Investigating Bank Capital on Firm Rating Analysis

. This paper examined the correlation between firm ratings of banks and financial risks assessment of PEFINDO using profitability, asset quality, and liquidity as independent variables. This paper also investigated the impact of bank capital as controlling variable on bank ratings. The data which have been observed are financial reports from publicly-held banks in Indonesia and firm rating analysis released by PEFINDO during 2017-2021 consecutively, then was analyzed with regression model. This paper finds that profitability, asset quality and liquidity have no correlation with bank ratings without the existence of bank capital. As bank capital are taken into account, the correlation analysis had a significant difference and bank capital becomes determining factor in bank rating analysis.


Introduction
Leverage is an essential part of banking and it is higher far compared to another industry sector due to intermediary role which banks are highly leveraged that are in the business of facilitating leverage for others [10]. Besides raising funds from deposits, banks could also raise funds from capital market, like offering shares to public or issuing debt securities.
Credit rating agencies help to translate the banks' ability to meet their debt obligations into a rating scale. Integrity of credit rating agency and also public accounting and tax firms are highly required so those institutions could mediate the asymmetry information between firms or banks and their stakeholder, like investors that use any financial published information from credit rating agency and public accounting firm as considerations to their strategic investment or financial decisions [8].
One of the rating agencies and acknowledged by the capital market regulator in Indonesia is PT Pemeringkat Efek Indonesia (PEFINDO). PEFINDO provide corporate ratings on non-financial institution and financial institutions such as banks, finance or insurance companies, securities firms, as well as any specific debt instruments issued. The rating analysis released by PEFINDO is only valid for twelve months and could be reassessed for the next period.
According to PEFINDO's rating criteria and methodology, the key success factors for banking industry are business risk assessment and financial risk assessment. This paper used financial risk assessment of profitability, asset quality, liquidity to examine their * Corresponding author : vica.kaparang@unima.ac.id correlation to firm ratings of banks and investigated bank capital to the extent controlling the bank ratings.
This paper is different from existed research which observed ratings of capital market instrument or securities issued by banks, like bond ratings, meanwhile firm rating of banks or bank ratings as dependent variable are used in this paper.

Literature Review
Signalling theory is a powerful theoretical foundation in management research as it assists in explaining decisions by managers since signal receivers will react when the signaller is credible due to information asymmetry [18]. Good governance reduced conflict between managers and stakeholders [3] and emphasizes the interest of stakeholder to have precise and trusted information from bank's management [11].
Regarding this paper, PEFINDO is appointed by bank to conduct credible rating analysis and rating results will be released as public information. The highest rating of Pefindo is idAAA and the lowest is idD and meanwhile minimum investment grade recognized by the regulator is idBBB-. These ratings define gradually the obligor's capacity to meet their long-term financial commitment relative to other obligors.
The probability of future cash flows to the firm determined firm credit ratings as when the likelihood default increase so the firms credit ratings will decline [3]. In contrary, the assessment on credit ratings highly tend to rely on historical data considering bank will be downgraded after unfavourable financial information were known by interested parties [8].

Profitability and Bank Ratings
Profitability analysis is one of the financial risk assessments of PEFINDO's rating criteria and methodology described as assessments on net interest income and margin, non-interest income, quality of earnings, cost structure, and also management strategy to control operational expenses and improve fee-based income of banks.
The correlation between bond ratings issued by banks listed on the Indonesia Stock Exchange and financial performance of banks using ROA which represented bank profitability had been examined then suggested that ROA has a positive and significant correlation with bond ratings [2]. The bond ratings determinants had been examined with ROA as profitability measurement and suggested that ROA has a positive correlation but insignificant [7].
Considering ROA were used in the previous studies as ratio to measure bank profitability so the first hypothesis proposed in this paper or H1 is: ROA has a positive correlation to bank ratings.

Asset Quality and Bank Ratings
PEFINDO defined asset quality analysis as intensive assessments on the bank's non-performing loans, bank's loan loss reserve policy and adequacy.
Existed papers on bonds rating of banks with NPL as one of independent variables then suggested that NPL have a significant and negative correlation to bond ratings [2], [14], [16]. Small NPL indicates that banks performed a well management of their assets [2].
Considering NPL were used in the previous studies as ratio to examine asset quality, so the second hypothesis proposed in this paper or H2 is: LDR has a negative correlation to bank ratings.

Liquidity and Bank Ratings
PEFINDO defined liquidity analysis as assessments on current market condition and its effect on the bank's liquidity, examination on the bank's liquidity management, bank's interest rate and maturity mismatches, net open position, loan to deposit ratio and evaluation on the proportion of the bank's liquid assets as compared to its short-term liabilities are also incorporated in the assessments.
Existed paper used LDR as independent variable and suggested that LDR had a positive and significant correlation with bonds rating [16] while other examined LDR which represent liquidity risk and its correlation with bond ratings then suggested that it is positive and significant [17].
Considering that LDR were used in the previous studies as a ratio to examine bank liquidity, so the third hypothesis proposed in this paper or H3 is: LDR has a positive correlation to bank ratings.

Bank Capital and Bank Ratings
PEFINDO defined capital analysis as assessments on the bank's capital composition, level of capital adequacy ratio (total and Tier 1), internal growth rate of capital, and capital in comparison with assets Risk-based capital adequacy ratios have been a base of Basel framework as measurement of sufficient capital of banks relative to risks which simply explains that banks should have higher capital to compensate risk as banks take higher risks from their business [10]. Bank capital could be used in forecasting actual bank ratings as well as assets, financial gain or losses from securities, operating income, and yield on earning assets [8].
Other paper examined banks size by using the logarithm of total assets as controlling variable and resulting a correlation between size and bond ratings and also suggested that CAR has a positive and significant correlation to bond ratings [16]. As well as CAR increases, good corporate governance also improves [5].
In contrary, another paper suggested CAR has significant negative correlation to bond ratings. Capital is definitely needed to absorb risks but excess of capital will influence banks' ability of profit making [2]. Meanwhile other paper argued that CAR has no correlation to bond ratings [14].
Currently in Indonesia, banks are grouped into 4 (four) KBMI or bank groups based on Tier 1 capital according to OJK Regulation No. 12/POJK.03/2021 concerning Commercial Banks. This paper used Tier 1 Capital as a controlling variable representing bank capital.

Results and Discussion
This paper used purposive sampling which observed only publicly-held banks in Indonesia and analyzed their financial report as of December 31, 2017-2021. This paper also observed corporate rating analysis released by PEFINDO during 2017-2021 consecutively. Based on those criteria, 12 out of 47 banks have been data sample.
The ratings and scales used as dependent variable in this paper are as follows: This paper used statistical descriptive and multiple regression analysis to examine the hypotheses.
Followings are descriptive statistical analysis as a cross-section data series Valid N 60 equivalent to the BBB+ rating and the maximum value is 7.00 equivalent to an AAA rating with a proportion of 58.33% or 35 of the total 60 observed samples.
The minimum value of profitability as measured by the ROA ratio is 0.07% and the maximum value of profitability is 4.22%. The average ROA ratio of data sample was 1.81%.
The minimum value of asset quality as measured by the NPL ratio is 1.12% and the maximum value of asset quality is 5.65%. The average NPL ratio of data sample was 2.92%.
The minimum liquidity value as measured by the LDR ratio is 56.47% and the maximum value of liquidity is 113.50%. The average LDR of data sample was 86.63%.
The minimum value of bank capital as measured by natural logarithm of Tier 1 Capital is 15.75 equivalent to Rp6.90 trillion. The maximum value of bank capital is 19.26 equivalent to IDR 231.98 trillion.
Based on the results of the Kolmogrov One-Sample test, residual unstandardized significance value of 0.2 was obtained from selected only 46 sample data which has been fit to normality data assumption with a significance value greater than 0.05.
Based on the results of the multicollinearity test, the independent variables and controlling variable are greater than 0.1 for tolerance values and less than 10 for VIF values and suggested that independent variables and controlling variable used in this paper has no correlation.
The heteroscedasticity test in this paper used the Glejser test which conducted by regressing independent variables and controlling variable to their absolute value of unstandardized residual. The significance value for all independent variables and controlling variable was obtained greater than 0.05 or 5% and suggested that this paper had met the assumption of heteroscedasticity. The value of Adjusted R2 without controlling variable is 0.00 or 0% and the value of Adjusted R2 with controlling variable is 0.570 or 57.0% and thus explained controlling variable contributes to the correlation between independent variables and dependent variable while remainders are influenced by other variables outside All independent variables have significance values greater than 0.05 or 5% with a coefficient for variable ROA of 0.113 and the coefficients of NPL and LDR by 0.08 and 0.008, respectively. These results confirmed to reject H0 from alternatives hypothesis in this paper, H1, H2, H3, which predict that there were correlations between independent variables and dependent variable. .065 This result confirmed that the coefficients and significance values of some independent variables have changed. ROA turned to has a negative coefficient of 0.39 with a significance less than 0.05 or 5%. NPL also turned to a negative coefficient of 0.147 with a significance value of 0.038. The Tier 1 Capital variable https://doi.org/10.1051/shsconf/202214903018 , 03018 (2022) SHS Web of Conferences 149 ICSS 2022 as a controlling variable has a coefficient of 0.489 with a significance less than 0.05 or 5%.

Conclusions
The results of this paper show that the existence of Tier 1 capital as controlling variables to the model resulted a significant difference of the correlation of ROA, NPL, and LDR as independent variables on the bank ratings as dependent variable which indicated by the changes in the constants and coefficients of the regression equation. Bank capital becomes the only determining factor in bank rating analysis even when profitability, asset quality, and liquidity are all zero by chance.
Rating agencies systematically assigned favourable ratings to larger banks [9]. The greater the growth, firms will have favourable investment grade of bond ratings [1]. The higher the value of Tier 1 Capital, the higher the risk management ability, profitability, and performance of the bank [16].
Banks as highly-leveraged financial institutions are faced with various challenges of business complexity and volatility as well as various internal and external risk exposures so that the bank's capital becomes vital to business sustainability. In addition, bank capital can be analogous to fuel to support the bank's business expansion, cover risks, and make a profit. Larger banks indicate a larger source of funding and higher ability to disburse financing or investment as well as funding access [16]. Greater bank capital not only reduces financial distress but also liquidity creation, otherwise an optimum bank capital structure trade-off of cost of bank distress and liquidity creation [4].
The results of this paper show that the correlation between ROA and the corporate rating are significant as if Tier 1 Capital as controlling variable are taken into account. Other paper did not find any correlation between ROA and bond ratings [14] while other also find that profitability has a negative correlation to bond ratings, but insignificant [6]. However, the results of this paper are contrast to existed paper which examined profitability using ROA ratio and had a positive correlation to bond ratings [2], [7].
The change in the ROA coefficient to negative in this paper can be explained that if assuming Tier 1 Capital reflects value of bank assets, then higher capital also reflect complexity in managing its assets and definitely would require a tighter supervision. This is related to the higher systemic impact of a bank with large asset value.
NPL has correlation with corporate rating and its coefficient changes to negative when bank capital are taken into account to the analysis. Other studies find negative coefficient of NPL [14], [16]. Less NPL ratio indicates bank is capable to manage assets well [2]. Meanwhile, banks with lower good corporate governance score tends to have larger NPL [12].
Regarding asset quality, quantitative calculations or using NPL ratio could not be sole analysis for corporate rating. In this case, qualitative justifications that are more reliable in ratings analysis such as credit portfolios by business sector or assessments of credit concentration, bank policies and procedures related to reserves or write-offs, and other qualitative disclosures that can determine the bank's asset quality.
LDR has no correlation with corporate rating, with or without controlling variable. In contrary, LDR has a positive and significant correlation to bond ratings [6], [7], [16].
LDR is ratio that represent the core business of banks so liquidity is assumed to be stable as long as there is no financial distress. This also assumed that the operational activities of banks regarding asset and liabilities management, particularly interest gaps, maturity mismatches and foreign exchange have been anticipated for liquidity risk aligned with OJK Regulation Number 42/POJK.03/2015 as an implementation of Basel III for bank liquidity standards.
The purpose this paper is to investigate the extent of Tier 1 capital controlling bank ratings using financial risk assessment. Profitability, asset quality and liquidity have no correlation with bank ratings without existence of bank capital. Bank capital has a significant role to the correlation of bank ratings and financial risk assessment rating using PEFINDO's rating methodology.
There are several extensions regarding the limitations of this paper. First, this paper only used banks with public ownership. Second, this paper used financial risk assessment as an independent variable without considering the business risk assessment, such as market position, infrastructure and quality of service, and corporate governance.