EARNINGS MANAGEMENT VIA DEPRECIATION AND ITS IMPACT ON MARKET REACTION IN THE CASE OF LISTED COMPANIES IN VIETNAM STOCK MARKET45433

EARNINGS MANAGEMENT VIA DEPRECIATION AND ITS IMPACT ON MARKET REACTION IN THE CASE OF LISTED COMPANIES IN VIETNAM STOCK MARKET Dr.. This paper will examine for the appearance of earnin

EARNINGS MANAGEMENT VIA DEPRECIATION AND ITS IMPACT ON MARKET REACTION IN THE CASE OF LISTED COMPANIES IN

VIETNAM STOCK MARKET

Dr Tran Thi Kim Anh, Foreign Trade University Email: anhttk@ftu.edu.vn MSc Hoang Ha Anh Foreign Trade University Email: anhhh@ftu.edu.vn

Abstract

The quality of financial reports and its impacts on the economic decisions of external users of financial statement have been the remarkable topic for many academic researchers Earnings management addresses the possibility that managers might adjust the components of financial reports to accomplish personal goals and to deceive other parties, which can be done by shifting any accounting and methods techniques There are two types of earnings management: Accrual-based earnings management (AEM) and Real earnings management (REM) This paper mainly focuses on accrual-based earnings management and earnings management via depreciation of assets Accounting for depreciation is based on accruals basis which can be managed by changing depreciation methods or useful lives of assets This paper will examine for the appearance of earnings management via depreciation and the relationship between abnormal depreciation and its impact on market reaction in the case of listed companies in Vietnam stock market from 2013 to 2016 The result indicates that earnings management via depreciation appears in listed companies in Vietnam stock market and the relationship between abnormal depreciation and its impact on market reaction in the case of Vietnam is significantly positive Since then, the authors suggest several ways for external users of financial statements to have another view about earnings management and its impacts, and also give recommendation for future researches

Key words: Depreciation, earnings management, market reaction, accrual

accounting

1 Introduction

Financial reporting quality has become a top prior concern since many accounting scandals have recently been detected In the world, it can be mentioned about the case of WorldCom, Enron, Olympus, Health South, Tyco International, Parmalat, and Toshiba In Vietnam, several companies have been pointed out to make accounting fraud such as Go Truong Thanh JSC, Hung Vuong JSC, Vien Dong Pharmaceutical JSC… These scandals have strongly influenced Vietnam stock market and raise an alert for every investor Even though the financial reports are audited by independence auditor every year, accounting frauds are still unable to be perceived earlier

Many researchers have studied this topic and found out many means for companies to manage their financial reports Earnings management addresses the

possibility that managers might adjust the components of financial reports to accomplish personal goals and to deceive other parties, which can be done by shifting any accounting and methods techniques There are two types of earnings management: Accrual-based earnings management (AEM) and Real earnings management (REM) AEM occurs because generally accepted accounting principles allow managers to

be flexible to choose accounting methods, accounting policies and accounting estimates (Healy and Wahlen, 1999) For example, provision and allowances for bad debts, assets revaluation, depreciation of assets…are some events that are affected by accounting estimates If these estimates are biased in order to distort the underlying real economic performance, AEM has been applied (Joosten, 2012) Meanwhile, REM happens when firms manage earnings through deviating from the normal business activities (Roychowdhury, 2006) Firms could deviate from business activities by, for example, altering expenditures, such as research and development expenditures, selling expenditures, administrative expenditures, selling fixed assets to receive a gain, changing cash discounts, trade discounts…

This paper mainly focuses on accrual-based earnings management and earnings management via depreciation of assets Accounting for depreciation is based on accruals basis which can be managed by changing depreciation methods or useful lives

of assets This study will examine the sample data of 34 listed companies in Vietnam stock market from 2013 to 2016 to find out the appearance of earnings management via depreciation and the relationship between abnormal depreciation and its impact on market reaction

2 Literature review

2.1 Earnings management and its motivation

Earnings management is a widely accepted term among academic literature It

is stated by Healy and Wahlen (1999) as “earnings management occurs when managers use their own judgments in financial reporting and in structuring transactions to alter financial reports to either mislead stakeholders about the underlying economic performance of the company” Similar to Healy and Wahlen

(1999), Dechow et al (2000) has identified the practice and method of earnings

management Earnings management addresses the possibility that managers might adjust the components of financial reports to accomplish personal goals and to deceive other parties, which can be done by shifting any accounting and methods techniques Meanwhile, earnings manipulation is “intentional misstatements or omissions of amounts and disclosures in financial statements to deceive users” (Arens and Loebbecke, 2000) Due to the similar in definition, sometimes the terms “earnings management” and “earnings manipulation” are used interchangeably However, earnings management is not the same as earnings manipulation The sharp line between them, according to Beneish (1999), is the violation of accounting standards If management uses their discretions which do not violate the GAAP or IFRS then it is called earnings management Otherwise, if managers violate the GAAP or IFRS then it

is called earnings manipulation or fraud accounting

There are two types of earnings management: Accrual-based earnings management (AEM) and Real earnings management (REM) This study mainly focuses on accrual-based earnings management Accrual accounting requires to records revenues and expenses when they are incurred, regardless of when cash is exchanged Therefore, accrual accounting creates opportunities for accountants and

managers to manage earnings based on non-cash transactions such as receivables, payables, depreciation, provision, and so forth According to Ayres (1994) there are three ways of managing earnings The first way is to modify earnings by varying items such as the probability to recover debts as provision and contingency The second way

is to alter the timing of obligatory accounting policies such as useful lives of fixed assets The third way is to change from one accounting method to another such as methods of calculating inventory, methods of calculating depreciation…

Whichever the ways to manage earnings are, the motivations for earnings management and manipulation are summed up for several reasons Many researchers has studied and empirically examined the reasons for managers to make up the accounting numbers The main reasons for managing earnings are capital market motivation, manager incentives, and reducing in income taxes

Firstly, capital market motivations happen when companies want to increase their stock price and trading volume, and especially, when announcing the initial public offerings, companies might manage to increase earnings to demonstrate their great economic performance Capital market motivations for earnings management has

been examined by many researchers such as Healy and Wahlen (1999), DuCharme et

al (2001) in the case of USA, Chou et al (2006), Kao et al (2009) in the case of China,

Ahmad-Zaluki et al (2009) in the case of Malaysia

Secondly, managers would act for earnings management if it helped increasing their reward and incentives (Guidry et al, 1999; Cheng and Warfield, 2005; Bergstresser and Philippon, 2006)

Thirdly, to delay paying tax, to reduce tax liability, or to gain tax incentives is one of the conspicuous reasons for managing earnings Income tax motivation for

earnings management has been proved by some researches such as Boynton et al

(1992) and Phillips et al (2001) in the case of the USA, Tang and Firth (2011) in the case of China

2.2 Earnings management via depreciation and its impact on market reaction

Depreciation is one of the accrual variables which has wisely impacted on company’s expense and profit numbers Since Vietnam Accounting Standard required all companies to apply accrual basis of accounting, depreciation becomes a variable in the financial statement that could be distorted for the purposes of managers to manage

or manipulate earnings It is flexible for firm to calculate depreciation so there is a window for managers to dress the accounts Managing earnings via depreciation is achieved in two ways The first way is to switch depreciation methods and the second way is to change estimation of assets’ useful life (Hillier and McCrae, 1998)

Myers (1967) is the very first paper studied about managing earnings through depreciation methods The paper investigated the relationship between any change in depreciation method and the firms’ earnings per share Later, Myers (1969) observed

20 enterprise and 24 enterprises that have changed depreciation from accelerated to straight-line method in 1967 and 1968 respectively Myers (1969) concluded that every change in depreciation lead to the increase in firms’ earning per share The impact of changes in depreciation method on capital market reaction has been examined by some researchers Archibald (1972) tested for the impact of changes in depreciation method on capital market reaction for 69 companies and did not find abnormal performance of the stock during the announcement of the change of depreciation method On the contrary with Archibald (1972), Teoh, Wong and Rao

(1998) examined depreciation estimates surrounding initial public offers They found that firms were more likely to switch to income-increasing depreciation policies in the

IPO year and for several subsequent years Peasnell et al (2000) stated that earnings

management through depreciation manipulation was “a somewhat transparent”, and

thus it made impacts to the market price of the firm Jackson et al (2009) found out

that firms that make such depreciation accounting changes make smaller capital investments in the post-change periods than in the pre-change periods The study concluded that firms’ depreciation method choice is likely to influence managers’ capital investment decisions

Depreciation management has been considered as a variable for earnings management on accruals basis in academic literature Numerous econometric models which is used to detect accruals earnings management has included depreciation as the main variables such as M-score model of Beneish (1999), Jones model (1991),

Modified Jones model by Dechow et al (1995), Kasnizh model (1999)

The important role of depreciation in managing earnings has been tested and confirmed by many researches However, the researches for earnings management via depreciation in particular are not plentiful This paper will examine earnings management via depreciation and its impact on market reaction in the case of listed companies in Vietnam stock market to fill in the gap of lacking of researches about depreciation management in particular and to flourish the researches for earnings management in general in Vietnam

3 Research methodology

3.1 Research method

This study is going to use quantitative research method Quantitative method is used as a synonym for any data collection technique and data analysis procedure that

generate numerical data (Saunders et al., 2012) Therefore, this study is going to

collect numerical data from reliable and reputational sources and then use highly structured quantitative testing to examine the data set in order to prove the hypothesis

3.2 Data collection method

As the results of using quantitative method, the study mainly relies on secondary data collected using computational techniques The use of secondary data has some drawbacks relevant to data availability In this study, all of the data is collected from financial reports of firms which are listed in Vietnam stock market The sample in this study contains data of 34 companies from 2013-2016 These 34 companies are chosen randomly in Vietnam stock market due to the available of data The total number of observations in the sample data is 102

3.3 Development of hypothesis

In the scope of this paper, it aims to examine for the appearance of earnings management via depreciation and its impact on market reaction through abnormal stock return for listed companies in Vietnam stock market Therefore, the hypothesis is:

- Earnings management via depreciation: Depreciation has been a frequent variable used in many econometric models to discover earnings management such as Jones model (1991), Modified Jones model (1995), M-score of Beneish (1999) It also

mentioned in many academic literature such as Dechow et al (1995), Dechow et al (2000), Chou et al (2006), Kao et al (2009), Jamal and Murray (2013)… Therefore,

the hypothesis for managing earnings via depreciation in listed companies in Vietnam stock market is:

H 1 : Listed companies in Vietnam stock market tend to use abnormal depreciation

to manage earnings

- The impact of abnormal depreciation on market reaction: There were several researches about the impact of abnormal depreciation on market reaction as mentioned

in literature review However, the author is unable to find any previous research about this topic in the case of Vietnam Therefore, to fill in the gap among literature about earnings management in Vietnam, this study will examine the following hypothesis:

H 2 : Abnormal depreciation has positively impacted on the abnormal stock return

3.4 Variables explanation

This study uses two different proxies to identify abnormal depreciation The first abnormal depreciation variable (DEPI_1) is designed following Marquardt and Wiedman (2004) paper It is based on the calculation of the expected value of depreciation, which is assumed to remain a constant in proportion with gross property, plant, and equipment variable (Gross PPE) The abnormal depreciation is the difference between expected depreciation and the real value of depreciation DEPI_1 measures the abnormal depreciation as proportion of total assets

DEPI_1 =

Whereas:

DEPI_1: abnormal depreciation 1

DEP: net depreciation

Gross PPE: gross property, plant, and equipment

TA: total assets

T: year

The second abnormal depreciation variable (DEPI_2) is calculated by using M-score model of Beneish (1999) The author indicates that a depreciation index greater than 1 is an indicator of the slowdown of the rate by which assets have been depreciated It happens because of two reasons: firm has switched its depreciation method in order to increase earnings or estimation of asset’s useful life has been raised

DEPI_2 =

Whereas:

DEPI_2: abnormal depreciation index 2

DEP: net depreciation

Net PPE: net property, plant, and equipment

T: year

This study aims to test market reaction change when depreciation is managed The determinants of market reaction could be stated by several factors such as abnormal earnings, abnormal sales, abnormal depreciation, abnormal dividend announcement… (Bajaj and Vijh, 1995; Palmrose et al, 2001).Therefore, this study

uses ordinary least square (OLS) method to find out the relationship between market reaction and depreciation index as stated below:

CAR = β0 + β1 x DEPI_1 + β2 x UE + β3 x US + β4 x SIZE (1)

CAR = β0 + β1 x DEPI_2 + β2 x UE + β3 x US + β4 x SIZE (2)

The dependent variable CAR is measured by calculating the difference between a portfolio's performance and the market return over a set period of time This study uses change in VN-Index as the benchmark for determining market stock return For example, if the average stock price increased by 5% and the average market increased

by 3%, then the abnormal return was 2% (5% - 3% = 2%) The independent variables

UE (unexpected earnings) and US (unexpected sales) are calculated by taking the difference between real earnings/sales and expected earnings/sales The independent variables SIZE represents that size of the company which is related to abnormal return (Cook and Rozeff, 1984) This study uses natural logarithm of total assets to embody the size of the company

3.5 Econometric model

In accordance with the sample data, this study will use technique for panel data

to analyze the data set There are three techniques to analyze panel data: fixed effects, random effects, and pooled OLS regression To choose the most appropriate technique

to apply in this data set, two pre-requisite tests will be conducted Firstly, Breusch-Pagan Lagrange multiplier test will be conducted to choose between pooled OLS regression and Random Effects If Random Effects technique is chosen after Breusch-Pagan Lagrange multiplier test, Hausman test will be conducted to choose between Fixed Effects and Random Effects The results show that pooled OLS regression model would be the more appropriate model for this case (result will be attached in the next section)

The pooled OLS regression model aims to explain every value of independent variable Y is associated with a value of dependent variable X The model for pooled OLS regressions is:

Yit = β1X1it + β2X2it +……… + βnXnit+ αi+ uit

Whereas

Yit is the dependent variable where i = entity and t = time

Xit is independent variable

αi (i=1….n) is the unknown intercept for each entity

β is the coefficient

uit is the error term

4 Findings

STATA software is used to run panel data model

Table 1: Descriptive statistic

Variable Mean Std Dev Min Max Observations CAR 17.56843 45.67969 -63.8 154.2 N = 102

n = 34

T = 3

DEPI_1 -.0065158 1280644 -.2994086 8330272 N = 102

n = 34

T = 3

DEPI_2 955168 3840101 1775302 2.483437 N = 102

n = 34

T = 3

UE 29498.58 184665.3 -780392 1097830 N = 102

n = 34

T = 3

US 251760.7 2142207 -5785549 1.81e+07 N = 102

n = 34

T = 3

SIZE 13.95122 1.348336 11.65979 17.19578 N = 102

n = 34

T = 3

From the descriptive data table, both DEPI_1 and DEPI_2 appeared in the case

of listed companies in Vietnam stock market Therefore, the hypothesis H1 “Listed

companies in Vietnam stock market tend to use abnormal depreciation to manage earnings” is accepted

DEPI_1 ranges from -0.29 to 0.83 with the mean of -0.006 DEPI_1 expressed the unexpected depreciation as proportion of the firm’s total assets The range from -0.29 to 0.83 means abnormal depreciation is calculated as from -29% to 83% of firms’ total assets which is very considerable It is necessary for investors to pay attention to these numbers because total assets of the firms included in this study have been managed significantly through depreciation

DEPI_2 ranges from 0.17 to 2.48 with the mean of 0.95 for the 34 listed companies in Vietnam stock market According to Beneish (1999), a depreciation index greater than 1 is an indicator of the slowdown of the rate by which assets have been depreciated It happens because of two reasons: firm has switched its depreciation method in order to increase earnings or estimation of asset’s useful life has been raised Therefore, for the firms that have depreciation index greater than 1, it

is essential for investors to take into account the abnormal depreciation to consider any firms’ earnings management via depreciation that might take place to mislead outside investors

After finding out the results for hypothesis 1, here is the result for the hypothesis

2 This study aims to explore the relationship between abnormal stock return and abnormal depreciation expressed through DEPI_1 and DEPI_2

Before running the main test, the sample data set needed overcoming other

pre-tests As such, correlation test and multicollinearity test will be presented

Table 3: Correlation test

| car depi_1 depi_2 size ue us

-

car | 1.0000

depi_1 | 0.2870 1.0000

depi_2 | 0.2885 0.7311 1.0000

size | -0.1035 -0.0740 -0.0521 1.0000

ue | 0.2480 0.0243 0.0664 0.2471 1.0000

us | -0.0615 0.0035 0.0034 0.2626 -0.1187 1.0000

Correlation test determines how strongly two variable's movements are associated Correlation coefficients are expressed as a value between +1 and −1, where

1 is total positive linear correlation, 0 is no linear correlation, and −1 is total negative linear correlation As the result in table 3, all of the variable in the sample data set weakly correlate with each other It is a decent indication that allows us to run the main test without worrying about correlation problem among variables

Table 4: Multicollinearity test

Variable | VIF 1/VIF

-

depi_1 | 1.01 0.991498

size | 1.18 0.845788

us | 1.12 0.894165

ue | 1.11 0.900530

-

Mean VIF | 1.10

Variable | VIF 1/VIF - depi_2 | 1.01 0.989395 size | 1.18 0.847309

us | 1.01 0.989395

ue | 1.12 0.895905 -

Mean VIF | 1.11

Multicollinearity test is to determine whether one predictor variable can be linearly predicted from the others with a substantial degree of accuracy Multicollinearity problem is quantified by the variance inflation factor (VIF) in an ordinary least squares regression analysis If VIF is more than 10, it indicates high collinearity problem in the data set As the results in table 4, VIF for all variables included in the two models is closely to 1 which indicates that there is no

multicollinearity problem in the sample data set It is a good signal that will lead to run the main test without worrying about collinearity problem among variables

Here are the main results of the study Firstly, the result relationship between abnormal stock return and abnormal depreciation expressed through DEPI_1 is represented

Breusch and Pagan Lagrangian test are represented to choose between pooled OLS regression and RE technique If the P-value is less than 0.05, RE is in favor

Table 5: Breusch and Pagan Lagrangian multiplier test for random effects

car[company,t] = Xb + u[company] + e[company,t]

Estimated results:

| Var sd = sqrt(Var)

-+ -

car | 2086.634 45.67969

e | 1848.181 42.99047

u | 0 0

Test: Var(u) = 0

chibar2(01) = 0.00

Prob > chibar2 = 1.0000

From the above results, pooled OLS model is chosen to analyze the relationship between CAR and DEPI_1

Table 6: Pooled OLS results

F( 4, 97) = 4.70 Prob > F = 0.0017 R-squared = 0.3623 Adj R-squared = 0.2278 Root MSE = 42.662

Model 34206.4344 4 8551.6086

Residual 176543.579 97 1820.0369

Total 210750.014 101 2086.6338

CAR Coef Std Err t P>|t| [95% Conf Interval]

DEPI_1 95.77605 33.28933 2.88 0.005 29.70595 161.8462

UE 0000696 0000242 2.87 0.005 0000216 0001177

US 2.56e-07 2.10e-06 0.12 0.903 -3.90e-06 4.42e-06

SIZE -5.297755 3.423345 -1.55 0.125 -12.09215 1.496636

cons 89.98389 47.56702 1.89 0.062 -4.423467 184.3912 Joint test on regression indicated F statistic = 0.0017 which is less than 1% significant level It indicates that the null hypothesis: “There is no significant relationship between dependent and independent variables” is rejected Therefore, there is significant relationship between dependent and independent variables so that our model is highly appropriate and the result is absolutely reliable

R is the multiple correlation of determination which represents the total correlation between all the predictors and the dependent variable while R-squared represents the total amount of variance accounted for the dependent variable by the predictors (Miles and Shevlin, 2001) In this case, it can be concluded that the predictors (DEPI_1, UE, US, SIZE) explain 36.23% of the variance in dependent variable CAR

The result from STATA shows that p-value for DEPI_1 is 0.005, for UE is 0.005, for US is 0.903, and for SIZE is 0.125 Therefore, abnormal depreciation DEPI_1 and unexpected earnings UE have significantly impacted on the abnormal stock return as their p-value is less than 5% The model then has significant meaning in econometrics The coefficient between DEPI_1 and CAR is 95.7765 As a result, the relationship between abnormal depreciation and abnormal stock return is positive Each unit increasing in abnormal depreciation causes 95.7765 units increasing in abnormal stock return It expresses the highly fluctuating in stock price that might happen due to the earnings management via depreciation in listed companies in Vietnam stock market

Secondly, here is the result for the relationship between abnormal stock return and abnormal depreciation expressed through DEPI_2

Table 7: Breusch and Pagan Lagrangian multiplier test for random effects

car[company,t] = Xb + u[company] + e[company,t]

Estimated results:

| Var sd = sqrt(Var)

-+ -

car | 2086.634 45.67969