting its financial statements. Research conducted by Radyasinta and Kusmuriyanto (2014) states that profitability has an effect on accounting conservatism. H4: Profitability has an effect on Accounting Conservatism.

RESEARCH METHOD Research Design
Based on the type of research, this research is a hypothesis testing research and causal research (Jo-giyanto 2015: 70). Based on the nature and type of data, the data used in this study is included in the metric data with quantitative value in which the type of scale or type of data value used in this study is the ratio scale and using secondary data (Jogiyanto 2015: 81). Secondary data used are the data obtained from financial statements, annual reports, and good corporate governance reports of banking industry companies listed on the Indonesian Stock Exchange period 2008-2015.

Research Limitation
There are several limitations in this study, such as: 1. This study examines the effect of good corporate governance, firm size, leverage, and profitability on the level of accounting conservatism only. 2. The industry used in this research is conventional banking industry only. 3. The period of research is from 2008 to 2015 only.

Variable Identification
The variable used in this study consists of two variables, namely dependent variable and independent variable. 1. The dependent variable (Y) is Accounting Conservatism 2. The independent variables (X) are Good Corporate Governance (X1), Firm Size (X2), Leverage (X3) and Profitability (X4).

Operational Definition and Variable Measurement Accounting Conservatism
Conservatism is an attitude or flow to take action or decision cautiously in the face of risk by being willing to sacrifice something in order to reduce or eliminate risk (Suwardjono 2013: 245). The conservatism in this research uses measurement of market-to-book value ratio, which refers to the research conducted by Fani and Kusmuriyanto (2015).

Company Size
Company size is one of indicators that can be used to determine the size of a company. In this study, the size of the company is proxied by the natural logarithm of total assets owned by the company, which refers to the research conducted by Jumratul and I Dewa (2014).

Leverage
Leverage is a ratio that shows the extent to which the company uses debt to finance the company's assets. In this study, the proxy used to measure leverage is total debt divided by total assets which refers to the research conducted by Ni Wayan and Ni Made (2015): = . (2)

Profitability
Profitability is the ratio used to assess the company's ability to generate profits through total assets owned by the company. In this study, the proxy used to assess profitability is using ROA (Return on Asset) that compares between Net Profit and Total Asset owned by the company in accordance with the research conducted by Radyasinta and Kusmuriyanto (2014): (3)

Population and Sampling Technique
The

Data Analysis Technique
This research used descriptive statistical analysis, multiple regression analysis, classical assumption test, and hypothesis test.

Descriptive Statistical Analysis
According to Imam (2016: 19), descriptive statistical analysis gives descriptions of data seen from the values of mean, standard deviation, variance, minimum, maximum, sum, range, kurtosis and skewness. This analysis provides an overview of the distribution and behavior of the sample data.

Classical Assumption Test Normality Test
Normality test is used to test whether in regression model, the dependent and independent variables have normal or abnormal distribution. This model is said to be good if the distribution is normal or close to normal. The statistical test used in this study is Kolmogorov-Smirnov Test. Kolmogorov-Smirnov Test has certain criteria in the assessment, such as: 1. If the significance is > 5%, then the data is normally distributed. 2. If the significance is <5%, then the data is not normally distributed.

Multicolinearity Test
In a good regression model, there should be no correlation between independent variables. This multicollinearity test is used for the research that has more than one independent variable. This can be seen from the value of tolerance and variance inflation factor (VIF). If the tolerance value is > 10% and VIF value is < 10%, then it can be concluded that there is no multicollinearity among independent variables in the regression model.

Autocorrelation Test
According to Imam (2016: 107), autocorrelation test aims to test whether in the linear regression model there is a correlation between the confounding error in period t and the disturbance error in period t 1 (previously). This study uses Run Test. The hypotheses used for this test are: 1. H 0 : Residual is random, which means that there is no autocorrelation between residual values 2. H a : Residual is not random, which means that there is autocorrelation between residual values. The data will be seen at significant value of > 0.05. If the significance value > 0.05, it can be concluded that in the regression model there is no autocorrelation.

Heteroscedasticity Test
According to Imam (2016: 134), heteroscedasticity test aims to test whether there is a variance inequality of residual from one observation to another in a regression model. A model can be said to be free from heteroscedasticity if the β parameter coefficient of the regression equation is not statistically significant or the probability value of significance is above 5 percent of confidence level.

Multiple Linear Regression Analysis
Regression analysis is performed to measure the strength of the relationship between two or more variables. In addition, it also shows the direction of the relationship between the dependent variable and the independent variables (Imam Ghozali 2011: 96). In this study, the researchers tested the influence of independent variables on dependent variable using multiple linier regression analysis method with mathematical equation as follows: KON_AK = α + β 1 GCG + β 2 SIZE + β 3 LEV + β 4 ROA + ε. (

DATA ANALYSIS AND DISCUSSION Descriptive Statistical Analysis
In this study, descriptive statistical analysis will explain and describe the data seen from the minimum, maximum, mean, and standard deviation values. The following is the explanation of the descriptive analysis from the result of descriptive statistical analysis (see Table 1): The maximum value of 3.6587 is obtained by Bank Central Asia Tbk (BBCA) in 2015. The high value of accounting conservatism is due to the high ratio of market-to-book value. The closing price of BCA bank was IDR 13,300, higher than the book value of IDR 3,635.162, resulting in a market-tobook value ratio of 3.6587. This indicates that the company has a high level of accounting conservatism in its financial reporting because it means the company records its book value lower than the market value of the company. Standard deviation value of accounting conservatism from is 0.6475693 with a mean value of 1.381490. The standard deviation value of accounting conservatism variable in this research is much smaller than the mean value obtained for accounting conservatism variable. This means that the variation of data in this research is considered low or homogeneous, so it can be concluded that the data distribution is good.
Of the 122 samples studied, 53% of companies have accounting conservatism value below the average during the period 2008-2015 and the remaining 47% of companies have accounting conservatism value above average, thus it can be interpreted that the average banking company has a low level of accounting conservatism. The lower ratio of market-to-book value makes the company recognize that the book value is higher than the market value of the company

Good Corporate Governance
Based on the output of descriptive statistical test (Table 1) The average value of good corporate governance in the banking industry based on Table 1 is 1.653238. Of the 122 samples studied, 48% of the companies have a composite value of GCG below the average, while the remaining 52% of the companies have a composite value of GCG above the average. The standard deviation of good corporate governance is 0.4255754. The standard deviation value, which is far below the average, indicates that the data variations in the study are low.
However, for the entire data processed shows that the average company has a good predicate of good corporate governance that can be seen from the result of statistical test with the average composite value of 1.653238. And the average composite value is still in good category because the composite value is still less than 2.5. So, it can be concluded that the average of banking companies that have been sampled in this research have good governance.

Firm Size
The minimum value of firm size is 27.9384 derived from the total logarithm of the total asset of Bank of India (BSWD) in 2008 with total asset of IDR 1,359,880,323,678. While the maximum value of the company size is 34.4445 came from the total natural logarithm of the total assets of Bank Mandiri (Persero) Tbk in 2015 with total assets of IDR 910.063.409.000.000.
Based on Table 1, it can also be seen that the mean value of the firm size of 122 samples is 31.763050 with a standard deviation value of 0.6475693. The standard deviation value of firm size in this research is much smaller than the mean value obtained for company size variable, which means that the data variations in this research are low or homogeneous data. So, it can be concluded that the data distribution is good.

Leverage
The minimum value of leverage is 0.3422 obtained by Bank QNB Indonesia Tbk in 2013 with total debt of IDR 3,781,586,000,000 and total assets of IDR 11,051,347,000,000. This means that the company's lowest ability to finance its assets using debt is 34.2%. While the maximum value of 0.9398 was obtained by Bank Artha Graha International Tbk in 2011 with total debt of IDR 18,031,095,213,642 and total assets of IDR 19,185,436,308,366. This means that the company's highest ability to finance its assets is 93.98%. Thus, it can be said that the companies with high leverage will get more supervision from creditors than the companies with low leverage.
Overall, the mean value of leverage is 0.873263 with a standard deviation value of 0.0663636. Standard deviation value can be used to see how far the distance between one data and another. The data variations of the leverage can be said to be low because the standard deviation is smaller than the mean value of 0.0663636.

Profitability
The minimum value of 0.0005 was obtained by Bank QNB Indonesia Tbk in 2011 with net profit before tax of IDR 15,550,000,000 and total assets of IDR 3,593,817,000,000. This means that the company's lowest ability to generate profits through its assets is 0.05%.
While the maximum value of .0446 was obtained by Bank Rakyat Indonesia (Persero) Tbk in 2013 with net profit before tax of IDR 27,910,000,000,000 and total assets of IDR 626,100,633,000,000. This means the company's highest ability to generate profit through assets owned by the company is 4.6%. From Table 1, it can also be seen that the mean value of profitability is 0.019599 with a standard deviation of 0.0106733. The standard deviation value which is smaller than mean value indicates that the data variations in the study are low or homogeneous data. So, it can be concluded that the data distribution is good.
High ROA (Return on Asset) ratio indicates that the level of company's ability to generate profits derived from the company's assets is getting better. This means that management is able to optimize the use and utilization of assets owned by the company. Conversely, low ROA (Return on Asset) ratio shows that the level of company's ability to generate profits derived from company assets is getting worse. This means that management is not able to optimize the use and utilization of assets owned by the company.

Normality Test
The results of normality calculation by using One-Sample Kolmogorov-Smirnov Test show that the initial data with a sample of 144 companies has the value of Asymp. Sig. (2-tailed) of 0.000. This indicates that the value of Asymp. Sig. (2-tailed) is 0.000 smaller than the significant coefficient value of 0.05 (0.000 < 0.05). So, it can be concluded that the residual data in the regression model is not normally distributed. Since the researchers use outlier test, the researchers have to re-test the data that is considered outlier. Outlier test in this research is done 7 times with total outlier data as many as 22 data and then generate final sample of 122 companies during period 2008-2015 with significance value of 0.065 so that the data have normal distribution.

Multicolinearity Test
The multicollinearity test is used to test whether there is a correlation between independent variables in the regression model (Imam 2016: 103). Tolerance value indicates that there is no independent variable having tolerance value less than 0.1, which means that there is no correlation between independent variables whose value is more than 95%. As for the calculation of Variance Inflation Factor (VIF) value also shows the same thing that, from the four independent variables, there is no single variable that has VIF value more than 10. So it can be concluded that there is no multicollinearity between independent variables in the regression model.

Autocorrelation Test
Autocorrelation test is used to test whether in the regression model there is a correlation between the confounding error in period t and the disturbance error in the previous period (t-1). In this study, the researches use Run Test to detect whether there is autocorrelation problem in the regression model used. The results of Run Test table 4.6 show the test value of -0.05809 with Asymp value. Sig. (2-tailed) of 0.585. It shows that the value of Asymp. Sig. (2tailed) is 0.585 greater than the significant coefficient value of 0.05 (0.585 > 0.05). This means the hypothesis is nil and residual data is random, so it can be concluded that in the regression model there is no autocorrelation.

Heteroscedasticity Test
Heteroscedasticity test in this research is done using Glejser test. The results of Glejser test show that the significance value of GCG variable is 0.252, company size is 0.232, Leverage is 0.749, and profitability is 0.070. The four independent variables have significance value above 0.05, which the value of Residual Absolute (AbsUt) of dependent variable. This means that the significant value statistically affects the absolute residual value (AbsUt) of dependent variable. So it can be concluded that heteroscedasticity does not occur in the regression model of this study.

Multiple Regression Analysis
In this study, multiple regression analysis is used to measure the strength of the relationship between independent variables (good corporate governance, firm size, leverage, and profitability) and dependent variable (accounting conservatism) either partially or simultaneously. From the results of data processing, it is obtained multiple regression equation as follows: KON_AK = 2.339 -0.181 GCG + 0.036 SIZE -2.8000 LEV + 32.316 ROA + ε.

Hypothesis Test F Test
F statistic test is performed to find out whether the regression equation model in this research is fit or not fit. The result indicates the value of F test is 23.310 with a significance value of 0.000 which is much smaller than the level of significance (α) 0.05. This means H 0 is rejected and H 1 is accepted. So it can be concluded that there is one of the independent variables (good corporate governance variable, firm size, leverage and profitability) that has an effect on accounting conservatism, then the regression model is said fit or good.

R 2 Test
Determination coefficient (R 2 ) is used to measure the ability of the model (the influence of independent variables) in explaining the variation of the dependent variable. The results indicate the value of adjusted R square of 0.424 which means that the variable of accounting conservatism can be explained by the variables of good corporate governance, firm size, leverage, and profitability with the value of 42.2%, while the remaining 57.8% is explained by other variables that are not investigated in this study.

T Test
T statistics test aims to show the influence of one (each) independent variable partially on the dependent variable. The results of the t test can be seen in Table 2.

a. The First Hypothesis Test
The first hypothesis test is conducted to examine the effect of good corporate governance on accounting conservatism. Based on table 8, the t count value is -1.387 with the significance level of 0.168. The significance level of 0.168 is greater than 0.05 and the t count value of -1.387 is smaller than the t table value of 1.9084. It can be concluded that in this study H 0 is accepted and H 1 is rejected, which means that good corporate governance (GCG) has no effect on accounting conservatism.

b. The Second Hypothesis Test
The second hypothesis is conducted to examine the effect of firm size on accounting conservatism. Based on table 8, the t count value is 1.040 with the significance level of 0.301. The significance level of 0.301 is greater than 0.05 and the t count value of 1.080 is smaller than the t table value of 1.9084. It can be concluded that in this study H 0 is accepted and H 1 is rejected which means that company size has no effect on accounting conservatism.

c. The Third Hypothesis Test
The third hypothesis test is conducted to examine the effect of leverage on accounting conservatism. Based on table 8, the t count value is 3.845 with the significance level of 0.000. The significance level of 0.000 is smaller than 0.05 and the t count value of 3.854 is greater than t table value of 1.9084. It can be concluded that in this study H 0 is rejected and H 1 is accepted, which means that leverage has an effect on accounting conservatism.

d. The Fourth Hypothesis Test
The fourth hypothesis is conducted to examine the effect of profitability on accounting conservatism. Based on table 8, the t count value is 6.427 with the significance level of 0.000. The significance level of 0.000 is smaller than 0.05 and the t count value of 6.427 is greater than t table value of 1.9084. It can be concluded that in this study H0 is rejected and H1 is accepted which means that profitability has an effect on accountancy conservatism.

Discussion The Effect of Good Corporate Governance on Accounting Conservatism
Based on the result of t test analysis, it is found that good corporate governance has no effect on ac- counting conservatism. Table 2 shows that the average composite value of good corporate governance in banking companies that apply accounting conservatism during the study period 2008 -2015 is 1.5472. Meanwhile, the average composite value of good corporate governance in banking companies that do not apply accounting conservatism is 1.8481. The results indicate that based on the data processed, both conservative companies and non-conservative companies have good GCG. This can be seen from the average composite value obtained from the observation, that is, from 1.5 to 2.00. In accordance with Bank Indonesia Circular Letter No. 09/12/DPNP/2007, companies that have such composite value are companies that implement good GCG. So it can be concluded that the implementation of good corporate governance does not affect the level of accounting conservatism applied by the company in the presentation of its financial statements.
The result of the test which states that there is no influence of good corporate governance on accounting conservatism, when it is related with the measurement of conservatism in this research with the ratio of comparison between closing price and book value per share, the composite value of good corporate governance will not affect the company stock price. It is because either big or small composite value obtained from the report of self assessment of good corporate governance is not the main focus of the investors. The main focus of investors is on the company's financial performance. The reason is that with a good financial performance, the investors will most likely get bigger dividends and returns. Therefore, the composite value has no effect on the fluctuation of the stock price causing the value of the market-to-book ratio, which is the proxy for the measurement of accounting conservatism, also has no effect.
The results of this study are consistent with the research conducted by Mariska Veres (2013) which can not prove the influence of good corporate governance on accounting conservatism but different from the results of the research conducted by Fani and Kusmuriyanto (2015) that good corporate governance has an effect on accounting conservatism .

The Effect of Firm Size on Accounting Conservatism
Based on the result of t test analysis, it is found that company size has no effect on accounting conservatism. Table 3 shows that the average value of    These results indicate that from the data processed, both conservative and non conservative companies have almost the same company size. This can be seen from the difference in the average value of natural logarithm of total assets between companies that implement accounting conservatism and companies that do not apply accounting conservatism, which is not too large. So, it can be concluded that company size does not affect the level of accounting conservatism.
If it is related to the measurement of conservatism in this study, which is the ratio between the closing price and the book value per share, the company size will not affect the company stock price because generally the main orientation of the investor is on dividend and return. Meanwhile, the size of company (big or small) will not necessarily benefit investors, because there is possibility that the profit earned by the company will be reinvested to the company for the purpose of business expansion. So it can be concluded that the size (big or small) of companies will not affect the fluctuation of stock prices so that the value of the ratio of market-to-book value, which is the proxy for measurement of accounting conservatism, will also has no effect.
The results of this study are consistent with the results of the research conducted by Mochammad and Darsono (2015) which can not prove the effect of firm size on accounting conservatism, but not consistent with the research conducted by Ni Wayan and Ni Made (2015) which indicates that firm size has an effect on accounting conservatism.

The Effect of Leverage on Accounting Conservatism
Based on the results of t test analysis, it is found that the leverage has an effect on accounting conservatism. Table 4 shows that the average value of leverage in banking companies that apply accounting conservatism during the study period 2008 -2015 is 0.8575. While in banking companies that do not apply accounting conservatism have an average leverage value of 0.9022.
The results indicate that, from the processed data, conservative companies have an average value of leverage of 0.8575 (85.75%) which is lower than the non-conservative companies which have an average value of leverage of 0.9022 (90.22%).
Based on these results, it can be concluded that leverage has an effect on the level of accounting conservatism. So the theory in this study which states that the higher the leverage of companies the more not conservative the company in presenting its financial statements. The results of this study are consistent with the research conducted by Ni Wayan and Ni Made (2015) which states that leverage has an effect on accounting conservatism but different from the results of the research conducted by Dini Prastiwi (2013) that leverage has no effect on accounting conservatism.

The Effect of Profitability on Accounting Conservatism
Based on the result of t test analysis, it is found that profitability has an effect on accountancy conservatism. Table 5 shows that the average value of profitability in banking companies that apply accounting conservatism during the study period 2008-2015 is 0.0235. While the average value of profitability in banking companies that do not apply accounting conservatism is 0.015.
These results indicate that, from the data processed, conservative companies have an average value of profitability of 0.0235 or 2.35% which is higher than non conservative companies that have an average value of profitability of 0.0125 or 1.25%. So, based on these data, it can also be proven that profitability has an effect on accounting conservatism.
This means that high level of profitability will bring a high tendency for companies to apply the principles of conservatism because the companies with high profitability try hard to keep the profit obtained does not look fluctuating. The results of this study are consistent with research conducted by Radyasinta and Kusmuriyanto (2014) that profitability has an effect on accounting conservatism, but the results of this study differ from the research conducted by Ika and Fachrurrozie (2015) which states that profitability has no effect on accounting conservatism.

CONCLUSION, IMPLICATION, SUGGES-TION, AND LIMITATIONS
This study aims to find out the effect of independent variables on accounting conservatism in banking companies. The data observed during 2008-2015 indicates that profitability variable and leverage variable have an effect on accounting conservatism. The large or small composite value obtained from the GCG self assessment report is not the main focus of the investors. The main focus of the investors is on the financial performance of the company so that the composite value will not affect the fluctuation of stock price which causes the value of market-to-book ratio, which is a proxy for measuring accounting conservatism, will also have no effect. In addition, the firm size shows that large companies generally will become the attention of various parties so that the large companies will create financial statements that describe the real condition of the company. Meanwhile, the principles of accounting conservatism are still considered by some parties as the principles that make the financial statements presented to be biased so that they have not been able to serve as a tool to evaluate corporate risk. Other results show that companies with higher leverage level have lower level of accounting conservatism. The companies with higher profitability have higher level of accounting conservatism.
This study has some limitations, among others are: 1) in this study, there are data outliers so that the results of the research achieved are less maximum than the expectation of the researchers; 2) the measurement of good corporate governance using eleven indicators has deficiencies. In 2013, 2014, and 2015, many banks used composite ratings for the GCG self-assessment due to the regulatory changes from BI.
Based on the results, conclusions and limitations of this study, future researchers are expected to: 1) use other statistical testing models that do not require the normal distribution data so that the data do not undergo the outlier process; 2) use a self assessment rating to measure good corporate governance variable in banking industry.