Abd. Ghafar b. Ismail Roselee Shah Shaharudin Ananda R. Samudhram Universiti Kebangsaan Malaysia. Abstract

Transcription

1 Abd. Ghafar b. Ismail Roselee Shah Shaharudin Ananda R. Samudhram Universiti Kebangsaan Malaysia Abstract The 1988 Basle regulations are meant to synchronise banking regulations worldwide and add to the stability of the global banking system. Basle II, due to be released in 2006, addresses some of the flaws in the 1988 accord. However, both versions do not close a loophole for managing earnings through discretionary adjustments of the loan loss provisions. Some researchers have found indications of income smoothing through loan loss provisions in banks, but others have concluded otherwise. All of these studies were conducted on banks outside of Malaysia. This study looks specifically at Malaysian banks, and uses a model based on bank-specific and macroeconomic factors that are peculiar to Malaysia. It concludes that Malaysian banks do not smoothen income. The likely reason is that good governance in Malaysian banks is driven more by regulatory measures imposed by the authorities than stock market discipline. Thus bankers do not seem to be concerned with managing earnings to present a rosy picture to investors. However, from a macroeconomic perspective, stacking up loan loss provisions during good times will help to release more credit during downturns in economic cycles, helping to soften the impact of recessions. Further research on whether Malaysian banks conduct capital management, to meet regulatory capital requirements, will help to shed more light on the behaviour of Malaysian banks. Keywords: earning management; loan loss provision; income smoothing; bank capital; 4 1

2 NATIONAL ACCOUNTING RESEARCH JOURNAL Introduction Banks have always played a key role in the economy, providing vital liquidity. But unlike other businesses the stability of the banking system is perception based and problems with single bank, real or assumed, can result in a run on all other banks, leading to credit crunches and escalation of national problems. With today's globally interlinked financial systems, problems with national economies can spread like wild fire over regional and international arenas, and rock the foundations of the world's economy. As such, it is not surprising that the banking sector tends to be heavily regulated across the globe. However, such regulations have generally been localized. This situation changed only in the late 1980s, when the G10 nations signed the 1988 Basle Accord that defined different categories of assets and associated risks, and set minimum regulatory capital requirements for internationally active banks. This Accord aims to synchronise minimum capital requirements for large banks across the globe, enhancing the stability of the worldwide banking system while warding off potentially damaging regional and global credit crunches. The 1988 Basle Accord has won praise for driving the adoption of risk-based capital ratios for measuring the stability of banks and propelling common regulatory standards in international banking (Santos, 2000). Despite compliance being voluntary, a large number of countries all over the world have adopted the Basle recommendations, compelling their banks to follow the Accord. However, this Accord is not without weaknesses. For instance, it applies the same risk weight to all commercial loans, whereas, in reality, loans given to large, established corporations are less risky than those disbursed to smaller, newer businesses. Since 1988, the Accord has undergone several amendments that addressed such weaknesses, incorporating the views of commentators and practitioners. In addition, a new Accord, known as Basle II, is nearing completion. Scheduled for release in 2006, Basle II will supersede the previous Accord. The new Accord tries to measure the risk attached to the different categories of assets more realistically. For example, commercial loans given to established organizations that enjoy strong ratings from external rating agencies will be assigned lower risk levels, compared with corporations saddled with a record of late payments or defaults. While the first Accord requires the banks to essentially meet and exceed minimum capital adequacy ratios, Basle II goes further. In addition to more sensitive risk assessment practices, it incorporates a supervisory process and the disclosure of pertinent information. The former allows for some flexibility in the use of the capital adequacy standards while the latter creates transparency, allowing the market to discipline poor governance. The capital adequacy requirements, supervisory and disclosure processes are known as the three pillars of the new Accord. 42

3 Do Malaysian Banks Manage Earnings Through Loan Loss Provisions? However, both the old and new Accords do not specifically address a loophole for manipulating the accounting numbers in financial statements, and presenting a picture of trends that are unrealistically attractive. Because bank managers have detailed information of their loan portfolios, which the regulators do not possess (information asymmetry), it is possible for them to adjust their loan loss provisions (LLPs) to smoothen income. They could provide lower reserves in bad years, resulting in lower overall expenses and higher overall profits, masking poor performance. In better years, managers could record higher reserves, with higher expenses that lower profits. The result is that the bank's published profits do not dip and rise according to actual performance, but show little fluctuation from year to year. However, these profits would reflect strategically created values, designed to bolster investors' confidence and the managements' compensation. Recent empirical research on banks, involving data from before and after the Accord period, does not always indicate the existence of income smoothing. The work of Beatty et al. (1995), Beaver and Engle (1996), Collins et al. (1995) and Rivard et al. (2003) suggests that banks do smoothen income while Ahmed et al (1999), Wetmore and Brick (1994) and Scheiner (1991) find otherwise. Empirical studies on income smoothing behaviour of Malaysian banks, especially in the post Accord period, are rare. This study is a preliminary exploration of such income smoothing behaviour in local banks, using empirical data and econometric modelling. The next section covers the hypothesis and methodology. The following section presents the results and the final part concludes. Hypothesis and methodology Hypothesis and model construction Unlike several earnings management studies on firms that use Jones model, the study of earnings management of banks among others by Ahmed et al. (1999) and Lobo and Yang (2001) use a single stage equation model to avoid under estimation of variable as suggested by Goldberger (1961). Based on the characteristics of data obtained, this study uses Generalized Least Squares method that is deemed more appropriate in estimating the relations of the variables The study hypothesizes that due to the roomy allocation of specific and general provisions. Malaysian banks managed earnings and smooth income by managing the largest accruals in its expense account, the loan loss provisions. 43

4 NATIONAL ACCOUNTING RESEARCH JOURNAL Discretionary and non-discretionary factors including the forces of macro economic environment can affect LLPs. To examine earnings manipulation, this study uses the following econometric model that includes bank specific and macroeconomic variables: LLP it = a + y/npl, + \ / 2 ALOANS + \ / 3 AROA (, + \ / 4 ROA POS jt + Where, 5 KLSE + ^; a, = constant ALLP /( = change in loan loss provisions for bank /' in the year t. ANPL /( = change in total non-performing loans for bank / in the year t. ALOANS /f = loan growth for bank /' in the year t AROA /f = earnings growth before provisions and taxes for bank / in the year t ROA POS lt = is the dummy variable for earnings before provisions and taxes for bank /'in the year t, that takes the value of 1 when earnings are positive and 0 otherwise. KLSE ;( = the annual average of the Kuala Lumpur Stock Exchange index for the year t Considering that the data cuts across different time periods and a host of different management styles within a cross section of banks, heteroscedacity (unequal variation) becomes a likely problem. A panel data regression model, using the GLS with fixed effects approach, takes into consideration these heteroscedacity concerns as well as the heterogeneous management and decision making styles of the different banks (Gujerati, 2003). The non-discretionary factors that influence LLPs are represented by ANPL and ALOANS. The amount of provisions attributable to these factors will conform to investors' expectations. Both higher NPL growths and higher loan growths will result in greater loan loss provisions, thus the ANPL and ALOANS will have a positive effect on LLPs. Managers can manipulate LLPs to attain specific earnings targets and strategies, including smoothing income. When ROA growth is negative, managers could understate LLPs and push up income, and vice versa. Thus AROA represents a discretionary factor that influences LLPs. A significant, positive relationship between LLPs and AROA would indicate income smoothing. The LLPs are also affected, in a discretionary manner, by a positive growth in earnings before provisions. In these circumstances, managers would be 44

5 Do Malaysian Banks Manage Earnings Through Loan Loss Provisions? motivated to increase LLPs, recording higher expenses and lower profits. Thus, income smoothing would be suggested by a positive relationship between LLPs and ROAPOS. The KLSE index represents the macroeconomic factors that affect LLPs. Higher KLSE indices indicate better economic conditions, leading to better business performances and lower defaults on loans. Lower KLSE indices have the opposite effect. Thus, theoretically, the KLSE index is negatively correlated with LLPs. Data source and description of analysis The data was obtained from the 1996 to 2002 annual reports, from a sample containing 21 Malaysian banks. The information was extracted from published annual reports as well as the banks' web sites. The annual averages were used for the KLSE stock index Results Table 1 presents the results of the analysis, using the Generalised Least Squares approach Table 1 : GLS Estimates of Coefficients Variable Coefficient t-statistic ANPL ALOANS AROA AROAPOS KLSE E ** * * * * Weighted Statistics Adjusted R-squared S.E. of regression F-statistic Prob(F-statistic) Durbin-Watson stat 'Significant at 5 % "Significant at 1 % 45

6 NATIONAL ACCOUNTING RESEARCH JOURNAL The results indicate that Malaysian banks generally do not smoothen income. In fact, income smoothing via loan loss provisions, as found in banks covered by current literature, is "neither illegal nor unethical" (Rivard et al. 2003). A consequence of not practicing income smoothing is that banks do not increase provisions in good times and stack up reserves that could allow for lower provisions in difficult times. Because limited provisions create a limitation on the loans that can be disbursed, this practice would dry up vital credit supply during economic downturns when demand for credit exceeds supply, particularly for promising smaller firms that would have had little trouble getting loans during booms. Since the supply of loans shrinks when it is most needed, this behaviour worsens the troughs in cyclical downturns, potentially leading to unnecessarily debilitating credit crunches in the normal economic down cycle. Conclusions Ongoing global level efforts to standardize banking have not specifically addressed the use of LLPs to smoothen income and paint an unrealistically attractive picture of banks' earnings. The management could be motivated to smoothen earnings with the view of improving market perception and enjoying performance based perks and rewards. Existing literature indicates the existence of income smoothing in banks, but lacks a consensus. Not much empirical research has been conducted on income smoothing behavior of Malaysian banks. Using data from a sample of Malaysian banks, this study finds that banks do not smoothen income. However, from a macro economic perspective, stacking up loan loss provisions during good times helps to improve the supply of loans during economic downturns and reduces the severity of recessions. Banking in Malaysia is tightly regulated, with the regulators imposing stringent conditions to ensure sound banking rather than the market disciplining poor management. Bank managers are likely more concerned with meeting regulatory requirements, such as relevant capital adequacy standards, than window-dressing accounting figures to embellish the appearance of financial reports during good times. Further research that investigates the presence of capital management'- adjusting accounting figures to meet capital requirement standards - could shed more light on the behaviour of Malaysian banks under the existing capital Accord, and give some indications on how the system might respond to the upcoming Basle II. Such information will help to provide some idea of the effectiveness of Basle II in strengthening the banking system, particularly in the Malaysian context. 46

7 Do Malaysian Banks Manage Earnings Through Loan Loss Provisions? References Ahmed A.S., Takeda, C. and Thomas, S (1999),. "Bank Loan Loss Provision: A Reexamination of Capital Management, Earnings Management and Signaling Effects,""Jouma/ and Accounting and Economics, 28, 1, pp Beatty, A, Chamberlain, S. and Magiiolo, J. (1995), "Managing Financial Reports of Commercial Banks: The Influences of Taxes, Regulatory Capital and Earnings," Journal of Accounting Research, 333, pp Beaver, W.H., Engel, E.E. (1996). "Discretionary Behaviour with Respect to Allowances for Loan Losses and Behaviour of Security Prices,""Journal and Accounting and Economics, 22,1, pp Collins, J.H., Shackelford, D.A. and Wahien, J.M. (1995). "Bank Differences in the Coordination of Regulatory Capital, Earnings and Taxes,"" Journal of Accounting Research, '33, 2, pp Goldberger, A. (1962) Best linear unbiased prediction in the generalized regression model. Journal of the American Statistical Association 57, Gujarati, D. (2003), "Basic Econometrics", 4,h Edition, McGraw-Hill/lrwin Lobo, G.J. and Yang, D.H. (2001) Bank managers heterogeneous decisions on discretionary loan loss provisions. Renew of Quantitative Finance and accounting, 16, Rivard, Bland and Morris (2003). "Income Smoothing Behavior of U.S. Banks Under Revised International Capital Requirements.""International Advances in Economic Research. V9, 14 pp Scheiner, J. (1991), "Income Smoothing: An Analysis in the Banking Industry,""Journal of Bank Research, pp Wetmore, J.L., Brick, J.R. (1994). "Loan-Loss Provisions of Commercial banks and Adequate Disclosure: Note,"''Journal of Economics and Business, 46, pp Santos, J. (2000), "Bank capital regulation in contemporary banking theory: A review of the literature", BIS Working Paper 90 47