The article Income Smoothing and Incentives: Empirical Tests Using Accounting Changes by 0. Douglas Moses written in April 1987 assumes that various firm-specific factors provide incentives for management to use accounting choices to smooth earnings; consequently, smoothing behavior varies across firms. Smoothing is associated with firm size, the existence of bonus compensation plans and the divergence of actual earnings from expectations. In addition, findings indicate that smoothing by accounting changes is related with the impact of the accounting change on the level of earnings. This is consistent with management recognizing and making trade-offs between the effect of accounting choices on both income levels and income variability. The hypothesis is as follows:
Various firm-specific factors provide incentives for management to use accounting choices to smooth earnings
This paper tests for differential smoothing behavior as a function of incentives by investigating the association between the degrees of apparent smoothing behavior exhibited within a sample of firms and firm-specific explanatory variables that proxy for factors that may motivate smoothing. A review of Accounting Trends and Techniques (ATT) for 1975-80 provided an initial sample of 1,483 firms reporting accounting changes. Firms were deleted from the sample due to various reasons. Deletions reduced the original sample to 231 discretionary accounting change events (16 percent of the ATT population). If a particular firm adopted more than one discretionary change in a given year, the effects were aggregated and the package was treated as a single event. This further reduced the sample size to 212 discretionary accounting change events. The variables are income smoothing, Firm Size/Political Costs, Market Share/Political Costs, Employee Costs, Bonus Compensation and Ownership Control. Income smoothing is the dependant variable and the rest are all independent variables. The purpose of this study was to examine the association between income smoothing and factors hypothesized to motivate it. The research question was "What factors explain variance in the smoothing measure SB?" Results suggest that smoothing is associated with large firms, firms with bonus plans, firms with a large pre change divergence of earnings from expectations, and firms whose accounting change tended to reduce earnings. The methodology used was t-test.
The article "Differential Valuation Implications of Loan Loss Provisions across Banks and Fiscal Quarters" written by Chi-Chun Liu, Stephen G. Ryan and James M. Wahlen in January 1997 states that loan loss provisions are positively related to bank stock returns and future cash flows, on the condition of less discretionary information about loan default. Increased discretionary loan loss provisions are good news only for banks which appear to have loan default risk problems. Moreover, prior literature also suggests that managers have incentives to delay income decreasing accruals until the fourth quarter when the audit occurs, implying that income decreasing accruals are more likely in the fourth quarter than in other fiscal quarters. The hypothesis of the article is as follows:
The market reaction to loan loss provisions is good news for "at risk" banks and bad news for "not at risk" banks.
Banks, Loan loss provisions, Stock returns and Cash flows are the variables in the article. Banks and loan loss provisions are independent variables whereas stock returns and cash flows are the dependant variables. Financial statement data and earnings announcement dates are obtained from the 1991 quarterly Compustat bank tape. All per share variables are adjusted for share distributions. Return data are obtained from the 1993 CRSP tape. Non-performing loan data for 31 quarters from the fourth quarter of 1983 to the second quarter of 1991 are obtained from Keefe, Bruyette, and Woods (KBW). Data for 104 banks result from merging the Compustat, CRSP and KBW data. The result of the finding was that the positive market reaction to loan loss provisions documented by prior research only obtains for fourth quarter loan loss provisions for "at risk" banks. It can be concluded that increased loan loss provisions of banks are perceived as good news by shareholders. The research question was "How do the costs and benefits of recognizing increased discretionary provisions vary across banks and across fiscal quarters?" The methodology used was a number of specification tests.
"Income smoothing behavior of U.S. banks under revised international capital requirements" by Richard J. Rivard & Eugene Bland & Gary B. Hatfield Morris in Nov, 2003 illustrates the use of the loan-loss provision to smooth reported income by large bank holding companies is a much investigated practice. To the extent that the variability of net income is a measure of risk, income smoothing may reduce the perceived riskiness of the bank and thus increase stock value. Managers may have added incentive to smooth income in response to the structure of their compensation package. The Basel Accord of 1988 phased in new definitions of regulatory capital for banks. These changes have increased the incentives for income smoothing. Most previous studies on income smoothing and loan-loss reserves predate the implementation of the Basel Accord. Others use data that include the transition period. This study revisits the subject, using only post-Accord data. Results of this study are compared with previous results. The evidence confirms the continued existence of income smoothing and supports the proposition that banks have become more aggressive in using loan-loss reserves as a tool for income smoothing.The hypothesis of this study is:
In the post-Basel Accord period, banks continued the practice of income smoothing using the provision for loan-losses.
Additionally, it is hypothesized that banks have increased their use of this accounting technique during the post-accord period. The added incentive to smooth income is generated by the division of retained earnings into Tier I capital and loan-loss reserves into Tier II capital, and from the limits on the use of loan loss reserves to meet capital requirements. As noted above, this change reduced the potential cost of income smoothing. This paper investigated the bank income-smoothing hypothesis in the aftermath of the 1988 Basel Accord. Using the same method employed by Greenawalt and Sinkey [1988] for the 1976-84 time periods, it analyzed the 1992-97 period. The present results confirm the earlier findings that large bank holding companies use the loan-loss provision as a means of leveling net earnings across time periods.
The results also suggest that in the post-Accord period, banks accelerated their use of this income-smoothing technique. By excluding the reserve for loan-loss from Tier I capital, and by restricting the amount which applies even to Tier II capital, the new system reduces the potential cost associated with understating the provision for loan-loss during periods of low earnings.
The article "Bank provisioning, business cycles and bank regulations: a comprehensive analysis using panel data" written by Meng-Fen Hsieh (Taiwan), Chung-Hua Shen (Taiwan) and Chien-Chiang Lee (Taiwan) in 2008, states that the with steady growth in both the economy and bank earnings, bank management will tend to increase loan loss provisions (LLP), whereas with a buoyant economy but negative growth in bank earnings, management will exhibit an inclination to reduce LLP. Regarding the influence of bank regulation on provisions, the evidence shows that under certain circumstances banks make more provision based on regulatory considerations. This explains why bank regulations regarding LLP across countries do have an effect on the banks' provisioning behavior. The variables in the study are loan loss provisions, pro-cyclicality, income-smoothing, dynamic panel data. Loan loss provision is the dependant variable and the rest are all independent variables. The hypothesis is as follows:
The intense income-smoothing effect is at play under taking regulations into consideration
The first objective of this article is to explore the relationships among business cycles, earnings and bank loan loss provisions by employing recent two step system GMM techniques developed for dynamic panels on a panel of 49 countries observed in the period of 1991-2002. The evidence shows that with steady growth in both the economy and bank earnings, bank management tends to increase LLP, whereas with a buoyant economy but negative growth in bank earnings, management shows a tendency to reduce LLP. In these scenarios, the income-smoothing effect appears to be held and less affected by business cycles. By contrast, when the economy is in a downward trend and banks suffer losses, management evidently increases LLP.\
The article "A signaling approach to the provision for loan losses" written by Joyce, William B, Woehrle and Stephen L in 2001 states that the provision for loan losses has a discretionary aspect. The bank manager has private information relative to the success of loan restructuring, and he/she has discretion over both the amount and the timing of the provision. If it is assumed that financial markets are not fully aggregating in the sense that market prices do not reflect all information, especially that which is not publicly available, then it is possible that managers may elect to use financial policy decisions to convey information to the market. Under certain circumstances, the manager may credibly signal his/her private information to financial markets through his/her provision policy. According to the hypothesis of the article:
If financial markets are not fully aggregating (in the sense that market prices do not reflect all information, especially that which is not publicly available), then it is possible that managers may elect to use financial policy decisions to convey information to the market.
The change in the provision for loan losses is advanced as a candidate for a signaling device.
The loan-loss reserve reflects the reserves needed to cover future loan losses. The level is increased with the provision for loan losses. The level is decreased by loans that are charged off. Because recoveries are made on some bad loans, only net loan losses deplete the level of the reserve. Banks increase their provision for loan-losses in order to write-- down loans closer to market values; the loan should be valued at the price it would command if traded in the open market. When banks do a better job of matching accounting income with economic income, the divergence between book and market values of loans is minimized or eliminated. The reserve for loan losses may also be influenced by the information content of the provision for loan losses. The information content or signaling of the provision for loan-losses is a means for bank managers to "signal" their private information with respect to loan quality.
The article "Determinants of Signaling by Banks through Loan Loss Provisions" written by Gerald J. Lobo and Dong-Hoon Yang on June 2003 discusses whether bank managers use their discretion in estimating loan loss provisions to convey information about their banks' future prospects. Bank Managers face different conditions and have different incentives. The variables used in the article are Signaling, Loan Loss Provision, Earnings Variability, Investment Opportunity Set and Income Smoothing. The hypothesis in the article states that:
The propensity to signal varies negatively with bank size it differs positively with earnings variability, future investment opportunities, and degree of income smoothing.
Propensity to signal is positively linked to the degree of information asymmetry and bank managers satisfy perceived undervaluation of their banks by communicating their private information about their banks' favorable future prospects. The sample is selected from those banks included in the 1997 Bank Compustat annual data files. Non-performing loans information for the period 1980-1997 is obtained from Keefe,Bruyette, and Woods (KBW). The 1997 Bank Compustat data file contains 14,080 bank-year observations for 705 banks for the period 1978-1997. Eliminating banks that lack the necessary financial data leaves 3,068 bank-year observations. From these, 1,410 bank-year observations are deleted due to first-differencing in total outstanding loans and one-year-ahead changes in cash flow as defined in the loan loss expectation model. The methodology used in the article is mostly secondary research. Prior research is used to make future decisions.
According to Gerald J. Lobo and Robert Mathieu, in the article "Managerial Incentives for Income Smoothing through Bank Loan Loss Provisions," illustrate the reasons as to why bank managers use loan loss provisions to smooth the reported income. The study predicts that for banks with good or poor current performance and expected poor or good future performance, managers will save income for the future by reducing or increasing current income through LLP. The three variables used in the study are income smoothing, bank loan loss provisions and job security. The difference in LLP between the two groups of banks is positive as hypothesized, indicating that:
Bank managers do save earnings through LLP in good times and borrow earnings using LLP in bad times.
Similar results are obtained for analysis using discretionary LLP. When bank managers are saving earnings for the future, we provide evidence that the need to obtain external financing is an important additional variable in explaining cross-sectional differences in the extent of income smoothing. Furthermore, whether or not a bank is well capitalized is also weakly significant in explaining cross-sectional differences in income smoothing. To measure income smoothing, the article follows the methodology used by DeFond and Park (1997). More specifically, it classifies bank-quarter observations into four quadrants based on (i) good-poor current performance and (ii) good-poor future performance. Good and poor performance banks are defined as banks with earnings before tax and loan loss provisions (deflated by beginning total assets) above and below the quarter's industry median earnings, respectively. Based on the analytical predictions of Fudenberg and Tirole (1995), we classify the good-poor and poor or good quadrants as the income-smoothing quadrants. Empirical analysis is based on 4,166 bank-quarter observations. The sample consists of US bank holding companies for the period 1987 to 2000. Quarterly information is obtained from the Call Reports filed by bank holding companies with the Federal Reserve Bank. Inclusion in the sample is based on the following three criteria: 1) the total assets exceed $300 million; 2) the Call Report is available for every quarter between 1987 and 2000 and 3) the bank is publicly listed. A total of 91 banks satisfied these conditions.
The article "Income Smoothing, Earnings Quality and Firm Valuation" written by BEN-HSIEN BAO AND DA-HSIEN BAO in December 2004 argues that lower variability of earnings does not guarantee income smoothers' higher firm values. Instead, smoothers' earnings should be more value-relevant if they are of high quality, i.e., earnings quality should be considered simultaneously. Sample firms are divided into four groups: quality earnings smoothers, quality earnings non-smoothers, non quality earnings smoothers, and non-quality earnings non-smoothers. Value relevance of reported earnings is then studied using both the levels and the changes approaches with indicator variables. Results show quality earnings smoothers have the highest price-earnings multiple while non-quality non smoothers have the lowest price-earnings multiple. The variables in the article are income smoothing, earnings quality, earnings multiple and firm valuation. According to Ben-Hsien Bao and Da-Hsien Bao:
Reported earnings of smoothers are more value-relevant if the reported earnings are of high quality.
1988 to 2000 sales and primary earnings per share before extraordinary items are used to calculate coefficient of variation for one period change in sales and coefficient of variation for one period change in earnings.14 Coefficient of variation is defined as standard deviation divided by mean value (Kendall and Stuart, 1958). Both standard deviation and mean value are determined using six years of data, e.g., 1988 to 1993 data are used to calculate five one-year changes, and coefficients of variation for 1993 are calculated using the five one-year changes. A firm in each of the eight years from 1993 to 2000 is classified as a smoother if the coefficient of variation for one-year change in earnings is smaller than that for one-year change in sales.
"Use of loan loss provisions for capital, earnings management and signaling by Australian banks," written by Asokan Anandarajan, Iftekhar Hasan and Cornelia McCarthy examines whether and to what extent Australian banks use loan loss provisions (LLPs) for capital, earnings management and signaling. It is examined in the research that there were changes in the use of LLPs as a result of the implementation of banking regulations consistent with the Basel Accord of 1988, which made loan loss reserves no longer part of Tier I capital in the numerator of the capital adequacy ratio. We find some evidence to indicate that Australian banks use LLPs for capital management, but we find no evidence of a change in this behavior after the implementation of the Basel Accord. Our results indicate that banks in Australia use LLPs to manage earnings. Furthermore, listed commercial banks engage more aggressively in earnings management using LLPs than unlisted commercial banks. We also find that earnings management behavior is more pronounced in the post-Basel period. Overall, we find a significant understating of LLPs in the post-Basel period relative to the pre-Basel period. This indicates that reported earnings might not reflect the true economic reality underlying those numbers. Finally, Australian banks do not appear to use LLPs for signaling future intentions of higher earnings to investors. The variables used in the study are Australian banks; Capital management; Earnings management and Signaling. The hypothesis states that the relation between LLPs and primary (Tier I) capital for commercial banks will be less negative in the post-Basel regime relative to the pre-Basel regime. Moreover, the relation between LLPs and earnings (before LLPs) will be more positive in the post- Basel regime relative to the pre-Basel regime. The data in our study are limited to commercial banks. During the period 1991-2001, the commercial banks in our sample possessed 87 per cent of the assets in the Australian banking industry and were subject to one major regulatory change. This change was the implementation of the capital adequacy guidelines of the Basel Accord of 1988.The methodology used is secondary research.
