Paper Title (use style: paper title)

2014 AAGBS International Conference on Business Management (AiCoMB 2014)

Bank Fragility and Its Determinants: Evidence From Malaysian Commercial Banks Nurul Farhana Mazlan1, Noryati Ahmad2, Norlida Jaafar3 Master in Science Candidature, AAGBS1 Universiti Teknologi MARA 40450 Shah Alam, Selangor, Malaysia [email protected]

Faculty of Business Management3 Universiti Teknologi MARA 40450 Shah Alam, Selangor, Malaysia [email protected]

Arshad Ayub Graduate Business School2 Universiti Teknologi MARA 40450 Shah Alam, Selangor, Malaysia [email protected]

BSFI is an index used to monitor the level of fragility among Malaysian domestic commercial banks [2]. The main components of BSFI are associated with three excessive risk factors (credit risk, liquidity risk and foreign-exchange risk) [3]. However, BSFI in this research uses market risk as suggested by [4] as one of the factors that affect the value of banks assets and liabilities in place of foreign-exchange risk. The fluctuation of all the indicators used in this research to compute BSFI is expected to explain the changes in the level of fragility in banking sector.

Abstract—When are banks considered fragile and what trigger them to be fragile? This paper attempts to answer those questions by measuring bank fragility and identify probable factors determining bank fragility of Malaysian commercial banking sector during the period of 1996-2011. This paper constructs the Banking Sector Fragility Index (BSFI) to measure fragility of commercial bank during the period studied. The index is then used to identify the determinants of commercial bank fragility. Results of BSFI show that the commercial banking sectors are in fragile condition from 1996 until 2000 and in a highly fragile stage between 1996 and 1998 since the BSFI is less than -0.5. In addition, findings based on the Logit regression analysis infer that the likelihood of Malaysian commercial banks to be fragile is significantly determined by total loans to total assets and interbank rate.

Banks play important role in the macro-economy with regard to monetary policy as highlighted in Keynesian Theory. Additionally, asymmetric information, bank run, adverse selection and moral hazard are the most common reasons that heighten the riskiness of bank fragility within banking sector [5]. Bank Negara Malaysia (BNM) defines financial stability as regard to the condition in which the financial system comprising financial intermediaries, markets and market infrastructures must capable of withstanding shock arise from risk associated. In other point of view, financial stability promotes confidence in the financial system. While banking fragility is closely identify as an unfavorable condition influence financial instability within banking sector (vulnerability to crisis) that may cause serious breakdown in market functioning such as disruption in financial intermediation, credit crunch or lack of financing for new investment and consumption activities as well as reduce the level of confidence among local and foreign investors in financial sector. In order to ensure financial stability, in particularly the banking sector, Malaysia used guideline of CAMELS framework or three pillar Basel accord consist of Basel I, II and III. Most of the previous study have used CAMELS to detect the fragility of the bank condition. However, [2] proposed Banking Sector Fragility Index (BSFI) to identify the sign of banking fragility as this index is able to identify different level of fragility. Besides, at the point this study is carried out, there is no study that has used BSFI to measure the bank fragility in Malaysia. Hence,

Keywords—bank fragility; CAMELS framework; BSFI, logit regression analysis

I.

Introduction

The banking sector plays an important role as financial intermediary and is a primary source of financing for the domestic economy. A bank is generally known as a financial firm which offers loan and deposit products on the market, and caters to the changing liquidity needs of its borrowers and depositors. Banks serve three important functions; as an intermediary, liquidity providers and payment servicer. These functions differentiate them from other financial institutions. On the other hand, the quest for profit remained as the principal objective of their existence similar to the other business. Hence, exposure to various types of risk will be detrimental to their profits and getting fragile. Eventually, it could increase their vulnerability to crisis and make them fragile. In this study, banking fragility is simply defined as vulnerability to crisis [1] that eventually could lead to serious breakdown in market functioning such as disruption in financial intermediation, credit crunch or lack of financing for new investment and consumption activities. It may also reduce the level of confidence among local and foreign investors in financial sector.

261


this study aims to fill up this gap by constructing the BSFI to detect the level of bank fragility within Malaysian commercial banks over the period of 1996 to 2011 In addition, this paper also integrates both macroeconomic and bank specific factors to unearth the determinants of fragility of Malaysian commercial banking sector. II.

development in harsh condition and from being fragile [18]. [19] in their studies reported that higher interbank rate will lead to fragility of banking sector

III.

Review of Literature

Conceptual Framework

In an attempt to analyze Malaysian commercial banking fragility, seven domestic commercial banks have been selected after taking into consideration the availability and consistency of data. The main features of the banks are that they must be listed and licensed by Bank Negara Malaysia (BNM) as shown in Table I.

A. Fragility Indices Past literatures have developed several fragility indices for banking system based on the indicators of fragility of individual banks [6]. The index is believed to be able to predict the future bank crisis. Whether crisis has occurred or the probability of it occurring could be determined via fragility index [7]. [8] developed Speculative Pressure Index (SPI) and Index of Currency Market Turbulence (ICMT) to detect currency crisis while [9] constructed Banking Sector Fragility Index (BSFI) and Excessive Risk Index (ERI) for banking crisis [9].

TABLE I.

MALAYSIAN DOMESTIC COMMERCIAL BANKS No Type of Bank 1 Affin Bank Bhd 2 Alliance Bank Malaysia Bhd 3 CIMB Bank Bhd 4 Hong Leong Bank Bhd 5 Malayan Banking Bhd 6 Public Bank Bhd 7 RHB Bank Bhd Source: Bank Negara Website

B. Bank Specific Factors Micro version of this framework will focus on individual bank’s balance sheet data to forecast banks fragility. The selected financial ratios are based on CAMELS framework that stand for capital adequacy, asset quality, management soundness, earnings and profitability as well as liquidity and sensitivity to market risk [2]. The application of financial ratios as the proxies for the bank specific provides information about the symptoms rather than the causes of financial difficulty as they provide leading indicators of incipient crisis [10]. As indicated revealed from previous studies, failed banks had significantly lower capital ratio than those of non-failed banks [11], asset quality shows the risk level of assets and rate of financial strength within a bank [12], ratio of deposit interest to total expenses as proxy for banks management is positively related to possibility of bank failure [13]. Banks earning proxied by Return on Assets (ROA) is associated with strong and healthy banks, which should decrease due to banks failure [14]. The liquidity level for every bank represents the capability of the banks to fulfil its respective obligation [15]. Higher liquidity level indicates that the bank is not in a fragile situation. It is found that larger banks have competitive advantages and due to fiercer competition, the larger banks will push smaller banks to take higher risks. This situation could explain why bank size will influence the probability of failure [16].

The independent variables identified for this paper consist of bank specific factors and macroeconomic factors (refer to Table II). These factors considered as potential leading indicators of banks failure [20]. TABLE II.

PROXIES FOR BANK FRAGILITY DETERMINANTS Variable/Proxy Measurement Expected Result to BSFI Bank-specific variables: Capital adequacy Capital Assets + Ratio(CAR) Asset quality Total Loans to Total Assets (TL/TA) Management quality Deposit Interest Expenses to Total Expenses (DIE/TE) Earning ability Net Income as a + Percentage of Total Assets (ROA) Liquidity Loans/ Customer Deposits (TL/TD) Sensitivity to market Size (SZ) + risk Macroeconomic variables: Interest rate Malaysian Interbank rate GDP Malaysian GDP growth -/+ rate (% Change in GDP)

C. Macroeconomic Factors as Determinants Macroeconomics is a branch of economic dealing with the performance, structure, behaviour and decision making of the whole economy. It is actually one of the two most general fields in economics which involves the sum total of economic activity dealing with the issues of growth, inflation, and exchange rate. The fluctuation of economic trend could affect the banking sector performance. Higher GDP growth will ensure the banking sector

Based on analysis and review of previous empirical studies, this paper, therefore, develops the following conceptual framework as display in Fig. 1.

262


Bank SpecificVariables 1. Capital Adequacy 2. Asset Quality 3. Management 4. Earning 5. Liabiilty 6. Sensitivity to market risk

Banking Sector Fragility Index (BSFI) 1= Bank is fragile (BSFI< 0) 0 = Bank is not fragile (BSFI>0)

Conceptual framework of bank fragility determinants

IV.

(1)

NPLt = ((TNPLt –TNPLt-1)/TNPLt-1)

(2)

DEPt = ((TDEPt –TDEPLt-1)/TDEPt-1)

(3)

DERt = ((TIERt –TIERt-1)/TIERt-1)

(4)

Once the level of banks fragility for individual bank is identified, this paper then uses the logit regression model to determine the bank specific and macroeconomics factors that are likely to cause bank fragility. Based on the constructed BSFI, the dependent variable is assigned a binary value equals to one if BSFI is < 0 and zero if BSFI is 0. The logit regression is shown in equation (5) and the expectationprediction evaluation by equation (6).

Macroeconomics Variable 1. GDP 2. Interbank Rate Fig. 1.

+((DERt-tier )/tier) 3

Methodology

(5)

The research methodology starts with the construction of bank sector fragility index for individual bank. Table III presents the three risks that are associated with the calculation of BSF index. BSF Index is calculated as an average of standardize value of NPL, DEP and TIER, where and stand for the arithmetic average and standard deviation of these variables respectively.

(6)

V. TABLE III. Economic Risks Credit Risk

The results of BSFI show the trend of fragility in Malaysian domestic commercial banks was in sequence with the event occurred (Fig. 2).

CONSTRUCTION VARIABLES FOR BSF INDEX Proxy Data Source Bank credit on private Bank Scope (2012) sector, NPL

The Liquidity Risk

Bank real total deposit, DEP

Bank Scope (2012)

Market Risk

Bank financial leverage, Time-Interest-Earned Ratio (TIER)

Bank Scope (2012)

The formula for the computation of BSFI computation is shown in equation (1). Different levels of bank fragility are determined as follows: a. b. c.

Findings

Fig. 2.

Not Fragile, if BSFI 0 Moderately Fragile, if Highly Fragile, if BSFI

Trend of average Banking Sector Fragility Index

The constructed BSFI revealed that the commercial banking sectors are in fragile condition from 1996 until 2000. They were particularly in a highly fragile stage between 1996 and 1998 since the BSFI is less than -0.5. This is not surprising because during that period, Malaysia financial institution was badly hit by the Asian financial crisis. The value of index was in a positive value in 2001 and 2002 before entering into slightly moderate fragile state in 2003. However, a year later in 2004, the commercial banking sector was not in the fragile zone where the highest BSFI of 0.86 points was recorded and was persistently above the zero level. However in 2009, the commercial banks entered into another fragile condition

Equation (2), (3) and (4) show the calculation of credit risks (NPLt), liquidity risks (DEPt) and market risk (TIERt) respectively. This paper uses annual data instead of using monthly data to avoid from the risk of deriving misleading interpretations toward fragility because of incoherent monthly data for selected variable [20]. BSFIt = ((NPLt -npl)/npl)+((DEPt -dep)/dep )

263


TABLE V. EXPECTATION-PREDICTION EVALUATION

when the BSFI was at -0.04 points. Between the year 2007 until 2008, the world was affected by the global financial crisis that led to the collapse of several large financial institutions. Fortunately, Malaysia financial institutions were not badly hit by it. The index rebounded back to positive value and maintains the positive value until 2011. Findings from this study proved that the measurement of BSFI given by previous studies like [2], [3] and [9] is a relevant approach in predicting of bank fragility.

Logit Regression Results of Commercial Banks Coeff.

Std. Error

Z-Stat

Prob.

Exp(B) a

Constant

99.2889

27.2811

3.6394

0.0003

CAR

-0.1234

0.1537

-0.8029

0.4220

0.8839

a

0.8057

TL/TA

-0.1260

30.2866

-3.9958

0.0001

DIE/TE

-3.6322

3.8075

-0.9539

0.3401

0.0265

ROA

-0.3130

0.4199

-0.7455

0.4559

0.7312

TL/TD

3.3473

1.8993

1.7623

0.0780b

28.4259

SZ

-0.3139

0.3566

-0.8801

0.3788

0.7306

IR

0.8120

0.2653

3.0607

0.0022a

2.2524

GDP

0.1662

0.0904

1.8386

0.0660b

1.1808

112

Obs with Dep=0

57

Obs with Dep=1

48

McFadden R-squared LR statistics Hosmer-Lemeshow statistics Prob. Chi-square

Total

% correct

78.95%

70.83%

75.24%

% incorrect

21.05%

29.17%

24.76%

VI.

Conclusion

Banking and financial sector is crucial for country development. It acts as an intermediary to allocate funds and controlled the level of liquidity for a country. However, banking and financial sector are faced with challenges to deal with the risks exposure during their business operations. Therefore, this research works on the determinants of bank fragility. Firstly, the non-parametric Banking Sector Fragility Index (BSFI) is used to measure individual banks’ fragility level. We tried to link the risk that associated with bank failure to measure bank fragility in term of index based. The results obtained from the constructed BSFI suggest are in tandem with the crisis events occurred in Malaysia for the period 1996-2011. Next, Logit regression is used to identify determinants that are likely to cause bank fragility in Malaysian banking sector during 1996-2011. Empirical evidence reveals interesting results for the Malaysian banking sector. It appears TL/TA and IR variables are the likelihood factors of bank fragility. The results indicate that the sign of fragility is determined by both bank-specific and economic factors. There are several suggestions for future research related to this study. Firstly, further research can be carried out by computing and comparing other indices such as speculative pressure index (SPI), index of currency market turbulence (ICMT) and excessive risk index (ERI) to test for the robustness of the indices. Secondly, researcher can also incorporate other independent variables that could be the factors for bank fragility. Lastly, other statistical methodology can be employed to investigate the causal relationship of bank fragility.

Summary Statistics for Logistic Regression Include Observation

Dep=1

Based on the expectation-prediction evaluation table (Table V), total prediction on bank is fragile and not fragile was correct at 75.24% and only 24.76% was incorrect. The overall model evaluation of the logistic regression shows that the model is well-specified. The p-value for the Hosmer and Lemeshsow test statistics is 0.4254 that indicates that the null hypothesis that there is no difference between the observed and predicted values is not rejected. MacFadden R-squared statistics (0.34) as well as the p-value of LR statistics that is statistically significant at 1% level also supported that the model is a good fit to the data.

TABLE IV. RESULTS OF ESTIMATED LOGIT REGRESSION MODEL Variable

Dep=0

0.34 50.0256 (0.0000)a 8.0834 (0.4254)

a

Significance at 1% level b Significance at 10% and

Based on results of the estimated logit model (Table II) the most significant bank-specific and macroeconomic factors (at 5% significance level) that influence the likelihood for bank fragility are asset quality (TL/TA) and interbank rate (IR). TL/TA shows a negative coefficient implying that a 1% increase in TL/TA led to the probability of bank being fragile reduces by about 0.81%. The finding concurs with the studies of [8],[9]and [13]. On the other hand, the interest rate (IR) has a positive coefficient, which indicates that, a 1 % increase in IR, the probability of bank being fragile increases by about 2.25%. [19] also reported similar relationship in their studies. They found that liquidity flow into the banking sector is affected when interbank rate increased and eventually led to fragility condition. Both liquidity (TL/TD) and GDP have positive relationship with BSFI but are statistically significant at 10% level.

Acknowledgment This paper will not be possible without the grant provided by Research Institute of Management, UiTM under the Research Intensive Faculty (RIF) Grant.

264


References [1] [2] [3]

[4] [5]

[6]

[7]

[8] [9]

[10] [11]

[12]

[13]

[14]

[15]

[16] [17]

[18] [19]

[20]

F. Allen and D. Gale, Bubble and crises. Economic Journal, pp. 236255, 2000. A. Kibritcioglu, Monitoring Banking Sector Fragility, The Arab Bank Review, Vol 5, No 2, 2003. C. H. Shen and C. F. Chen, Causality between banking and currency fragilities: A dynamic panel model. Global Finance Journal, 19(2), pp. 85-101, 2008. J. Tian, Banking sector fragility: Market failure, liquidity crunch and credit shock. State University of New York at Albany, 2010. F. Song and A. V. Thakor, Relationship banking, fragility, and the asset-liability matching problem. Review of Financial Studie , 20(6), pp. 2129-2177, 2007. B. Gonzalez-Hermosillo and R. Billings, Determinants of banking system fragility: A case study of Mexico. IMF Economic Review, 44(3), pp. 295-314, 1997 R. Duttagupta and P. Cashin, Anatomy of banking crises in developing and emerging market countries. Journal of International Money and Finance, 30(2), pp. 354-376. 2011. Eichengreen, Barry, A. Rose and C. Wyplosz, Contagious Speculative Attacaks, NBER Working Paper, No: 5681, 1996. A. Bayrakdaroglu, Financial crises and pressure indices: Case of turkey. International Journal of Economics and Finance Studies, Vol 2, No 1, pp. 1309-8055, 2010. J. F. Sinkey, A multivariate statistical analysis of the characteristics of problem banks, The Journal of Finance, 30(1), pp. 21-36, 1975. K., Männasoo and D. G. Mayes, Explaining bank distress in Eastern European transition economies, Journal of Banking & Finance, 33(2), pp. 244-253, 2009. R. A. Cole and J. W. Gunther, Separating the likelihood and timing of bank failure, Journal of Banking & Finance, 19(6), pp. 1073-1086, 1995. H. Dincer, G. Gemcer, N. Orhan and K. Sahinbas, A performance evaluation of the Turkish banking sector after the global crisis via CAMELS ratios, Procedia – Social and Behavioral Sciences, 24, pp. 1530-1545, 2011. M. Arena, Bank failures and bank fundamentals: A comparative analysis of Latin America and East Asia during the nineties using bank-level data, Journal of Banking & Finance, 32(2), pp. 299-310, 2008. C. A. Molina, Predicting bank failures using a hazard model: the Venezuelan banking crisis, Emerging Markets Review, 3(1), pp. 31-50, 2002. H. Hakenes and I. Schnabel, Bank size and risk-taking under Basel II, Journal of Banking & Finance, 35(6), pp. 1436-1449, 2011. D. Erdinç, Does a credit boom increase bank fragility?: Transitional challenges in Bulgarian Banking, 1999-2008, Eastern European Economics, 48(3), 68, 2010. G. Jacob and P. D. Wooldridge, Interbank rate fixings during the recent turmoil, Bis Quarterly Review, March 2008. N. Gunsel, Financial ratios and the probabilistic prediction of bank failure in North Cyprus, European Journal of Scientific Research 18(2), pp. 191-200, 2007. F. Sufian, Assessing the impact of financial sector restructuring on bank performance in a small developing economy, Ekonoska Istrazivanja, 23(2), pp. 11-33, 2010.

265