Credit Risk Management and Profitability in Commercial Banks

Credit Risk Management and Profitability in Commercial Banks in Sweden Ara Hosna, Bakaeva Manzura and Sun Juanjuan Graduate School Master of Science in Accounting Master Degree Project No. 2009:36 Supervisor: Inga-Lill Johansson Acknowledgements After several months of hard work our thesis has been finished. Now it is time to thank everyone warmly who provided their kind assistance to us. First of all, we would like to thank our supervisor Inga-Lill Johansson, Associate Professor of our University, for her guidance all through our work. We would like to thank Andreas Hagberg, PhD Candidate, as well for giving us his constructive suggestions.
We are grateful to Johan Sjomark, Credit Risk Control Department officer in Swedbank, for providing us helpful interview by using his wealthy knowledge in credit risk management area. The same appreciations are given to the risk management department in Swedbank, for arranging this interview. Furthermore, we would like to direct our appreciations to our opponent groups for providing us useful feedbacks. Last but not least, we are thankful to Eva Gustavsson and Wajda Irfaeya from Gothenburg University for facilitating us regarding the statistical analysis.
We also would like to express our thanks to IT and Library Services of the School, for providing professional software programs, books and databases. Without them, our thesis would not be finished. The dearest appreciations are directed to our families and friends, for giving us great support and help during these months. Special thanks to Sevara for being patient and “compassionate” to her mommy. Gothenburg, 24th of May 2009 ______________________ __________________________ ___________________________ Ara Hosna Bakaeva Manzura Sun Juanjuan ii Abstract

Credit risk management in banks has become more important not only because of the financial crisis that the world is experiencing nowadays but also the introduction of Basel II. Since granting credit is one of the main sources of income in commercial banks, the management of the risk related to that credit affects the profitability of the banks. In our study, we try to find out how the credit risk management affects the profitability in banks. The main purpose of our study is to describe the impact level of credit risk management on profitability in four commercial banks in Sweden.
The study is limited to identifying the relationship of credit risk management and profitability of four commercial banks in Sweden. The results of the study are limited to banks in the sample and are not generalized for the all the commercial banks in Sweden. Furthermore, as our study only uses the quantitative approach and focuses on the description of the outputs from SPSS, the reasons behind will not be discussed and explained. The quantitative method is used in order to fulfill the main purpose of our study. We have used regression model to do the empirical analysis.
In the model we have defined ROE as profitability indicator while NPLR and CAR as credit risk management indicators. The data is collected from the sample banks annual reports (2000-2008) and capital adequacy and risk management reports (2007-2008). The findings and analysis reveal that credit risk management has effect on profitability in all 4 banks. Among the two credit risk management indicators, NPLR has a significant effect than CAR on profitability (ROE). The analysis on each bank level shows that the impact of credit risk management on profitability is not the same.
The credit risk management of Nordea and SEB has relatively similar impact on their profitability. While Handelsbanken’s results indicate that NPLR and CAR are very weak or incapable of predicting ROE. In case of Swedbank NPLR and CAR explains the variances in ROE with very low probability. Basel II application has strengthened the negative impact of NPLR on ROE. Unlike effect of Basel I, CAR has positive and insignificant effect on ROE. Keywords: credit risk management, profitability, banks, Basel II iii Abbreviations Adj. R2 BCBS CAR CCF Coef. CRD FIRB FSA ICAAP IFRS IRB LGD N NI NPL NPLR PD P-value R2 ROA ROE RORAC RWA SFSA Signif.
TL TSE Adjusted R-squared Basel Committee on Banking Supervision Capital Adequacy Ratio Credit Conversion Factors Coefficient Capital Requirements Directives Foundation Internal Rating-based Financial Supervisory Authority Internal Capital Adequacy Assessment Process International Financial Reporting Standards Internal Rating-based Loss Given Default Number (of Observations) Net Income Non-performing Loan Non-performing Loan Ratio Probability of Default Probability Value R-squared Return on Assets Return on Equity Return on Risk Adjusted Capital Risk Weighted Asset Swedish Financial Supervisory Authority Significance Total Loan
Total Shareholders’ Equity iv Table of Contents 1. Introduction ……………………………………………………………………………………………………………… 1 1. 1 1. 2 Problem Discussion …………………………………………………………………………………………………………. 3 1. 3 Research question ……………………………………………………………………………………………………………. 4 1. 4 Purpose ………………………………………………………………………………………………………………………….. 1. 5 Delimitation ……………………………………………………………………………………………………………………. 4 1. 6 2. Background…………………………………………………………………………………………………………………….. 1 Disposition ……………………………………………………………………………………………………………………… 5 Methodology ……………………………………………………………………………………………………………… 2. 1 Research approach …………………………………………………………………………………………………………… 6 2. 2 Sampling ………………………………………………………………………………………………………………………… 6 2. 3 Data Collection ……………………………………………………………………………………………………………….. 6 2. 4 Data analyzing instruments ………………………………………………………………………………………………. 2. 5 Applied regression model …………………………………………………………………………………………………. 7 2. 5. 1 Dependent variable…………………………………………………………………………………………………… 7 2. 5. 2 Independent variables……………………………………………………………………………………………….. 7 2. 5. 3 Regression analysis explained ……………………………………………………………………………………. 8 2. 6 2. 3. Reliability and validity …………………………………………………………………………………………………… 10 Limitations of the study ………………………………………………………………………………………………….. 10 Framework ……………………………………………………………………………………………………………… 12 3. 1 Previous Studies ………………………………………………………………….. ……………………………………….. 2 3. 1. 1 ROE – profitability indicator ……………………………………………………………………………………. 12 3. 1. 2 Credit risk management indicators …………………………………………………………………………… 13 3. 2 Theories ……………………………………………………………………………………………………………………….. 17 3. 2. 1 Risks in banks ………………………………………………………………………………………………………… 17 3. . 2 Credit risk management in banks ……………………………………………………………………………… 17 3. 2. 3 Bank Profitability …………………………………………………………………………………………………… 19 v 3. 3 Regulations …………………………………………………………………………………………………………………… 20 3. 3. 1 3. 3. 2 4. Swedish regulation of banks …………………………………………………………………………………….. 0 The Basel Accords ………………………………………………………………………………………………….. 20 Empirical Findings and Analysis ……………………………………………………………………………… 23 4. 1 Overview of the banks studied ………………………………………………………………………………………… 23 4. 1. 1 Nordea ………………………………………………………………………………………………………………….. 24 4. . 2 SEB ………………………………………………………………………………………………………………………. 24 4. 1. 3 Svenska Handelsbanken…………………………………………………………………………………………… 25 4. 1. 4 Swedbank ………………………………………………………………………………………………………………. 26 4. 2 4. 3 The relationship between credit risk management and profitability in Nordea ……………………….. 28 4. 4
The relationship between credit risk management and profitability in SEB …………………………… 31 4. 5 The relationship between credit risk management and profitability in Svenska Handelsbanken .. 33 4. 6 The relationship between credit risk management and profitability in Swedbank …………………… 36 4. 7 Basel II application affect ……………………………………………………………………………………………….. 39 4. 8 5. The relationship between credit risk management and profitability in four banks…………………… 7 Summary of the findings ………………………………………………………………………………………………… 42 Concluding remarks ………………………………………………………………………………………………… 43 References ……………………………………………………………………………………………………………………… 44 APPENDIX 1 …………………………………………………………………………………………………………………. 8 APPENDIX 2 …………………………………………………………………………………………………………………. 49 APPENDIX 3 …………………………………………………………………………………………………………………. 59 APPENDIX 4 …………………………………………………………………………………………………………………. 61 vi List of Tables Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 Table 10 Table 11 Table 12 Table 13 Table 14 Table 15
Table 16 Table 17 Table 18 Table 19 Table 20 Table 21 Table 22 Table 23 Table 24 Overview of studied banks Coefficients summary table, 4 banks Model summary table, 4 banks Coefficient summary table, Nordea Model summary table, Nordea Coefficients summary table, SEB Model summary table, SEB Coefficient summary table, Handelsbanken Model summary table, Handelsbanken Coefficient summary table, Swedbank Model summary table, Swedbank Coefficient summary table, Basel II effect Model summary table, Basel II effect Summary of regression results Required ratios for SPSS analysis
Regression results with NPLR and CAR as independent variables in 4 banks Simple regression results with NPLR as independent variable in 4 banks Simple regression results with CAR as independent variable in 4 banks Regression results with NPLR and CAR as in dependent variables in Nordea Regression results with NPLR and CAR as independent variables in SEB Regression results with NPLR and CAR as independent variables in Swedbank Regression results with NPLR and CAR as independent variables in Svenska Handelsbanken Regression results with NPLR and CAR as independent variables in 4 banks before Basel II Regression results with NPLR and CAR as independent variables in 4 banks after Basel II 23 27 27 30 30 33 33 35 35 38 38 40 40 42 49 50 51 52 53 54 55 56 57 58 vii List of Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 Figure 21 Figure 22 Figure 23
Figure 24 Net Interest Income in 4 banks, 2000-2008 Share of deposit among banks in Sweden NPLR of Nordea 2000-2008, in % ROE of Nordea 2000-2008, in % CAR of Nordea 2000-2008, in % NPLR of SEB 2000-2008, in % ROE of SEB 2000-2008, in % CAR of SEB 2000-2008, in % NPLR of Handelsbanken 2000-2008, in % ROE of Handelsbanken 2000-2008, in % CAR of Handelsbanken 2000-2008, in % NPLR of Swedbank 2000-2008, in % ROE of Swedbank 2000-2008, in % CAR of Swedbank 2000-2008, in % Regression equation applied TLs amount in 4 banks 2000-2008, in mln SEK NPLs amount in 4 banks 2000-2008, in mln SEK ROE of 4 banks 2000-2008, in % NPLR of 4 banks 2000-2008, in %
CAR of 4 banks 2000-2008, in % Excerpt from Handelsbanken Annual Report 2007: Total Loan amount calculation Excerpt from Handelsbanken Annual Report 2005 (collection of ROE and CAR) Excerpts from Annual Reports: Handelsbanken 2007 and Swedbank 2002 SEB Risk composition per division 2008 24 24 28 28 29 31 31 32 34 34 34 37 37 38 49 59 59 60 60 60 61 61 61 62 viii 1. Introduction In this chapter, we present the background of the thesis followed by the problem statement. The discussion also contains the motivation for our thesis. Finally, we present the research question, the purpose of this thesis and limit the area of the study. 1. 1 Background
The world has experienced remarkable numbers of banking and financial crises during the last thirty years. Caprio and Klingebiel (1997) have identified 112 systemic banking crises1 in 93 countries since the late 1970s (Ibid. ). Demirguc-Kunt and Detragiache (1998) have identified 30 major banking crises that are encountered from early 1980s and onwards. Though most of those were experienced in the developing countries, the authors have noted that three Nordic countries Norway, Finland and Sweden – have also gone through similar crises in the late 1980s and early 1990s2. Interestingly, the majority of the crises coincided with the deregulatory measures that led to excessively rapid credit extension.
In the long run, continuous increases in asset prices created bubble3. At some point, the bubble burst and the asset markets experienced a dramatic fall in asset prices coupled with disruption. Finally, widespread bankruptcies accompanied by non-performing loans, credit losses and acute banking crises were observed. Very recently, the US subprime mortgage sector has observed one of the worst financial crises in 2007-2008. Subsequently, the global financial market is going through a turbulent situation. This has necessitated a close examination of the numerous issues related to the operation of financial markets to identify the root of the problem.
Various issues such as the capital adequacy levels in the banking system, the role of rating agencies in financial regulation and the fair-value assessment of banking assets are the most debated ones. In response to the banking crises, significant reformations have been carried out in the banking regulatory system. The most important ones are Basel Accord(s), Basel I and II, which refer to the banking supervision accords issued by Basel Committee on Banking Supervision (BCBS). Basel I, also known as 1988 Basel Accord, implemented a framework for a minimum capital standard of 8% for banks. This was enforced by law in the G 104 countries in 1992.
Basel I with focus on credit risk considers the minimum capital requirement as the main tool to prevent banks from taking excessive risk. The main reason was the belief that a well-designed structure of incentives is more effective than structural controls. Basel I contributed to the financial stability by creating conditions for equal competitions amongst banks across borders. However, several issues such as lack of risk sensitive measures of the creditworthiness and weak incentives for banks to strengthen risk management system emerged as shortcomings. These stimulated significant opportunities for regulatory arbitrage such as the increase of off balance-sheet exposure. It was revealed that Basel I was unable to provide an adequate response to the changing global context. 1
Systemic risk is the risk of collapse of an entire system or entire market and not to any one individual entity or component of that system. Steigum (1992) and Vihriala (1997) discusses on the Norwegian and Finnish cases Englund P. (1999), The Swedish Banking crisis: Roots and Consequences, Oxford review of Economic Policy, Vol. 15, no. 3, pages 80-97. 4 The Group of Ten consists of eleven industrial countries (Belgium, Canada, France, Germany, Italy, Japan, the Netherlands, Sweden, Switzerland, the United Kingdom and the United States) which advise and co-operate on economic, monetary and financial matters. 2 3 Consequently, Basel II came into effect to better reflect banks’ underlying risk and response to financial innovation like securitization.
It was argued that Basel II improved risk management practices that were not evident in Basel I5. Ironically, the frequency of crises did not decrease despite the introduction of successive reforms. Why? There are many contributing factors, mainly, political and economical conditions. It can thus be self evident that the improved risk management does not improve the banking business. Moreover, Jean-Charles Rochet (2008) states that key factors to successful reform are independence and accountability of banking supervisors. As long as banking supervisors represent political and economical interests of their respective countries, it is not possible to implement international regulation successfully.
The current global financial crisis indicates that risk management of the financial institutions is not adequate enough. This leads to the failure of the banks in highly challenging financial market. Furthermore, the discussion of financial crisis in mass media and among scholars mentions the risk management as omissions or neglect of risk measurement signals. They state that more attentive participants could avoid the tremendous affect of the financial meltdown6. Therefore, Risk Management as a discipline is being taking seriously nowadays7. Nevertheless, the financial storm teaches several key lessons which can assist to improve the risk management in future.
As a result, risk has become a very challenging area of studies. This has motivated us to conduct our thesis on this area of interest. 5 Wahlstrom G (2009), Risk management versus operational action: Basel II in a Swedish Context, Management Accounting Research, Vol. 20, page 53-68. 6 Joe Nocera, “Risk management and financial crisis”, Herald Tribune, January 4, 2009 7 A. E. Feldman Associates, Inc. , US Consulting firm report “On financial crisis and its effect” http://www. aefeldman. com Accessed February 9, 2009 2 1. 2 Problem Discussion “Keeping regulations too long is a danger, getting rid of them too quickly is another” Lars Nyberg, Deputy Governor of the Sveriges Riksbank
The banks of Sweden as well as the other countries of the EU are required to follow Basel II capital adequacy framework from 2007. The Basel II therefore replaced Basel I for international banks. Basel II aims to build on a solid foundation of prudent capital regulation, supervision, and market discipline, and to enhance further risk management and financial stability8. The banking sector has become more complex over the last decades due to the development of financial security market. As a result, banks are getting involved in compound transactions without fully realizing the risk level. Consequently, the risk bearing side gets blurred and risk exposure splits on everybody. This causes systemic failure – the economical system of the countries breaks down.
Government influences the situation and tries to stabilize economy through the regulatory mechanisms. However, it is worth mentioning that regulatory and deregulatory transitions usually end up with the same result. For example, financial instabilities and crises all over the world, the US banking crises, Swedish banking crisis and so on. The exposed risk – the main and most difficult one to identify – is the credit risk in the particular current case. The importance of this risk is increased by the fact that it is linked to the problem of collateral. Therefore, it is in need of being deliberately examined and studied. For this reason, Basel II considers varieties of credit risk measurement techniques, wider than Basel I did.
The goal is to improve the credit risk management quality without constraining banks’ competitiveness. Regulations should be interactive or flexible to be successful because of rapidly changing technological, political, and economical circumstances. Credit risk measurement tools presented in Basel II intended to be flexible. The banks can either choose from the proposed options or employ their own as long as it gives sound and fair results. The importance of the credit risk management and its impact on profitability has motivated us to pursue this study. We assume that if the credit risk management is sound, the profit level will be satisfactory.
The other way around, if the credit risk management is poor, the profit level will be relatively lower. Because the less the banks loss from credits, the more the banks gain. Moreover, according to Johan Sjomark from Credit Risk Management Department in Swedbank, profitability is the indicator of credit risk management. The central question is how significant is the impact of credit risk management on profitability. This thesis is an endeavor to find the answer. 8 Basel Committee on Banking Supervision. “Implementation of Basel II: Practical Considerations”. July 2004 3 Initially, we intend to describe not only the impact of credit risk management on profitability but also to determine how the banks in Sweden monitor and assess credit risk.
Since we believe that credit risk management is a very complex issue, it requires a deliberate qualitative study supplemented with quantitative study to achieve the goal. However, it was not possible because of failure in primary data collection. Unfortunately, the respondent banks refused to participate in the survey and to disclose the required information. Therefore, we are left with the data available from the annual report to carry out our study based on the quantitative approach. We choose four wellknown commercial banks in Sweden: Nordea, Handelsbanken, Swedbank and SEB. 1. 3 Research question The discussed background and problem formulation make us to have the following research question: – How does credit risk management affect the profitability in commercial banks in Sweden? 1. 4 Purpose
The purpose of the research is to describe the impact level of credit risk management on profitability in four commercial banks in Sweden. 1. 5 Delimitation The research is limited on identifying the relationship of credit risk management and profitability of commercial banks in Sweden. Thus, the other risks mentioned in Basel Accords are not discussed. Due to the unavailability of information in annual reports, our sample only contains four largest commercial banks and their 9 years’ annual reports from 2000 to 2008 respectively. Since the banks in sample rejected to participate in our internet based survey, the primary data was not possible to obtain. However, we were able to arrange one telephone interview with one of the risk control officers in Swedbank.
Considering the above mentioned circumstances, the results of the study are limited to four commercial banks in the sample and are not generalized for the commercial banks in Sweden. Finally, as our study only uses the quantitative approach and focus on the description of the outputs from SPSS, we will not go deep to discuss the reasons and give our own explanation. 4 1. 6 Disposition Chapter1 •INTRODUCTION: In this chapter, we present the background of the thesis followed by the problem statement. The discussion also contains the motivation for our thesis. Finally, we present the research question, the purpose of this thesis and limit the area of the study. Chapter 2 •METHODOLOGY: In this chapter, we widely describe the HOW part of our study. The chapter comprises In esearch approach, sampling, data collection, data anyzing instruments and the description of applied regression model. The chapter is finalized by reliability and validity and limitation of our study. Chapter 3 Chapter 4 •FRAMEWORK: In this chapter, we provide theoretical foundation to our study by presenting relevant In literature. •EMPIRICAL FINDINGS AND ANALYSIS: In this chapter, we present the results of our regression model. We EMPIRICAL analyze the results and describe the impact of credit risk management on profitability. •CONCLUDING REMARKS: In this chapter, we conclude on our study and give suggestions for further studies. CONCLUDING Chapter 5 5 2. Methodology In this chapter, we widely describe the HOW part of our study.
The chapter comprises research approach, sampling, data collection, data anyzing instruments and the description of applied regression model. The chapter is finalized by reliability and validity and limitation of our study. 2. 1 Research approach Our study is conducted by using deductive approach as we refer to the research question and do not intend to go beyond it. Also, we base our research question on previously existed theories and studies in this area. The method of our study is quantitative. We use regression model to analyze data collected from the annual reports of the sample banks. Based on the regression outputs we conduct the analyses and answer our research question. The analyses are presented by using descriptive approach.
Since we only describe the regression results without providing further explanation on the issues. 2. 2 Sampling We have selected four major commercial banks in Sweden: Nordea, SEB, Svenska Handelsbanken and Swedbank. We have used annual reports from 2000 to 2008 of each bank to collect the data. Therefore, there are total 36 observations in the regression analysis. Theoretically, the number of observations should be 20:1 (20 observations per one independent variable) in the regression analysis and as low as 5:19. In our case, we have 36 observations and two independent variables which are satisfactory with respect to standard. 2. 3 Data Collection The data source for our study is Annual Reports for 9 years, 2000-2008.
Our study necessitates looking into credit risk management disclosure, financial statements and notes to financial statements within the annual reports of the sample banks. We interviewed Johan Sjomark, credit risk control department officer in Swedbank with 8 years of working experience in credit risk management. The interview helped us to enhance our regression model and has been limited to that. The interview answers have not been used to produce analysis or the conclusion of this paper. The interview questions are presented in the Appendix 1. 2. 4 Data analyzing instruments We use multiple regression analysis in our study: the relation of one dependant variable to multiple independent variables. The regression outputs are obtained by using SPSS.
In addition, we apply MS Excel 2007 to confirm the accuracy of the results. 9 Princeton University. Data and Statistical Services online. http://dss. princeton. edu/online_help/analysis/interpreting_regression. htm. Accessed 2009-05-06. 6 2. 5 Applied regression model We have revealed from early studies that the determinant for profitability is ROE (Net Income/Total Shareholders’ Equity) and for credit risk management are NPLR (Non-performing Loans/Total Loans) and CAR [(Tier I + Tier II)/Risk Weighted Assets] respectively. We use multiple regression model with two independent variables in this study. In the regression model, we have considered the following: 2. 5. Dependent variable We have decided to use ROE as the indicator of the profitability in the regression analysis because ROE along with ROA has been widely used in earlier research. Initially, we have considered the ratios ROE and RORAC10 (Profit after Tax/Risk Adjusted Capital). RORAC is a measure for relative performance of the banks and could have been used in our regression analysis. However, we have not used RORAC because it is usually used by the banks with internally available information, for example, risk-adjusted capital, and we do not have access to such required information. Therefore, we have decided to use ROE as the indicator of profitability.
In this case, the required information is available in the annual reports of the banks under Key figures section. 2. 5. 2 Independent variables We have chosen two independent variables namely NPLR and CAR because these two are the indicators of risk management which affect the profitability of banks. NPLR, in particular, indicates how banks manage their credit risk because it defines the proportion of NPL amount in relation to TL amount. NPLR. NPLR is defined as NPLs divided by TLs. To calculate this ratio, we have used data provided in the annual reports of each bank. From 2000 until 2005, NPL amount has been presented using different names, such as, impaired loans, problem loans, doubtful claims and loan allowances. 1 However, the definitions of those are similar to the definition of NPLs. Banks provide more precise categorization of NPLs after the adoption of IFRS in 2005. NPL amount is provided in the Notes to financial statements under Loans section. TL amount, the denominator of the ratio, has been gathered by adding two types of loans: loans to institutions and loans to the public. We have collected the loan amount provided in the balance sheet of the banks in their annual reports12. Thus, calculation of the NPLR has been accomplished in following way: NPLR = (NPL amount) ? (TL amount) 790 ? 1478137 = 0,0513 CAR. CAR is regulatory capital requirement (Tier 1 + Tier 2) as the percentage of RWAs.
We have chosen CAR as an independent variable and justified the choice in the literature review 10 Return On Risk Adjusted Capital Excerpt from annual reports with definitions under Figure 23 12 Excerpt from the annual report is presented under Figure 21 in appendix. 13 Numbers are obtained from Handelsbanken Annual Report 2007 11 7 section of this study. CAR is taken from the banks’ annual reports under Key figures14 section. We have not done any calculation to obtain CAR. We find it useful for our study to discuss if the adoption of Basel II has influenced the credit risk management and its effect on profitability in four banks. 2. 5. 3 Regression analysis explained
The regression analysis is conducted to find out the following: a. The relationship between credit risk management and profitability in four banks: we use 9 year period (2000-2008) for 4 banks which in total gives 36 observations; b. The relationship between credit risk management and profitability in each bank: we use 9 year period (2000-2008) for each bank which in total gives 9 observations; c. The Basel II application affect in four banks: we use 2 years before (2005-2006) and 2 years after (2007-2008) the adoption of new regulation for all four banks which in total gives 8 observations. We have employed the multivariate regression model which is presented below: Y= ? +? 1X1+ ? 2X2+…+ ? nXn+ ? Standard
Y – the value of dependent variable; ? – the constant term; ? – the coefficient of the function; X – the value of independent variables: Our application Y: ROE- profitability indicator X1: NPLR –credit risk management indicator X2: CAR –credit risk management indicator ? – the disturbance or error term. Thus the regression equation becomes: ROE= ? +? 1NPLR+ ? 2CAR+ ? It is the regression function which determines the relation of X (NPLR and CAR) to Y (ROE). ? is the constant term and ? is the coefficient of the function15, it is the value for the regression equation to predict the variances in dependent variable from the independent variables. This means that if ? oefficient is negative, the predictor or independent variable affects dependent variable negatively: one unit increase in independent variable will decrease the dependent variable by the coefficient amount. In the same way, if the ? coefficient is positive, the dependent variable increases by the coefficient amount. ? is the constant value which dependent variable predicted to have when independent variables equal to zero (if X1, X2=0 then ? =Y). Finally, ? is the 14 15 Excerpt from the Annual Report is presented under Figure 22 in Appendix ? represents the independent contributions of each independent variable to the prediction of the dependent variable. 8 isturbance or error term, which expresses the effect of all other variables16 except for the independent variables on the dependent variable that we use in the function. Regression analysis output contains values which we discuss below: R2 is the proportion of variance in the dependent variable that can be predicted from independent variables. There is also adjusted R2 which gives more accurate value by avoiding overestimation effect of adding more variables to the function. So, high R2 value indicates that prediction power of dependent variable by independent variables is also high. Adjusted R2 is calculated using the formula 1-((1-R2)*((N-1)/(N-k17-1))18. The formula shows that if the number of observations is small the difference between R2 and adjusted R2 is greater han 1 since the denominator is much smaller than numerator. Adjusted R2 sometimes gives negative value. Since R2 is adjusted to find out how much fit probably happen just by luck: the difference is amount of fit by chance. Also, negative values of adjusted R2 occur if the model contains conditions that do not help to predict the response (ROE) or the predictors (NPLR and CAR) chosen are wrong to predict ROE. R2 is generally considered to be secondary importance, unless the primary concern is of using regression equation to make accurate predictions. R2 is an overall measurement of the strength of association, and does not reflect how any independent variable is associated with the dependent variable.
The Probability value (P-value) is used to measure how reliably the independent variables can predict the dependent variable. It is compared to the significance level which is typically 0,05. If the P-value is greater than 0,05, it can be said that the independent variable does not show a statistically significant relationship with the dependent variable. The F-value calculated as (R2/1)/((1-R2/n19-2)) and associated P-value shall be looked at to measure the effect of the group of independent variables on dependent variable. The resulted Fvalue should be compared to the critical F-value (Fv1, v2) which is taken from the F distribution table. Both V1 and V2 are called as degrees of freedom.
V1 is number of independent variables and V2 is number of observations minus number of independent variable minus 1. For instance, in our case, we have two independent variables and 36 observations, then V1=2, and V2=n-k-1=36-21=33. Thus the critical value of F (3,32)20 can be found in the distribution table accordingly. If the resulted F-value exceeds the critical F-value, it can be said that the regression as a whole is significant. 16 Variables that are not included in the function but could have effect on dependent variable k – number of independent variables 18 UCLA Academic Technology Services. Annotated Stata Output. Regression Analysis. http://www. ats. ucla. edu/stat/stata/output/reg_output. htm. Accessed 009-05-06. 19 n=number of observations 20 Anerson. D. R & Sweeney. D. J & Williams. T. A (2008) Statistics for business and economics, 10th edition, F-distribution Table. pp. 928 17 9 2. 6 Reliability and validity Reliability and validity are often used by the scientific researchers in their studies, both qualitative and quantitative. Reliability refers to the consistency and accuracy of the research results. In the quantitative research, reliability can be illustrated as the stability of the measurement over time, the similarity of the measurements during the given period, and also the degree to the same results of the measurement given repeatedly.
Validity means the accuracy of the measurement of which it is intended to be measured and how truthful the results of the research are. 21 In our study, we have collected the data from the peer reviewed scientific articles, journals, books, the audited annual reports by the authorized accounting firms. In addition, we have used the capital adequacy and risk management reports of banks to collect our data. Furthermore, ROE and CAR are taken from the annual reports directly in order to avoid the mistakes of calculation. However, NPLR is not available for all the banks in the annual reports. So we have taken the amount of NPLs and the TLs from the financial statements and the related notes, and then, used the formula of NPLR (NPLs/TLs) to obtain the value.
Moreover, interview with the credit risk officer helped us to ensure application of NPLR as an independent variable. To ensure the accuracy of the results, we have triple checked the data collection and calculation processes. Next, we have used the statistical analysis tool SPSS to obtain results and conduct analysis of the regression model that we have adopted in our study. The reliability of the SPSS results has been proved by many researchers in their studies. We have also used several articles to get the idea how to analyze the SPSS outputs. It is worth to mention that we have compared the regression analysis results of SPSS with the results of MS Excel to ensure correctness.
Also, we have checked correlation between independent variable to prevent multicollinearity effect in our result. Thus, the theories, the findings and the results obtained through regression analysis of our study are replicable which consequently guarantee the reliability and validity of our study. 2. 7 Limitations of the study The study includes only four commercial banks in Sweden. Therefore, the results cannot be generalized to all commercial banks. The number of observations is very small which is not preferable while conducting studies of this kind. Moreover, only nine observations of each bank are used in the regression analysis which might not provide accurate results.
Nordea is not comparable to other banks in sample since it is twice as large as each of the other bank in the sample. Therefore, the analysis and the comparison of the banks are subjective to some extent. Moreover, unlike other three banks in the sample, the Swedish government is the major shareholder in Nordea. This might have introduced difficulties in comparability within the sample. In addition, this also can influence the credit risk management approach, described as more prudent than that in other banks. 21 Nahid Golafshani (2003) Understanding Reliability and Validity in Qualitative Research, The Qualitative Report Volume 8 Number 4 PP. 97607 http://www. nova. edu/ssss/QR/QR8-4/golafshani. pdf accessed 2009-05-14 10 NPLs are subjective amount (years 2000-2004) since banks’ annual reports have different names for such types of loans and definitions are not always provided. Therefore, some NPL amounts before year 2005 might not be accurate. Also, not many previous studies have treated NPLR as one of the variables in the regression models. The number of independent variables could be more than two since there are other factors affecting ROE with the same prediction level. Additionally, the independent variables are chosen by our initiative and therefore the regression model might be subjective.
Finally, due to instability of current financial market it is not the ideal situation to conduct this type of analysis since ROE and NPLR are affected more than that of normal conditions. 11 3. Framework In this chapter, we provide theoretical foundation to our study by presenting relevant literature. 3. 1 Previous Studies 3. 1. 1 ROE – profitability indicator ROE as an important indicator to measure the profitability of the banks has been discussed extensively in the prior studies. Foong Kee K. (2008) indicated that the efficiency of banks can be measured by using the ROE which illustrates to what extent banks use reinvested income to generate future profits. 2 According to Riksbank’s Financial Report (2002)23, the measurement of connecting profit to shareholder’s equity is normally used to define the profitability in the banks. Furthermore, the paper “Why Return on Equity is a Useful Criterion for Equity Selection”24 has mentioned that ROE provides a very useful gauge of profit generating efficiency. Because it measures how much earnings a company can get on the equity capital. The ROE is defined as the company’s annual net income after tax divided by shareholder’s equity. NI is the amount of earnings after paying all expenses and taxes. Equity represents the capital invested in the company plus the retained earnings25. Essentially, ROE indicates the amount of earnings generated from equity.
The increased ROE may hint that the profit is growing without pouring new capital into the company. A steadily rising ROE also refers that the shareholders are given more each year for their investment. All in all, the higher ROE is better both for the company and the shareholders. 26 In addition, ROE takes the retained earnings from the previous periods into account and tells the investors how efficiently the capital is reinvested. In accordance with the study Waymond A G. (2007), profitability ratios are often used in a high esteem as the indicators of credit analysis in banks, since profitability is associated with the results of management performance.
ROE and ROA are the most commonly used ratios, and the quality level of ROE is between 15% and 30%, for ROA is at least 1%27. The study of Joetta C (2007) presented the purpose of ROE as the measurement of the amount of profit generated by the equity in the firm. It is also mentioned that the ROE is an indicator of the efficiency to generate profit from equity. This capability is connected to how well the assets are utilized to produce the profits as well. The effectiveness of assets utilization is significantly tied to 22 Foong. K. K (2008) Return-on-equity ratio can show how efficient banks are. Malaysian Institute of Economic Research. http://biz. thestar. com. y/news/story. asp? file=/2008/11/24/business/2590590&sec=business accessed 2009-04-02 23 Riksbank Financial Stability Report (2002), ”The major Swedish banks in an international comparison”, Issue 1 24 Jensen Investment Management (2008) why return on equity is a useful criterion for equity selection www. jenseninvestment. com/documents/WPIssue7_ROE_Aug08. pdf accessed 2009-04-03 25 Retained earnings is calculated as net income minus dividend paid out. 26 Jensen Investment Management (2008) why return on equity is a useful criterion for equity selection www. jenseninvestment. com/documents/WPIssue7_ROE_Aug08. pdf accessed 2009-04-03 27 Waymond. A .
G (2007) Credit analysis of financial institutions 2nd Edition, Published by Euromoney Books pp. 197 12 the amount of assets that the company generates for each dollar of equity. 28 Thus, after we brought the evidence of ROE being used as the profitability indicator, we can move to the discussion of credit risk management indicators. 3. 1. 2 Credit risk management indicators In response to recent corporate and financial disasters, regulators have increased their examination and enforcement standards. In banking sector, Basel II has established a direct linkage between minimum regulatory capital and underlying credit risk, market risk and corporate risk exposure of banks.
This step gives an indication that Capital management is an important stage in risk mitigation and management. However, development of effective key risk indicators and their management pose significant challenge. Some readily available sources such as policies and regulations can provide useful direction in deriving key risk indicators and compliance with the regulatory requirement can be expressed as risk management indicators. A more comprehensive capital management framework enables a bank to improve profitability by making better riskbased product pricing and resource allocation. The purpose of Basel II is to create an international standard about how much capital banks need to put aside to guard against the types of risk banks face.
In practice, Basel II tries to achieve this by setting up meticulous risk and capital management requirements aimed at ensuring that a bank holds capital reserves appropriate to the risks the bank exposes itself to. These rules imply that the greater risk which bank is exposed to, the greater the amount of capital a bank needs to hold to safeguard its solvency. The theoretical banking literature is, however, divided on the effects of capital requirements on bank behavior and consequently, on the risks faced by the institutions. Some academic works point toward that capital requirement clearly contributes to various possible measures of bank stability. On the contrary, other works conclude that capital requirements make banks riskier institutions than they would be in the absence of such requirements.
Jeitshko and Jeung (2005) have discovered numerous aspects that explain the differing implications of portfolio-management models for the responsiveness of bank portfolio risk to capital regulation. Results depend on banks being either value-maximizing or utility-maximizing firms; bank ownership (if limited liability) and whether banks operate in complete or incomplete asset markets. Moreover, the effects of capital regulation on portfolio decisions and therefore on the banking system’s safety and soundness eventually depend on which perspective dominates among insurers, shareholders, and managers in the principal-agent interactions. Capital and profitability
Theory provides contradictory forecast on whether capital requirements limit or enhance bank performance and stability. The soundness of the banking system is important because it limits 28 Joetta. C(2007) Credit risk management: how to avoid lending disasters and maximize earnings 3rd edition pp. 144 13 economic downturn related to the financial anxiety. Also, it avoids unfavorable budgetary consequences for governments which often bear a substantial part of bailouts cost. Prudential regulation is expected to protect the banking system from these problems by persuading banks to invest prudently. The introduction of capital adequacy regulations strengthen bank and therefore, enhance the resilience of banks to negative shocks.
However, these rules may cause a shift of providing loans from private sector to public sector. Banks can comply with capital requirement ratios either by decreasing their risk-weighted assets or by increasing their capital. Goddard, Molyeux and Wilson (2004) analyzed the determinants of profitability of European banks. The authors found a considerable endurance of abnormal profits from year to year and a positive relationship between the capital-to-asset ratio and profitability. Demirguc-Kunt and Huizinga (1999) examined how capital requirement alter the incentives that banks face. An increase in capital requirement necessitates banks to substitute equity for deposit financing, reduce shareholder’s surplus.
The decline in surpluses intensifies the probability of loss, driving a rise in the cost of intermediation to sustain profitability. In support of this hypothesis, authors have provided empirical evidence showing a significant effect on interest margins pursuant to higher capital holdings and the share of total assets held by banks. The evidence also supports higher net interest margins and more profitability for well-capitalized banks29. This is in harmony with the fact that banks with high capital ratio have low interest expenses due to less probable bankruptcy costs. Samy and Magda (2009) focus on the impact of capital regulation on the performance of the banking industry in Egypt.
The study provides a comprehensives framework to explicitly measure the effects of capital adequacy on two specific indicators of bank performance: cost of intermediation and profitability. The results provide a clear illustration of the effects of capital regulations on the cost of intermediation and banks’ profits. As CAR internalizes the risk for shareholders, banks increase the cost of intermediation, which supports higher return on assets and equity. These effects appear to increase progressively over time, starting in the period in which capital regulations are introduced and continuing 2 years after the implementation. Nonetheless, the evidence does not support the hypothesis of a sustained effect of capital regulations over time, or variation in the effects with the size of capital across banks.
The authors have concluded that a number of factors contributed positively to banks’ profitability in the post-regulation period: higher capital requirements, the reduction in implicit cost, and the increase in management efficiency. Countering effects on banks’ profitability were attributed to the reduction in economic activity and, to a lesser extent, to the reduction in reserves. An improvement of cost efficiency is not reflected in a reduction in the cost intermediation or an 29 Naceur Samy and Kandil Magda, The impact of capital requirements on banks’ cost of intermediation and performance: The case of Egypt, Journal of Economics & Business, (2009), Vol. 61, 70-89. 14 improvement in profit. The effect of better efficiency is likely to have been absorbed in banks’ fees and commissions.
Non-performing loans Why NPL occurs? The IMF30 paper (2001) presents two main reasons for that: poor risk management and plain bad luck in form of external independent factors. The inflation, deregulation and special market conditions can lead to poor credit lending decision which in turn leads to NPLs. In fact, many NPL studies are conducted in the countries with financial market recession31. In prior studies, NPL is usually mentioned in East Asian countries’ macroeconomic studies, while they run into serious economic downturn, as one of the financial and economical distress indicators. Japan and China, are those of most mentioned in this regard32.
Moreover, IMF working paper from December 2001 encourages better account of NPL for macroeconomic statistics which makes NPL to be widely used in macroeconomic statistics. Moreover, Hippolyte F. (2005) advocates that macroeconomic stability and economic growth are associated with declining level of NPLs, while the adverse macroeconomic situation is associated with rising scope of NPLs. Ongoing financial crises suggest that NPL amount is an indicator of increasing threat of insolvency and failure. However, the financial markets with high NPLs have to diversify their risk and create portfolios with NPLs along with Performing Loans, which are widely traded in the financial markets. In this regard, Germany was one of the leaders of NPL markets in 2006 because of its sheer size and highly competitive market.
Also, Czech Republic, Turkey and Portugal are noticeable NPL markets in EU according to Ernst &Young’s Global Non-performing Loan report (2006). Nonetheless, not many studies have done research on NPL market in Western Europe or Scandinavia. Empirical study of Petersson J et al. (2004), states that during the crises in the early 1990’s in Sweden, the Swedish government created the workout units in order to improve the situation with loan losses in banks and succeeded. The same paper claims that no NPL market exists in Sweden since four major banks33 showed loan losses below 0. 25% for the year 2003. We wonder whether the situation has changed nowadays. Brewer et al. (2006) use NPLR as a strong economic indicator.
Efficient credit risk management supports the fact that lower NPLR is associated with lower risk and lower deposit rate. However it also implies that in long run, relatively high deposit rate increases the deposit base in order to fund relatively high risk loans and consequently increases possibility of NPLR. Therefore, the 30 International Monetary Fund World Bank Policy Research Working Paper 3769, November 2005 32 Ernst&Young. Global Nonperforming Loan Report 2006. 33 Paper presents banks: Foreningssparbanken, Handelsbanken, Nordea and SEB. 31 15 allocation of the available fund and its risk management heavily depend on how the credit risk is handled and diversified to decrease the NPL amount. NPL is a probability of loss that requires provision.
Provision amount is “accounting amount” which can be further, if the necessity rises, deducted from the profit. Therefore, high NPL amount increases the provision amount which in turn reduces the profit. The above stated discussion proves that NPLR and CAR are reasonably considered as credit risk management indicators. Thereby, they can be used in our study. 16 3. 2 Theories 3. 2. 1 Risks in banks Risks are the uncertainties that can make the banks to loose and be bankrupt. According to the Basel Accords, risks the banks facing contain credit risk, market risk and operational risk. Credit risk is the risk of loss due to an obligator’s non-payment of an obligation in terms of a loan or other lines of credit. 4 The Basel committee proposes two methodologies for calculating the capital requirements for credit risk, one is to measure the credit risk in a standardized manner and the other is subject to the explicit approval of the bank’s supervisor and allows banks to use the IRB approach. Market risk is defined as the risk of losses in on and off-balance sheet positions arising from movements in market prices. 35 The capital treatment for market risk addresses the interest rate risk and equity risk pertaining to financial instruments, and the foreign exchange risk in the trading and banking books. 36 The value at risk (VaR) approach is the most preferred to be used when the market risk is measured.
Operational risk is defined as the risk of direct or indirect loss resulting from inadequate or failed internal processes, people and systems or from external events. There are three approaches applied to the operational risk measurement: Basic Indicator Approach (BIA), Standardized Approach (SA), and Advanced Measurement Approach (AMA). 3. 2. 2 Credit risk management in banks Bank loan is a debt, which entails the redistribution of the financial assets between the lender and the borrower. The bank loan is commonly referred to the borrower who got an amount of money from the lender, and need to pay back, known as the principal. In addition, the banks normally charge a fee from the borrower, which is the interest on the debt. The risk associated with loans is credit risk.
Credit risk is perhaps the most significant of all risks in terms of size of potential losses. Credit risk can be divided into three risks: default risk, exposure risk and recovery risk. As the extension of credit has always been at the core of banking operation, the focus of banks’ risk management has been credit risk management. It applied both to the bank loan and investment portfolio. Credit risk management incorporates decision making process; before the credit decision is made, follow up of credit commitments including all monitoring and reporting process37. The credit decision is based on the financial data and judgmental assessment of the market outlook, borrower, management and shareholders.
The follow-up is carried out through periodic reporting reviews of the bank commitments by customer. Additionally, “warning systems” signal the deterioration of the condition of the borrower before default, whenever possible. 34 Basel II (2006) International Convergence of Capital Measurement and Capital Standards, A Revised Framework Comprehensive Version. Basel II (2006) International Convergence of Capital Measurement and Capital Standards, A Revised Framework Comprehensive Version. 36 Basel I http://www. bnm. gov. my/guidelines/01_banking/01_capital_adequacy/02_basel1. pdf accessed 2009-03-14 37 Joel Bessis (1998) Risk Management in Banking 35 17 Loans that are in default or close to being default become NPLs38.
The terms of the default rate in loans are defined by each bank. Usually, loan becomes non-performing after being default for three months but this can depend on contract terms. NPLR shows the proportion of the default or near to default loans to the actual performing loans. It indicates the efficiency of the credit risk management employed in the bank. Therefore, the less the ratio the more effective the credit risk management. Measurement of credit risk Usually, bank can project the average level of credit losses it can reasonably expect to experience. These losses are referred to: a. Expected Losses (EL): perceived as cost of business undertaking by financial institutions; b.
Unexpected Losses (UL): losses above expected level when banks anticipate their occurrence though the timing and severity cannot be known beforehand. A few portions of unexpected losses might be absorbed by the interest rate charged on credit exposure although market will not support adequate prices to cover all unexpected losses. c. Loss Given Default (LGD): the amount of fund that bank can lose when the borrower defaults on a loan. Therefore, capital is needed to cover the risks of such losses. Banks have an incentive to minimize capital they hold since reducing capital frees up economic resources that can be directed to profitable investment. In contrast, the less capital a bank holds, the greater is the likelihood that it will not be able to meet its own debt obligations, i. e. hat losses in a given year will not be covered by profit plus available capital, and that the bank will become insolvent39. Accordingly, banks must carefully balance the risks and rewards of holding capital. A number of approaches exist to determine how much capital a bank should hold. The IRB approach adopted by Basel II focuses on the frequency of bank insolvencies (the case of the bank failing to meet its senior obligations) arising from credit losses that supervisors are willing to accept40. Through IRB approach, the Basel Committee intended to develop a framework which is credible, prudentially sound and reflect healthy risk management practices.
Banks have made use of internal rating systems for very long time as a means of categorizing their exposure into broad, qualitatively differentiated layers of risk41. 38 Special term dictionary for investors: www. investopedia. com, accessed on 2009-03-13 An Explanatory Note on the Basel II IRB Risk Weight Functions, http://www. bis. org/bcbs/irbriskweight, accessed 2009-03-14 40 http://www. bis. org/bcbs/irbriskweight accessed 2009-03-14 41 Internal Rating-Based Approach, http://www. bis. org, Accessed on 13. 03. 09 39 18 3. 2. 3 Bank Profitability In our study, we try to examine the profitability of the banks. The profitability in our case is presented and measured using ROE42.
In other words, the amount of NI returned as a percentage of TSE. We choose it as profitability indicator because ROE comprises aspects of performance, such as profitability and financial leverage. ROE in banks The measurement of bank performance has been developed over time. At the beginning, many banks used a purely accounting-driven approach and focused on the measurement of NI, for example, the calculation of ROA43. However, this approach does not consider the risks related to the referred assets, for instance, the underling risks of the transactions, and also with the growth of off-balance sheet activities. Thus the riskiness of underlying assets becomes more and more important.
Gradually, the banks notice that equity has become the scarce resource. Thereby, banks turn to focus on the ROE to measure the net profit to the book equity in order to find out the most profitable business and to do the investment. 44 ROE is commonly used to measure the profitability of banks. The efficiency of the banks can be evaluated by applying ROE, since it shows that banks reinvest its earnings to generate future profit. The growth of ROE may also depend on the capitalization of the banks and operating profit margin. If a bank is highly capitalized through the risk-weighted capital adequacy ratio (RWCAR) or Tier 1 capital adequacy ratio (CAR), the expansion of ROE will be retarded.
However, the increase of the operating margin can smoothly enhance the ROE45. ROE also hinges on the capital management activities. If the banks use capital more efficiently, they will have a better financial leverage and consequently a higher ROE. Because a higher financial leverage multiplier indicates that banks can leverage on a smaller base of stakeholder’s fund and produce higher interest bearing assets leading to the optimization of the earnings. 46 On the contrary, a rise in ROE can also reflect increased risks because high risk might bring more profits. This means ROE does not only go up by increasing returns or profit but also grows by taking more debt which brings more risk.
Thus, positive ROE does not only represent the financial strength. Risk management becomes more and more signifi