Quality of Reported Earnings: An Empirical Study of Indian Banking Industry

Abstract

Earnings management (EM) practices by bank managers can prove to be very precarious in smooth running of the financial system of a country. The failure of financial system shocks the entire economy. The present paper aims to assess the quality of earnings in the Indian banking industry. EM is estimated by employing a bank-specific model that measures EM through loan loss provision (LLP) and realized security gains and losses (RSGL). The findings exhibit that public banks practice income increasing and private banks practice income decreasing EM, whereas, combined result reinforces the practice of income decreasing EM. The results indicate that public sector banks use both LLP and RSGL to manage earnings whereas private sector banks increasingly rely on RSGL. Further, direction of EM is gauged by classifying EM on the basis of quartiles. This study has implication for regulators, investor and depositors. Regulators should be stricter regarding policies of LLP. Apart from earnings, investor and depositors should be considered other measures of stability of banks like capital adequacy ratio because earnings may be manipulated.

Keywords

Earnings Management Banking Industry Quality of Earnings Loan Loss Provision Realized Security Gains and Losses

Introduction

The main pillar of economic prosperity in any country is its financial sector. The strength and efficiency of financial sector are directly related to economic growth. Financial sector fills the gap between lenders and borrowers by providing a place where money flows from the people who are in excess of funds to the people who needed it most. Banks are main element of a financial sector. They play a vital role in mobilization of household savings via disbursing credit to various sectors of an economy. Banks have special role to play in India because as per India Brand Equity Foundation (2020) they dominate financial sector, accounting for more than 64% of total assets held by financial system.

The current financial system of India is swinging amidst despair and hope. One of the articles in daily newspaper Business Standard (Nayyar, 2019, p. 8) reported ‘How ironic it is that a decade on, it is India’s supposedly well-regulated and tightly controlled banking sector that is in a complete mess, paralyzed even as the once excess-ridden banks of the West have got their mojo back’. Now the concern is what happened within a decade that made Indian financial system fragile? Why did it happen? The reason is that in this decade banks faced a series of defaults and scams such as Vijay Mallya–PNB, Nirav Modi–PNB, PMC Bank, Yes Bank and defaults by Jet Airways, HDIL Group. In all these cases some insiders (top-level managers or executives) were involved with the main perpetrators. Managers produced desired financial numbers by managing earnings to divert away the regulator’s attention from irregularities. They do not identify bad loans at an earlier stage and make inadequate loan loss provision (LLP) that is meant to absorb default loan losses in future. The application of managerial discretion in reported accounting numbers to produce desired financial statement for their personal motive is termed as earnings management (EM). Management of earnings may produce desired results in short term but has serious negative repercussions like bankruptcy in long term. Business Standard has reported that Yes bank which is at the verge of bankruptcy has bad loans that are four times what Yes Bank had reported in its audited financial statements (Mukherjee, 2019, p. 4). EM results in poor financial performance of banks in long run (Alhadab & Al-Own, 2017; Ashfaq & Saeed, 2017; Ujah et al., 2017; Umoren et al., 2018; Wu et al., 2016). Thus, it becomes inevitable to detect EM at early stages. The main objective of the present paper is to assess whether Indian banks manage earnings or not?

Review of Literature

The stakeholders want financial assurance and stability of the company to which they are going to transact with and this assurance is provided by financial statements. Financial statements of a company reflect the information of earnings which are symbol of strength of the company. All the companies have pertinent reasons that they should be perceived as financially sound and stable. Managers accomplish this motive by practising EM. They use the discretion provided by various accounting standards and laws in financial reporting to produce desirable financial statements. Although EM results in manipulated financial numbers, it cannot be directly identified through the financial statements. Sun and Rath (2010) discuss a number of approaches, that is, total or discretionary accruals, specific accruals, changing accounting choices, real activities, earnings distribution that are used to measure EM. The most common approach is discretionary accruals which are widely used in literature. To quantify discretionary accruals as proxied for EM, a number of models—that is, Healy (1985), DeAngelo (1986), Jones (1991), Dechow et al. (1995) and Modified Jones model—have been developed in literature. Modified Jones model for calculating discretionary accruals is widely appreciated in literature (Myers & Skinner, 2000; Peasnell et al., 2000; Evans et al., 2012; Osma & Noguer, 2007; Sarkar et al., 2008). But in case of banking industry, these models do not seem to fit in perfectly.

The nature of business of banking industry differs from non-financial firms. In banks, a single accrual, that is, LLP has great significance as it forms major part of total accruals. LLP is reported in income statement on expense side, set aside by bank managers on the basis of their assessment of future loan losses (Cornett et al., 2009). Accumulated LLP is accounted in balance sheet as loan loss allowance (LLA) which is a type of capital and during bad times it can be used to absorb losses (Grougiou et al., 2014). LLP has two parts, nondiscretionary and discretionary parts. Nondiscretionary LLP brings LLA to an admissible level and discretionary LLP (DLLP) provides room for managerial discretion, being the reason for its close monitoring by regulators (Cornett et al., 2009). In banking EM literature, majority of the studies exhibit that banks manage earnings via LLP (Anandarajan et al., 2003; Anandarajan et al., 2007; Curcio & Hasan, 2015; Ghosh, 2007; Kanagaretnam et al., 2003, 2004; Leventis & Dimitropoulos, 2012; Ozili, 2017; Ozili & Outa, 2018; Vishnani et al., 2019). To the contrary, Ahmed et al. (1999) and Beatty et al. (1995) did not find any evidence of EM by banks through LLP. Some studies reported that earnings can be managed through changing the time of security gains and losses (Beatty et al., 1995; Beatty & Harris, 1998; Collins et al., 1995). Yet another chunk of literature uses both LLP and security gains and losses in combination (Beatty et al., 2002; Bratten et al., 2017; Cheng, 2012; Cohen et al., 2014; Cornett et al., 2009; Grougiou et al., 2014; Kumari & Pattanayak, 2017) to measure EM.

Beatty et al. (2002) developed a bank-specific model to measure EM using LLP and realized security gains and losses (RSGL) in combination and found that in order to avoid small earnings decreases public banks are more prone to use DLLP and discretionary realized security gains and losses (DRSGL) than private banks.

Discussing EM in context with Indian banks, Chipalkatti and Rishi (2007) observed that banks that are weak in terms of profitability and capital adequacy ratio engage in aggressive EM to compensate for low earnings. Ghosh (2007) opined that banks listed on recognized stock exchange more aggressively indulge in EM via LLP than unlisted counterparts. Kumari and Pattanayak (2017) found evidence of income increasing EM in Indian commercial banks through LLP and RSGL. Vishnani et al. (2019) also exhibit that Indian banks practice EM. The review of EM in the Indian banking industry indicates that banks are actively engaged in managing earnings. So, there is a need to explore management of reported earnings by banks in India. Presence of few studies on EM in the Indian banking industry is the prime motivation for this paper. The main objective of this paper is to evaluate the quality of reported earnings of Indian banks.

Research Design

Sample Selection and Data Collection

The sample examined in the present paper includes all the scheduled public and private sector commercial banks operating in India during six-year time period starting from financial year (FY) 2013–2014 to 2018–2019. The sample consists of 18 public and 22 private banks. Two private banks, that is, Bandhan Bank Ltd and IDFC First Bank Ltd, were incorporated in 2015. So, the final data include in total 236 banking years observation.

The data for DLLP estimate were obtained from ProwessIQ, a database of Centre for Monitoring Indian Economy (CMIE) and data for DRSGL estimate were obtained from Reserve Bank of India (RBI) website.

Measurement of Earnings Management

EM cannot be directly measured from financial statements of companies. It is estimated via some proxies of managerial discretion employed by managers. The widely used measure of EM is Modified Jones model (Dechow et al., 1995). But this model does not seem to be fit to banking industry as banking industry differs from other industries in terms of structure, nature of business and regulations. The accruals used as proxy for EM also differ in banking industry. So, in the present paper, EM is measured through Beatty et al. (2002) model that is compatible with the unique features of banks. In Beatty et al. (2002) model, two measures are used to calculate EM, namely DLLP and DRSGL.

Earnings of a bank primarily vary in accordance with the performance of its loan portfolio. ‘Loans over 90 days past due and still accruing interest as well as loans no longer accruing interest are observable measures of the current loans at risk of default’ and a provision is set aside on these bad loans called LLP that provide ample discretion to managers for creating provision on bad as well as good loans (Cohen et al., 2014). ‘LLPs are an expense item reported on the income statement reflecting bank managers’ current period assessment of the level of future loan losses’ while accumulated LLP is accounted in balance sheet as LLA which is a type of capital and during bad times it can be used to absorb losses (Grougiou et al., 2014). LLP has two parts, nondiscretionary part and discretionary part. Nondiscretionary LLP brings LLA to an admissible level and DLLP provides room for managerial discretion, being the reason for its close monitoring by regulators (Cornett et al., 2009). Thus, managers may manage earnings by applying their discretion in making provision on bad loans. They may overreport (underreport) LLP to decrease (increase) earnings. Grougiou et al. (2014) defined ‘RSGLs is the difference between the most recent mark-to-market price and the proceeds from the sale or redemption of the security’. When sale proceeds are greater than its most recent mark-to-market price then security gain occurs and when sale proceeds are less than the most recent mark-to-market price then loss occurs (Grougiou et al., 2014). Managers may increase or decrease earnings by changing the timing of sale or redemption of an investment security. Further, shareholders, auditors or regulators cannot challenge the decision of managers of sale of an investment security because it is an unaudited and unregulated discretionary management action (Cornett et al., 2009).

The first measure of EM, DLLP is estimated by following Beatty et al. (2002) and Cornett et al. (2009). The method of financial reporting of Indian banks differs from U.S. banks. Thus, classification of loans given in original model is substituted by loan classification presented in Indian banks’ financial statements. A new variable, bad debt written off (BDW) is incorporated in the model as per Kumari and Pattanayak (2017) as it may be significant for the Indian banking industry. To estimate DLLP data are pooled across all years. The modified version of model can be described as:

\begin{array}{l} L L P_{i t} = α + β_{1} L A S S E T_{i t} + β_{2} N L P_{i t} + β_{3} B D W_{i t} + \\ β_{4} L L R_{i t} + β_{5} T L O A N_{i t} + β_{6} L O A N T A_{i t} + β_{7} L O N G_{i t} + \\ β_{8} L O A N P S_{i t} + β_{9} A D V P S_{i t} + β_{10} L O A N F_{i t} + ε_{i t} \end{array}

(1)

Where:

i = Banking company identifier ranging from 1 to 40;

i = Year ranging from FY2013–2014 to FY2018–2019;

LLP = Loan loss provision as a % of TL;

LASSET = Natural log of TA;

NPL = Non-performing loans as a % of TL;

BDW = Bad-debt written off as a % of TL;

LLR = Loan loss allowance as a % of TL;

TLOAN = Term loan as a % of TL;

LOANTA = Loans secured by tangible assets as a % of TL;

LOANG = Loans secured by bank/govt. guarantee as a % of TL;

LOANPS = Loans to priority sector as a % of TL;

ADVPS = Advances to public sector as a % of TL;

LOANF = Loans to foreign country as a % of TL;

ε = Error term.

The fitted value in above regression Equation (1) imitate normal loan losses as per the loan portfolio composition and the error term is treated as abnormal or DLLP (Beatty et al., 2002; Bratten et al., 2017; Cheng, 2012; Cohen et al., 2014; Cornett et al., 2009; Grougiou et al., 2014).

As per Cornett et al. (2009), DLLP (error term of Equation (1)) is estimated as percentage of total loans. But the second measure DRSGL is standardized by total assets. Thus, DLLP is also transformed and defined as:

\frac{D L L P_{i t} = ε_{i t} \times L O A N S_{i t}}{A S S E T S_{i t}}

(2)

Where:

LOANS_it = TL of ‘ith’ bank in the year ‘t’;

ASSETS_it = TA of ‘ith’ bank in the year ‘t’.

To measure DRSGL, again Beatty et al. (2002) and Cornett et al. (2009) are followed and data are pooled across all bank-years. The model is defined as:

R S G L_{i t} = α + β 1 L A S S E T_{i t} + β 2 U N G L_{i t} + ε_{i t}

(3)

Where:

i = Banking company identifier ranging from 1 to 40;

i = Year ranging from FY2013–2014 to FY2018–2019;

RSGL = Realized security gains and losses as percentage a % of TA;

LASSET = Natural log of TA;

UNGL = Unrealized security gains and losses as a % of TA.

Here, TA refers to total assets and TL refers to total loans.

The error term of the regression Equation (3) is termed as DRSGL (Beatty et al., 2002; Bratten et al., 2017; Cheng, 2012; Cohen et al., 2014; Cornett et al., 2009; Grougiou et al., 2014; Kumari & Pattanayak, 2017)

Finally, Cornett et al. (2009) explained that DRSGL has positive relation with EM and DLLP has negative relation with EM. Thus, EM is defined as:

E M_{i t} = D R S G L_{i t} - D L L P_{i t}

(4)

High level of earning management implies higher DRSGL and lower DLLP which results in increased earnings. On the other hand, low level of EM shows that there is overreporting of LLP and less security gains are realized which results in decreased earnings.

Results and Discussions

Descriptive statistics of variables that are used to measure DLLP and DRSGL over a period of six years (2014–2019) are reported in Table 1. Descriptive statistics of LOSS, NPL, BDW, LLR, TLOAN, LOANTA, LOANG, LOANPS, ADVPS and LOANF are shown as percentage of TL while RSGL and UNGL are shown as a percentage of TA. The mean value of NPL of all the banks is 3.932, suggesting that out of total loans, 3.932% are at default risk. For these bad loans, on an average 2.409% LLP (LOSS) is made on total loans. The accumulated loss represented by LLR has a mean value of 3.599%. Mean value of BDW is 1.447% exhibiting that out of total loans 1.447% loans are deemed uncollectible and hence written-off net of recovery. The minimum level of NPL and LOSS is 0 (recorded by Nainital Bank Ltd) while the maximum level of NPL and LOSS are 16.691% and 20.138% respectively (recorded by IDBI Bank Ltd).

The standard deviation of NPL and LOSS are 3.421% and 2.511% indicates that the level of bad loans and their provision are largely dispersed between banks. LLR is between 0.069% and 23.973% with a standard deviation of 3.695%.

The dispersion of BDW ranges from 0 to 14.506%. The mean values of different loan classification indicate that in a bank’s loan portfolio on an average 56.281% are disbursed as term loans, 84.312% loans are secured by tangible assets, 3.713% loans are secured by bank or government guarantee, 36.06% loans are disbursed to priority sector, 4.889% to public sector and 5.047% to foreign countries.

The mean value of RSGL suggests that on an average, banks report realized gain of 0.226% of total assets while unrealized loss of 0.002% of total assets is presented on average. The minimum value of RSGL is −0.289% (realized loss reported by Dhanlaxmi Bank Ltd) and maximum value is 1.207% (realized gain reported by CSB Bank Ltd) with a standard deviation of 0.202% suggesting that the level of RSGL is not much dispersed between banks. The level of UNGL varies between −0.089 to 0.058% with a standard deviation of 0.012%.

The results of pooled OLS regression of estimating DLLP and DRSGL are presented in Table 2. The adjusted R-squared in estimating DLLP is 0.917 and 0.018 in estimating DRSGL.

The results exhibit that NPL and LLR are positively and significantly related with LLP, consistent with Beatty et al. (2002), Cheng (2012), Grougiou et al. (2014) and Cohen et al. (2014). It suggests that when the problem of bad loans and amount of LLA increases, then there is an increase in LLP also. Cornett et al. (2009) found negative coefficient of LLR with LLP. The significant and positive coefficient of BDW shows that when the probability of un-collectability of loans increases then LLP also increases, consistent with Cheng (2012). Further, LLP is positively correlated with bank size (LASSET), term loan (TLOAN), loan secured by tangible assets (LOANTA) and negatively with loan secured by bank/government guarantee (LOANG), priority sector loan (LOANPS), public sector loan (ADVPS) and foreign countries loan (LOANF).

The results of estimating DRSGL regression model exhibit that RSGL are negatively and significantly correlated with unrealized security gains and losses, contrary with Beatty et al. (2002), Cornett et al. (2009), Cheng (2012), Grougiou et al. (2014) and Cohen et al. (2014). This might be attributed to difference in reporting standards among these countries. For example, Cornett et al. (2009) computed URSGL as ‘unrealized security gains and losses (includes only unrealized gains and losses from available-for-sale securities)’. But in annual reports of Indian banks, specified reporting of unrealized gains and losses of different securities is not available as these banks have not shifted to the new reporting standards, that is, Ind-AS.

Table 1.

Descriptive Statistics of Variables Used to Measure DLLP and DRSGL, 2014–2019

	Mean	Median	Std. Dev	Minimum	Maximum	N
LOSS (%)	2.409	1.535	2.511	0	20.138	236
NPL (%)	3.932	2.907	3.421	0	16.691	236
BDW (%)	1.447	0.911	1.738	0	14.506	236
LLR (%)	3.599	2.299	3.695	0.069	23.973	236
TLOAN (%)	56.281	54.484	16.122	12.436	99.32	236
LOANTA (%)	84.312	86.375	10.314	26.287	99.654	236
LOANG (%)	3.713	2.175	5.83	0	63.905	236
LOANPS (%)	36.06	35.741	12.407	3.567	98.575	236
ADVPS (%)	4.889	3.806	4.796	0	26.819	236
LOANF (%)	5.047	0	7.771	0	31.816	236
RSGL (%)	0.226	0.172	0.202	−0.289	1.207	236
UNGL (%)	−0.002	0	0.012	−0.089	0.058	236
Total Assets (₹ Million)	3,090,000	1,970,000	4,640,000	53425.9	36,800,000	236
Total Loans (₹ Million)	1,860,000	1,200,000	2,780,000	24370.4	21,900,000	236

Source: The authors.

Table 2.

Regression Analysis of Loan Loss Provision and Realized Security Gains and Losses, Sample Period 2014–2019

Independent Variables	Loan Loss Provision Coefficient Estimate (P-value)	Realized Security Gains and Losses Coefficient Estimate (P-value)
LASSET	0.0827 (0.160)	0.0109 (0.286)
NPL	0.120*** (0.000)
BDW	0.637*** (0.000)
LLR	0.303*** (0.000)
TLOAN	0.0044 (0.198)
LOANTA	0.0085 (0.183)
LOANG	−0.00198 (0.838)
LOANPS	−0.0026(0.583)
ADVPS	−0.0217* (0.0618)
LOANF	−0.0056 (0.556)
UNGL		−2.236** (0.0469)
Constant	−1.981* (0.0839)	0.067 (0.645)
Observations	236	236
Adjusted R-squared	0.917	0.018

Source: The authors.

Note: ***, ** and * presents significant at 1%, 5% and 10% level, respectively.

Table 3 reports summary statistics of the variables present in regression Equation (4), estimated over the time period of 2014–2019. The variables reported in Table 3 are presented as a percentage of TA. Panel A presents EM variables on aggregate basis of both public and private sector commercial banks. The results exhibit that mean of DRSGL is 0.000% and of DLLP is 0.007%, resulting in EM as difference between DRSGL and DLLP is −0.007%. Negative EM indicates that Indian banks practice earnings decreasing EM via overreporting DLLP. Note that mean of DRSGL is 0 because it is measured as regression residuals of Equation (3) and as per one of the assumptions of OLS regression the mean value of error term should be zero for a regression model to be correctly specified (Gujarati et al., 2018). DLLP is also estimated as regression residual of Equation (1) but it is non-zero because it is transformed as per Equation (2). The level of DRSGL varies between −0.483 to 1.009% with a standard deviation of 0.199%, suggesting that DRSGL is not much dispersed between banks. DLLP varies between −1.077 and 1.594% with a standard deviation of 0.399%. It indicates that level of DLLP is more dispersed than DRSGL between banks, some banks underreport DLLP and some overreport it to manage the earnings in desired direction. Although the level of EM of sample banks during the period of six-year is close to zero, but note that these figures are scaled by total assets and in their raw form these play a significant role in manipulating earnings.

Table 3.

Summary Statistics on Earnings Management Variables, 2014–2019

	Mean	Median	Std. Dev	Minimum	Maximum	N
Panel A: Aggregate
DRSGL (%)	0.000	−0.054	0.199	−0.483	1.009	236
DLLP (%)	0.007	0.009	0.399	−1.077	1.594	236
EM (%)	−0.007	−0.048	0.459	−1.723	1.448	236
Panel B: Public Sector Commercial Banks
DRSGL (%)	0.015	−0.051	0.204	−0.335	0.843	108
DLLP (%)	−0.035	−0.024	0.452	−1.077	1.459	108
EM (%)	0.05	0.005	0.538	−1.723	1.448	108
Panel C: Private Sector Commercial Banks
DRSGL (%)	−0.012	−0.055	0.195	−0.483	1.009	128
DLLP (%)	0.042	0.036	0.346	−1.062	1.594	128
EM (%)	−0.055	−0.059	0.376	−1.518	0.92	128

Source: The authors.

Panels B and C of Table 3 report summary statistics of EM variables of public and private sector commercial banks respectively. The positive mean value of EM (0.05%) exhibits that public sector commercial banks practice income increasing EM via realizing more security gains (0.015%) and underreporting DLLP (−0.035%). While negative mean value of EM (−0.055%) of private sector commercial banks shows that these banks practice income decreasing EM via realizing more security losses (−0.012%) and overreporting DLLP (0.042%). One possible reason for this might be that public sector banks are regulated by not only regulator but from government also. So, managers of public sector banks are more pressurized than private banks to meet earnings targets. That is why they manage earnings upward.

Table 4 reported the quality of earnings in the Indian banking industry for the sample year. Significance level of mean and median are checked through two-tailed t-test and Wilcoxon test respectively. Panels A, B and C report results of both (public and private sector banks) on aggregate basis. Public sector banks show significant results, while private sector banks’ results are significant only for DRSGL indicating that public banks are actively engaged in EM than private banks.

The results indicate that public sector banks use both LLP and RSGL to manage earnings whereas private sector banks increasingly rely on RSGL. The results of EM of public banks are significant in almost all years, however, they did not follow any particular trend with regard to direction of EM. They practice negative and positive EM in alternative years. On the other hand, private banks practice negative EM. Per cent positive values of EM also shows that public banks practice income increasing EM (0.778 in year 2019) more than private banks (0.318 in year 2019). Aggregate results of both sectors show negative EM in earlier years and positive in recent years. Year 2018–2019 does not produce significant results because public banks practice income increasing and private banks practice income decreasing EM. Further, proportion test results stated that earlier, less banks practice income increasing EM but later more banks are engaged in positive EM as depicted by 0.8, 0.35 and 0.525% in 2017, 2018 and 2019 years respectively. The magnitude and significance level of DRSGL and DLLP exhibit that in earlier years public banks use more DRSGL but later they shift to DLLP, while, private sector bank uses more DRSGL than DLLP. As the results of public banks are more powerful than private banks, so, aggregate results also match with public bank’s results.

Table 4.

Quality of Earnings in Indian Banking Industry, 2014–2019

	Variables	2013–2014	2014–2015	2015–2016	2016–2017	2017–2018	2018–2019
Panel A: Aggregate
No. of observations		38	38	40	40	40	40
Mean	EM	−0.095**	−0.005	−0.186***	0.327***	−0.141*	0.054
Median	EM	−0.107***	−0.019	−0.109***	0.272***	−0.121*	0.016
Per cent positive	EM	0.237***	0.5	0.275***	0.8***	0.35*	0.525
Mean	DRSGL	−0.074***	−0.004	−0.043**	0.205***	0.038	−0.125***
Median	DRSGL	−0.089***	−0.029	−0.063***	0.142***	−0.0001	−0.139***
Per cent positive	DRSGL	0.132***	0.395	0.175***	0.85***	0.5	0.1***
Mean	DLLP	0.021	0.0007	0.143**	−0.122**	0.178**	−0.179**
Median	DLLP	0.051	0.002	0.033**	−0.082**	0.138**	−0.176**
Per cent positive	DLLP	0.632	0.526	0.575	0.4	0.675**	0.325**
Panel B: Public Sector Commercial Banks
No. of observations		18	18	18	18	18	18
Mean	EM	−0.173***	0.028	−0.268**	0.559***	−0.132	0.286
Median	EM	−0.151***	−0.058	−0.216***	0.507***	−0.142	0.443**
Per cent positive	EM	0.111***	0.500	0.167***	1***	0.444	0.778**
Mean	DRSGL	−0.087***	−0.013	−0.031	0.236***	0.077	−0.095*
Median	DRSGL	−0.101***	−0.068*	−0.061**	0.197***	0.029*	−0.116***
Per cent positive	DRSGL	0.111***	0.222**	0.167***	1***	0.722*	0.167***
Mean	DLLP	0.086**	−0.041	0.238**	−0.324***	0.209*	−0.381**
Median	DLLP	0.083*	0.014	0.154**	−0.356***	0.164	−0.586**
Per cent positive	DLLP	0.667	0.556	0.722*	0.056***	0.611	0.222**
Panel C: Private Sector Commercial Banks
No. of observations		20	20	22	22	22	22
Mean	EM	−0.024	−0.033	−0.120	0.137	−0.148	−0.136
Median	EM	−0.062	−0.020	−0.092*	0.061	−0.109*	−0.135**
Per cent positive	EM	0.350	0.500	0.364	0.636	0.273**	0.318*
Mean	DRSGL	−0.062***	0.005	−0.054**	0.179***	0.005	−0.15***
Median	DRSGL	−0.068***	0.001	−0.064***	0.109***	−0.052	−0.143***
Per cent positive	DRSGL	0.15***	0.550	0.182***	0.727**	0.318*	0.045***
Mean	DLLP	−0.039	0.038	0.066	0.043	0.153	−0.014
Median	DLLP	0.047	−0.001	−0.019	0.044	0.127**	−0.042
Per cent positive	DLLP	0.600	0.500	0.455	0.682*	0.727**	0.409

Source: The authors.

Notes: Mean, median and per cent positive significance level have been checked through two-tailed t-test, Wilcoxon sign test and Proportion test respectively.

*, ** and *** represent significance level at 10%, 5% and 1%, respectively.

Table 5.

Classification of EM on the Basis of Quartiles

	Mean	Directions of EM
Quartile I	–0.529***	Negative EM
Quartile II	–0.134***	Close to
Quartile III	0.058***	Zero
Quartile IV	0.577***	Positive EM

Source: The authors.

Notes: One-tailed t-test has been used to check the significance level of mean.

*** represents the significance level at 1%.

Table 5 present the classification of EM on the basis of quartile division. To erect quartiles, all the observations are sorted in increasing order and then scaled by four. First 59 observations are regarded as first quartile, next 118 observations are taken as second and third quartile, last 59 observations come under the purview of fourth quartile. One-tailed t-test shows that all the quartiles have significant mean value.

Quartile I have negative mean value (−0.529%) indicating that the banks that came under this quartile are indulged in income decreasing EM. Quartile IV is just opposite to first quartile with a mean value of 0.577% reflecting that the banks take place in this quartile practice income increasing EM. On the other hand, second and third quartile includes values that are close to zero, however, second quartile has negative mean value (−0.134%) and third quartile has positive mean value (0.058%). The close to zero mean values reflects moderate level of EM.

Conclusion

The paper examines the quality of reported earnings of Indian public and private sector banks by using bank-specific model, that is, Beatty et al. (2002) during a time period of six years (2014–2019). The findings suggest that Indian banks are actively engaged in EM practices by applying their discretion in LLP and in realization of security gains and losses. Public banks indulge in income increasing EM (0.05%) while private banks practice income decreasing EM (−0.055%). To understand it more precisely, quality of earnings is studied year-wise. The results exhibit that the direction of EM keeps fluctuating, in some years there is income increasing EM and, in some years, there is income decreasing EM. One possible reason might be that managers manage earnings upward or downward as the situation demands or as per the business cycle. Another reason may be the reversal nature of accruals, implying that earlier when impaired loans are not recognized then DLLP is less than needed resulting in positive EM, later on when these impaired loans are classified as NPA, the bulky loss provisions result in negative EM. Further, first and fourth EM quartiles show negative and positive EM respectively and second and third EM quartiles exhibit close to zero values.

The findings of the study provide deeper insights to regulators, investors and depositors. Regulators should make the policies regarding LLP more specific and stricter and keep a close watch on high-value loans also. Investors and depositors should not regard earnings as full-proof security to their investment and deposits as earnings may be manipulated. They should consider other measures like capital adequacy ratio which provide cushion to banks to absorb unexpected losses.

The present study is confined to public and private sector commercial banks over a time period from 2014 to 2019. Future researchers may increase sample size by including foreign banks, cooperative banks and increase time period simultaneously. Second, the current research measures real activity EM through DRSGL. Researcher can also go for another measure of real activities EM over which managers have discretion such as commission and fee income.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

References

Ahmed

A. S.

, Takeda

, & Thomas

(1999). Bank loan loss provisions: A reexamination of capital management, earnings management and signaling effects. Journal of Accounting and Economics, 28(1), 1–25.

Alhadab

M. M.

, & Al-Own

(2017). Earnings management and banks performance: Evidence from Europe. International Journal of Academic Research in Accounting, Finance and Management Sciences, 7(4), 134–145.

Anandarajan

, Hasan

, & Lozano-Vivas

(2003). The role of loan loss provisions earnings management, capital management, and signaling: The Spanish experience. Advances in International Accounting, 16(3), 45–65.

Anandarajan

, Hasan

, & McCarthy

(2007). Use of loan loss provisions for capital, earnings management and signalling by Australian banks. Accounting & Finance, 47(3), 357–379.

Ashfaq

, & Saeed

M. A.

(2017). Financial performance of banking industry of Pakistan: The role of corporate governance index and earnings management practices. Asian Journal of Scientific Research, 10(2), 97–103.

Beatty

, & Harris

D. G.

(1998). The effects of taxes, agency costs and information asymmetry on earnings management: A comparison of public and private firms. Review of Accounting Studies, 4(3–4), 299–326.

Beatty

, Chamberlain

S. L.

, & Magliolo

(1995). Managing financial reports of commercial banks: The influence of taxes, regulatory capital, and earnings. Journal of Accounting Research, 33(2), 231–261.

Beatty

A. L.

, Ke

, & Petroni

K. R.

(2002). Earnings management to avoid earnings declines across publicly and privately held banks. Accounting Review, 77(3), 547–570.

Bratten

, Causholli

, & Myers

L. A.

(2017). Fair value exposure, auditor specialization, and banks’ discretionary use of the loan loss provision. Journal of Accounting, Auditing & Finance, 35(2), 318–348.

10.

Cheng

(2012). Managing specific accruals vs. structuring transactions: Evidence from banking industry. Advances in Accounting, 28(1), 22–37.

11.

Chipalkatti

, & Rishi

(2007). Do Indian banks understate their bad loans? The Journal of Developing Areas, 40(2) 75–91.

12.

Cohen

, Cornett

M. M.

, Marcus

A. J.

, & Tehranian

(2014). Bank earnings management and tail risk during the financial crisis bank earnings management and tail risk during the financial crisis. Journal of Money, Credit and Banking, 46(1), 171–197.

13.

Collins

J. H.

, Shackelford

D. A.

, & Wahlen

J. M.

(1995). Bank differences in the coordination of regulatory capital, earnings, and taxes. Journal of Accounting Research, 33(2), 263–291.

14.

Cornett

M. M.

, McNutt

J. J.

, & Tehranian

(2009). Corporate governance and earnings management at large US bank holding companies. Journal of Corporate Finance, 15(4), 412–430.

15.

Curcio

, & Hasan

(2015). Earnings and capital management and signaling: The use of loan-loss provisions by European banks. European Journal of Finance, 21(1), 26–50.

16.

DeAngelo

L. E.

(1986). Accounting numbers as market valuation substitutes: A study of management buyouts of public stockholders. The Accounting Review, 61(3), 400–420.

17.

Dechow

P. M.

, Sloan

, & Sweeney

A. P.

(1995). Detecting earnings management. The Accounting Review, 70(2), 193–225.

18.

Evans

, Houston

, Peters

, & Pratt

(2012). The Financial Reporting Environment, Reporting Discretion, and Earnings Management: US GAAP vs. IFRS [Working Paper]. Indiana University. http://citeseerx.ist.psu.edu/viewdoc/download;jsessionid=E9A218741CA43E7B0BC7354BE19C2C06?doi=10.1.1.466.9790&rep=rep1&type=pdf

19.

Ghosh

(2007). Loan loss provisions, earnings, capital management and signalling: Evidence from Indian banks. Global Economic Review, 36(2), 121–136.

20.

Grougiou

, Leventis

, Dedoulis

, & Owusu-Ansah

(2014). Corporate social responsibility and earnings management in US banks. Accounting Forum, 38(3), 155–169.

21.

Gujarati

N. D.

, Porter

C. D.

, & Gunasekar

(2018). Basic Econometrics (5th ed.). McGraw Hill Education Private Limited.

22.

Healy

P. M.

(1985). The effect of bonus schemes on accounting decisions. Journal of Accounting and Economics, 7(1–3), 85–107.

23.

India Brand Equity Foundation (2020). Indian financial services industry report. https://www.ibef.org/industry/financial-services-india.aspx

24.

Jones

J. J.

(1991). Earnings management during import relief investigations. Journal of Accounting Research, 29(2), 193–228.

25.

Kanagaretnam

, Lobo

G. J.

, & Mathieu

(2003). Managerial incentives for income smoothing through bank loan loss provisions. Review of Quantitative Finance and Accounting, 20(1), 63–80.

26.

Kanagaretnam

, Lobo

G. J.

, & Mathieu

(2004). Earnings management to reduce earnings variability: Evidence from bank loan loss provisions. Review of Accounting and Finance, 3(1), 128–148.

27.

Kumari

, & Pattanayak

J. K.

(2017). Linking earnings management practices and corporate governance system with the firms’ financial performance: A study of Indian commercial banks. Journal of Financial Crime, 24(2), 223–241.

28.

Leventis

, & Dimitropoulos

(2012). The role of corporate governance in earnings management: Experience from US banks. Journal of Applied Accounting Research, 13(2), 161–177.

29.

Mukherjee

(2019, September 20). Yes bank’s race against crisis puts India on warning. Business Standard. https://www.business-standard.com/article/finance/bank-s-race-against-crisis-puts-india-on-warning-119091700118_1.html

30.

Myers

, & Skinner

(2000). Earnings momentum and earnings management [Working Paper]. University of Michigan.

31.

Nayyar

(2019, October 24). As safe as the banks? Business Standard. https://www.business-standard.com/article/opinion/as-safe-as-the-banks-119102301470_1.html

32.

Osma

B. G.

, & Noguer

B. G.

(2007). The effect of the board composition and its monitoring committees on earnings management: Evidence from Spain. Corporate Governance: An International Review, 15(6), 1413–1428.

33.

Ozili

P. K.

(2017). Discretionary provisioning practices among Western European banks. Journal of Financial Economic Policy, 9(1), 109–118.

34.

Ozili

P. K.

, & Outa

E. R.

(2018). Bank income smoothing in South Africa: Role of ownership, IFRS and economic fluctuation. International Journal of Emerging Markets, 13(5), 1372–1394.

35.

Peasnell

K. V.

, Pope

P. F.

, & Young

(2000). Detecting earnings management using cross-sectional abnormal accruals models. Accounting and Business Research, 30(4), 313–326.

36.

Sarkar

, Sarkar

, & Sen

(2008). Board of directors and opportunistic earnings management: Evidence from India. Journal of Accounting, Auditing & Finance, 23(4), 4517–551.

37.

Sun

, & Rath

(2010). Earnings management research: A review of contemporary research methods. Global Review of Accounting and Finance, 1(1), 121–135.

38.

Ujah

N. U.

, Brusa

, & Okafor

C. E.

(2017). The influence of earnings management and bank structure on bank performance: International evidence. Managerial Finance, 43(7), 761–773.

39.

Umoren

, Ikpantan

, & Ededeh

(2018). Earnings management and financial performance of deposit money banks in Nigeria. Research Journal of Finance and Accounting, 9(22), 94–100.

40.

Vishnani

, Agarwal

, Agarwalla

, & Gupta

(2019). Earnings management, capital management and signalling behaviour of Indian banks. Asia-Pacific Financial Markets, 26(3), 285–295.

41.

Y. C.

, Wei Kiong Ting

, Lu

W. M.

, Nourani

, & Kweh

Q. L.

(2016). The impact of earnings management on the performance of ASEAN banks. Economic Modelling, 53, 156–165.