# Discretionary Accruals, Market Capitalization, and Risk

### Text-only Preview

ISSN: 1450-2889 Issue 5 (2009)

© EuroJournals Publishing, Inc. 2009

http://www.eurojournals.com/MEFE.htm

**Discretionary Accruals, Market Capitalization, and Risk**

**Mohammad Fawzi Shubita**

*Assistant Professor, School of Management*

*New York Institute of Technology, Amman – Jordan*

**Moade Fawzi Shubita**

*Corresponding Author, Associate Professor, Regional Associate Dean*

*School of Management: Middle East*

*New York Institute of Technology*

*PO Box 13893 Code 11942. Amman – Jordan*

E-mail: [email protected]; [email protected]

**Abstract**

Managers can use their discretionary power in the financial reporting process and in

structuring transactions. Managers convey private information to stakeholders about the

underlying economic performance of the company or attempt to influence contractual

outcomes that depend on the reported accounting income assuming market efficiency.

This study aims at achieving the following objectives: Introducing evidence about

the information content of discretionary accruals from the Jordanian market, enriching the

literature on discretionary accruals relationships in the capital market research and

examining the explanatory power of the stock price to future profitability. The study

sample consists of (44) industrial companies listed in the Amman Stock Exchange (Amman

Bourse), during the period (2000-2007).

Regression analysis is employed to examine the study's hypotheses (five models).

(Adjusted-R2) was used to indicate the incremental information content for the study

variables. The study found the following: - Stock prices can't predict future profitability in

Jordanian companies; current profitability has incremental information relative to stock

prices when predicting future profitability; discretionary accruals have incremental

information content to current profitability and stock prices when predicting future

profitability; discretionary accruals are associated with the level of systematic risks and are

associated with the level of insolvency risks. Thus, the level of discretionary accruals

should have impact on the market valuation of company risk (naive investor hypothesis).

**Keywords:**Discretionary accruals, Future profitability, Information Content, Stock prices,

and Systematic risk.

**Jel Classification Codes:**G150, G320, N250, G32 and M490

**Introduction**

The subject of discretionary accruals in accounting literature grew in popularity during the last decade

of the 20th century. Numerous studies investigated theoretically and empirically different hypotheses

related to discretionary accruals. Some researchers aimed simply to provide evidence of earnings

51

*Middle Eastern Finance and Economics - Issue 5 (2009)*

management; others investigated its role in agency problems, financial markets and information

asymmetry (Babalyan, 2004).

Clearly, high earnings are not as important as high-quality earnings: those that are repeatable,

controllable and bankable. Earnings that experience a surge because of a one-time, uncontrollable

event are not earnings that are inherent to the activities of the business. These earnings are the result of

luck, which is never a reason to invest. Finally, those businesses that generate revenue but not cash are

not engaging in profitable activities (McClure, 2006).

We expect shareholders to have incentives at all times, either to support income-increasing or

income-decreasing earnings management. Highly concentrated shareholders have controlling power,

therefore, have greater power to influence the decision pertaining to either income-increasing or

decreasing earnings management (Varma, et al. 2009). Misstated financial results can mask underlying

trends in revenue and earnings growth. Thus, overstatements of revenues and earnings are likely to

distort expectations of growth by those unaware of the misstatement (McNichols and Stephen, 2008).

The expected trend of future earnings determine the direction of employing the discretionary

accruals; if managers believe that future earnings of the firm will be low, they use discretionary

accruals to manage current earnings upwards (Yasuda, et al. 2004).

Correspondingly, managers use discretional accruals to manage current earnings downwards, if

they believe that future earnings will be high. Earnings management is also forward-looking, because

there may be several subsequent years of low/high earnings, and the managers need to consider the

level of earnings of future years when making the earnings management decisions (Jones, 1991).

Companies' managers may misrepresent the reported earnings to investors in order to make

their financial positions look better by controlling the discretionary portion of accruals. However,

"rational investors would correctly anticipate this manipulation by the companies' managers; thus, the

level of discretionary accruals should have no impact on the market valuation of companies' risk

(rational investor hypothesis)" (Yasuda et al. 2004).

Conversely, naive investors would misinterpret high reported earnings as being favorable

information about company health, and undervalue companies' risk1. If so, "the discretionary accruals

should be negatively associated with the level of company risk (naive investor hypothesis)" (Yasuda, et

al. 2004).

To distinguish the two hypotheses, the relationship between discretionary accruals and

company risk will be investigated. This research studies whether the market rationally or naively

interprets risk-irrelevant information contained in the discretionary accruals.

**Research Problem**

To study the relationship between discretionary accruals on one hand and profitability and risk in the

other, the research problem can be expressed in the following questions:-

1. Do discretionary accruals have incremental information relative to current profitability and

stock prices when predicting future profitability?

2. Do discretionary accruals associate with the level of systematic risks?

3. Do discretionary accruals associate with the level of insolvency risks?

**Research Objectives**

This study aims to achieve the following objectives:

1. Introducing evidence about the information content of discretionary accruals from the

Jordanian market.

1 This hypothesis does not fit with Jordan environment because naïve investors in Jordan do not read financial statements.

*Middle Eastern Finance and Economics - Issue 5 (2009)*52

2. Enriching the literature on discretionary accruals relationships in the capital market

research.

3. Examining the explanatory power of the stock price to future profitability.

4. Showing the incremental information content of discretionary accruals relative to stock

prices when predicting future profitability.

5. Reporting the relationship between discretionary accruals and systematic risks.

6. Investigating the relationship between discretionary accruals and insolvency risk.

**Previous Studies**

The issue of discretionary accruals has been discussed in academic or the literature for a long period.

Subramanyam (1996) and Ali et al. (2000) found that discretionary accruals enhanced the ability to

predict future earnings. Ali et al. (2000) aimed to explore whether the association between accruals and

future returns is due to naïve investors and found, by running some regression models on 86 USA

firms, that the ability of accruals to predict future profitability was not lower for large firms followed

by analysts, indicating that the ability of accruals to predict future profitability was not due to the

inability of market participants to understand value-relevant information.

Healy (1985) found that managers were more likely to choose income-decreasing accruals

when the upper or lower bounds of their bonus plans were binding, and income-increasing accruals

when these bounds were not binding. Jaggi and Lee (2002) found that managers used income

increasing discretionary accruals if they were able to obtain waivers from lenders for the violation of

debt covenants, but use income decreasing discretionary accruals if debt restructuring took place or

debts were renegotiated because waivers are denied.

There are several articles that study the relationship between distressed companies and

discretionary accruals and the relationship between companies surrounding risks and discretionary

accruals. Yasuda, et al. (2004) aimed to show empirically the relationship between bank risk and

discretionary accruals and showed using Jones (1991) model on 48 Japanese banks that bank risk is

negatively associated with discretionary accruals, indicating that investors misinterpreted high reported

earnings as favorable information about bank financial health. Peltier-Rivest (1999) selected 127

troubled firms in USA that reduce dividends due to operating risks and found that these firms adopt

income-decreasing accounting polices. Smith et al. (2001) categorized bankruptcy firms as failed and

non-failed. They predicted that managers of distressed firms that subsequently failed would adopt

accounting choices to reflect the underlying economic performance of the firm (on average income-

decreasing) due to the high ex-post settling costs. Bergman and Callen (1991), Neo and Wang (2000)

and Jaggi and Lee (2002) suggested that managers of a distressed firm during debt renegotiation were

focused on convincing creditors to extract concessions so that the firm can over-come its financial

difficulties that would derive benefits to both parties.

Another strand of the literature related to this research is on the determination of risk behavior

at companies. Saunders, et al. (1990) found that managerial ownership was positively associated with

firm risk and that this relationship was more pronounced during periods of deregulation. Anderson and

Fraser (2000) also found that firm risk was increasing in managerial ownership in periods of relative

deregulation, but that managerial ownership is negatively related to firm risk. Chen et al. (1998) also

found a negative relationship between managerial ownership and company risk. Demsetz and Strahan

(1997) found that larger companies are better diversified.

In many countries some industries such as banking, insurance and utility face regulations often

based on accounting measures. There is considerable evidence that banks close to capital adequacy

requirements overstate reserves or recognize book gains selling securities (see e.g. Moyer (1990),

Collins et al. (1995)). A number of studies examined the incentives to avoid anti-trust investigations, to

obtain government protection or subsidy by appearing less profitable. In her well-known paper, Jones

(1991) reported that firms in the industries seeking import relief tend to recognize negative abnormal

53

*Middle Eastern Finance and Economics - Issue 5 (2009)*

accruals to defer income in the year of application. Key (1997) examined accruals for USA firms in the

cable TV industry at the time when deregulation plans were being debated in Congress. The evidence

is again consistent with earnings understatement.

**Research Hypotheses**

To achieve the research objectives five hypotheses will be tested; the hypotheses will be expressed

using the null form as follows:-

**H01:**

Stock prices can't predict future profitability.

**H02:**

Current profitability doesn’t have incremental information relative to stock prices when

predicting future profitability.

**H03:**

Discretionary accruals don’t have incremental information relative to current

profitability and stock prices when predicting future profitability.

**H04:**

Discretionary accruals are not associated with the level of systematic risks.

**H05:**

Discretionary accruals are not associated with the level of insolvency risks.

**Research Models and Variables**

To investigate information content for discretionary accruals, the following pooled OLS regressions

are estimated (Yasuda, et al. 2004):

PROFit= β0+ β 1STCPRIit-1+ e1it

(Model 1)

Where:

PROFit = Future profitability of the firm i at year t which is measured by the ratio of net income

to Sales at year t.

STCPRIit-1 = the stock price multiplied by the number of shares outstanding (market value of

equity) in year t-1 dividing by lagged total assets.

β 0 and β 1 = Coefficients.

e1it = Error term.

The market value of equity is used instead of stock prices, because other variables are measured

at the firm level rather than in per share form (Kallunki and Martikainen, 2003). The market value of

equity is scaled by lagged total assets to reduce heteroscedasticity. The method used in the models is

dividing all the variables by the lagged total assets instead of adding the natural log of the book value

of total assets because the market value of equity is an absolute figure not a ratio figure. In this model,

the relationship between stock prices and future profitability will be investigated to examine the ability

of stock prices to predict future profitability. This will be done by executing a simple regression and

show the significance of STCPRIit-1 and Adjusted-R2 factors. This model is used to test the first

hypothesis.

To test the second hypothesis, "Current profitability doesn’t have incremental information

relative to stock prices when predicting future profitability,” Model (2) will be used.

PROFit= β'0+ β'1PROFit-1+ β'2STCPRIit-1+ e2it (Model

2)

Where

PROFit-1 = Current Profitability of the firm at year t-1 which is measured by the ratio of net

income to Sales at year t-1.

In this model, the incremental information of current profitability relative to stock prices when

predicting future profitability is investigated. This will be done by executing a multiple regression and

show the significance of PROFit-1, STCPRIit-1 and Adjusted-R2 factors, and compare Adjusted-R2

factors between models (1) and (2). This model is used to test the second hypothesis.

To test the third hypothesis, “Discretionary accruals don’t have incremental information

relative to current profitability and stock prices when predicting future profitability,” Model (3) will be

used.

*Middle Eastern Finance and Economics - Issue 5 (2009)*54

PROFit= γ'0+ γ'1PROFit-1+ γ'2STCPRIit-1+ γ'3 DISCit-1+ e3it (Model

3)

The discretionary portion of total accruals is used in this study rather than the discretionary

portion of a single accrual because total accruals should capture a larger portion of managers'

manipulation (Jones, 1991).

Total accruals (TA) are defined as the difference between net income (NI) and cash flow from

operating activities (OCF): -

TA= NI-OCF

The following expectation model will be used for total accruals to control for changes in the

economic circumstances of the firm (Jones, 1991):

TAit/Ait-1= a1 [1/ Ait-1] + a2 [Δ REVit/ Ait-1] + a3 [PPEit/ Ait-1] + ěit

Where:

TAit = total Accruals in year t for firm (difference between net income (NI) and cash flow from

operating activities);

Δ REVit = revenues in year t less revenues in year t-1 for firm i;

PPEit = gross property, plant and equipment in year t for firm i;

Ait-1 = total assets in year t-1 for firm i;

ěit = error term in year t for firm i2;

a1, a2 and a3 = Model Factors

I = 1,…, N

T = 1,…, ti, year index for the years included in the estimation period for firm i.

In this equation, change in revenues, and gross property, plant and equipment are included in

the expectation model to control for changes in nondiscretionary accruals caused by changing

conditions.

Total accruals (TA) include changes in working capital accounts, such as accounts receivable,

inventory and accounts payable that depend to some extent on changes in revenues. Revenues are used

as control for the economic environment of the firm because they are an objective measure of the firms'

operations before managers' manipulations.

Gross property, plant and equipment are included as control variable for the portion of total

accruals related to nondiscretionary depreciation expense. The residual of this equation will be

estimated the discretionary accruals in year t-1 (Jones, 1991).

In this model, the incremental information of discretionary accruals relative to current

profitability and stock prices when predicting future profitability is investigated. This will be done by

executing a multiple regression and show the significance of PROFit-1, STCPRIit-1, DISCit-1 and

Adjusted-R2 factors, and compare Adjusted-R2 factors between models (2) and (3). This model is used

to test the third hypothesis.

To test the fourth hypothesis, “Discretionary accruals are not associated with the level of

systematic risks,” Model (4) will be used.

SESRISKit= v0+ v1 DISCit-1 +v2ASSETit-1+ e4it (Model

4)

Where:

SESRISKit = systematic risk for firm i at month t.

MANAGit-1, ASSETt-1 = as defined in models (1) and (2).

Systematic risk from market model will be derived by computing beta that represents the

systematic risk from the following model using monthly data for the study period (1995-2004):-

Rit = ai + βi Rmt + eit

Where:-

Rit

=

Market Return for company i share at year t.

ai

= Constant

factor.

βi.

=

Stock beta for firm i.

2 An estimation of the discretionary accruals.

55

*Middle Eastern Finance and Economics - Issue 5 (2009)*

Rmt

= Market

return.

eit

= error

term.

This model investigated if discretionary accruals associated with the level of systematic risk.

This model is used to test the fourth hypothesis that is described in the international studies as "Naive

Investor Hypothesis" (Yasuda, et al 2004).3

The natural log of the book value of total assets to models (4) and (5) is added to control main

factors that may affect the systematic risk; since larger firms have better access to capital markets, they

may have more flexibility to deal with unexpected liquidity shortfalls. Furthermore, larger firms may

be more capable of diversifying risk (Demsetz & Strahan, 1997). Therefore, the firm size may be

negatively associated with the level of firm risk.

To test the fifth hypothesis "Discretionary accruals are not associated with the level of

insolvency risks,” Model (5) will be used.

Zit= q0+ q1 DISCit-1 +q2ASSETit-1+ e5it

(Model 5)

Where:

Zit= "Z-score", a measure of insolvency risk.

It can be calculated by the following equation (Boyd et al 1993):-

Z= (π/ A +E/A)/ Sr

Where π is total profit, E is total equity, A is total assets, and Sr is a standard deviation of return

on equity (ROE).

Insolvency risk, "Z-score," is a statistics indicating the probability of bankruptcy. This model

investigated if earning management associated with the level of insolvency risk. This model is used to

test the fifth hypothesis.

**Population**

The population will consist of the Industrial Jordanian shareholding companies listed in the first and

second markets in Amman Stock Exchange for the study period (2000-2007). Insurance and banking

sectors were excluded because of their financial nature. There are (120) companies listed from these

two sectors in Amman Stock Exchange.

**Sample**

All industrial and service shareholding companies that satisfy the following conditions will be included

in the study sample:

1. Share prices data are available during the study period (2000-2007) and there is an

availability of data required to calculate study variables.

2. The company didn’t enter in a consolidation process or allocated free shares because these

events affect the company figures such as earnings.

(44) Companies will represent the study sample.

**Study Period**

The study covers the period from 2000 to 2007, some data required for more than this period to

computing some variables such as standard deviation for return of equity and beta, the required data

include the following:-

1. Net Income from 1996 to 2007.

2. Net Sales from 1998 to 2007.

3 (Yasuda, et al 2004) used the same model in his study to investigate the relationship between profitability and risks in

Japanese market.

*Middle Eastern Finance and Economics - Issue 5 (2009)*56

3. Total Assets from 1999 to 2008.

4. Monthly Closing Prices from 1999 to 2008.

5. Net Cash Flow from Operating Activities from 2000 to 2007.

6. Total Fixed Assets from 2000 to 2007.

7. Total Shareholders Equity from 1995 to 2007.

8. Total outstanding Shares from 2000 to 2007.

9. Market Capitalization from 2000 to 2007.

10. Market Index from 2000 to 2007.

**Data Sources**

1. Jordanian shareholding companies guide issued by Amman Stock Exchange for the period

(1995-2008).

2. Financial reports of Jordanian shareholding companies.

**Statistical Analysis Tools**

Several statistical tools have been used in this research. Firstly, descriptive analysis has been used in

order to describe the data of the research. Next, simple and multi regressions were used for testing the

study hypotheses. Lastly, (Adjusted-R2) was used to indicate the incremental information content for

the study variables.

First, the descriptive statistical tools for the main variables are shown, and then the main

statistical problems in these regression models are discussed, lastly the regression models’ statistical

results are provided.

**Descriptive Statistics**

Tables (1) shows the descriptive statistics for the main variables, the descriptive measures include the

mean, median, standard deviation, minimum value, maximum value, percentile 1, and percentile 99.

**Table 1:**

Descriptive Measures

**Variable Mean**

**Median**

**Std.**

**Deviation**

**Minimum Maximum Percentile**

**1 Percentile**

**99**

Net Income 2,479,334 495,419

11,918,971

-68,855,313 150,191,000 -6,470,885 52,662,983

DISC -0.028

-0.001

0.121 -0.714 0.723 -0.460 0.302

STCPRI 0.974

0.703 1.192 367,500 17.42 549,008 4.984

Beta 0.523

0.386

0.974 -8.42 6.34

-1.359

3.902

Z-Score 22.52

16.39 31.06

-2.01 417 -0.669 163.27

The minimum values for the variables are near the percentile (1) and the maximum values are

near the percentile (99), which means that the data are normally distributed, and the convergence

between the mean and the median leads to the same conclusion. The minimum values of net income

are very high negative figures because net income for some companies is high negative numbers.

The discretionary accruals have a negative mean, which means that the total accrual was

negative in the majority of the sample. This happened because the average operating cash flow was

more than the average net income, which is normal because there are much expenditure that have been

deducted from income and have not been included in the operating cash flow such as depreciation and

amortization. The mean and median for discretionary accruals are near zero because some

discretionary accruals figures are positive and others are negative.

57

*Middle Eastern Finance and Economics - Issue 5 (2009)*

Table

**(2) shows Pearson (and Spearman)’s4 correlation matrix: the strongest relationship was**

between current and past profitability 0.303 (0.644), then between future and current profitability 0.192

(0.581), then between future and past profitability 0.126 (0.456).

These correlation results are normal and agree with previous studies, such as Jones (1991) and

Kallunki and Martikainen (2003), and mean that there are strong relationships among company's

profitability over the years.

**Multicollinearity Problem**

As shown in Table (2), there is a relationship among the main variables that may lead to

multicollinearity problem, which will affect the model power and its ability in explaining the results.

Variance Inflation Factor (VIF) has been used, which refers to actual disparity percentage to total

disparity, and if this factor is less than (3) this means that there is no multicollinearity problem (Fox,

1991).

**Table 2:**

Pearson (Spearman) correlation matrix

**Variable Factor**

**PROF**

**it-1**

**PROFit-2**

**DISC**

**STCPRI**

Pearson 0.303***

0.126**

0.083

0.024

**PROFit**

Spearman

(0.644)*** (0.456)*** (0.178)*** (0.38)***

Pearson

0.192***

0.034

-0.057

**PROFit-1**

Spearman

(0.581)*** (0.151)*** (0.366)***

Pearson

0.025

-0.149***

**PROFit-2**

Spearman

(0.102)*

(0.287)***

Pearson

0.214***

**DISC**

Spearman

(0.267)***

**Table 3:**

Regression Model Variance Inflation Factors (VIF)

**Regression Model**

**VIF factor***

1. PROFit= β0+ β 1STCPRIit-1+ e5it 1

2. PROFit= β'0+ β'1PROFit-1+ β'2STCPRIit-1+ e6it

1.007

3. PROFit= γ'0+ γ'1PROFit-1+ γ'2STCPRIit-1+ γ'3 DISCit-1+ e7it

1.06

4. SESRISKit= v0+ v1DISCit-1+v2ASSETit-1+ e8it 1

5. Zit= q0+ q1DISCit-1+q2ASSETit-1+ e9it 1

**Note:*** These values are the highest values in the models

As shown from Table (3), all (VIF) factors are less than (5), so there is no multicollinearity

problem in the regression models.

**Autocorrelation Problem**

The autocorrelation among regression model residuals have been tested using Durbin-Watson factors,

if Durbin-Watson factors are between (1) and (3) there is no autocorrelation problem

**(Alsaeed, 2005).**

4 Pearson correlation factors are for parametric tests and Spearman factors are for non-parametric tests.

*Middle Eastern Finance and Economics - Issue 5 (2009)*58

**Table 4:**

Regression Model Durbin-Watson Factors

**Regression Model**

**Durbin-Watson Factors**

1. PROFit= β0+ β 1STCPRIit-1+ e5it 1.7

2. PROFit= β'0+ β'1PROFit-1+ β'2STCPRIit-1+ e6it

2.3

3. PROFit= γ'0+ γ'1PROFit-1+ γ'2STCPRIit-1+ γ'3DISCit-1+ e7it

2.4

4. SESRISKit= v0+ v1 DISCit-1+v2ASSETit-1+ e8it 1.8

5. Zit= q0+ q1DISCit-1+q2ASSETit-1+ e9it 1.2

As shown in Table (4), all Durbin-Watson factors are between (1) and (3), so there is no

autocorrelation problem in the regression models.

**Hypotheses Testing**

Table (5) reports the results of regressing future profitability on the current profitability, stock price

and discretionary accruals

Table 5:

Table 5:

Relationship between future profitability, stock price and discretionary accruals

**Model #**

**Constant**

**STCPRI**

**PROFit-1 DISCit-1 F-statistic**

**Adj**

**R2**

-1.94 3.74

1

(0.207)

(-1.529) (0.46)

*-0.002*

-1.68 6.624 0.21

2

(19.12)***

(-1.38) (0.84)

(6.17)***

*0.089*

-1.92 1.19 0.208 1.01

3

(12.223)***

(-1.12) (0.90)

(5.831)*** (1.16)

*0.091*

* The Factor is significant at the 0. 1 level.

** The Factor is significant at the 0.05 level.

*** The Factor is significant at the 0.01 level.

**Hypothesis number 1:**Stock prices can't predict future profitability.

**:- PROFit= β0+ β 1STCPRIit-1+ e1it**

Model number 1

Model number 1

In this model, we try to study the explanatory power for stock price relative to future

profitability. As shown in Table (5), the stock price coefficient is not significant and Adj R2 equals

(0%) when running the regression model for the whole study period, which means that null hypothesis

will be accepted so stock prices can't predict future profitability. This result means that the increase in

stock prices doesn’t lead to an increase in future profitability in the Jordanian companies.

**Hypothesis number 2:-**Current profitability doesn’t have incremental information relative to stock

prices when predicting future profitability.

**Model number 2**:-

**PROFit= β'0+ β'1PROFit-1+ β'2STCPRIit-1+ e2it**

We add current profitability in this model to study its incremental information content relative

to stock prices when predicting future profitability. We expect on the basis of the literature review that

current profitability predicts future profitability.

To test this hypothesis, Adj-R2 for Model (2) and Model (1) will be compared, as displayed in

Table (5) the Adj-R2 increases, which means that current profitability has incremental information

relative to stock prices when predicting future profitability.

**Hypothesis number 3:-**Discretionary accruals don't have incremental information relative to current

profitability and stock prices when predicting future profitability.

**Model number 3**:-

**PROFit= γ'0+ γ'1PROFit-1+ γ'2STCPRIit-1+ γ'3DISCit-1+ e3it**

We add lagged discretionary accruals (DISCit-1) to study its effect on future profitability, and to

investigate its incremental information content to current profitability and stock prices when predicting

future profitability.

59

*Middle Eastern Finance and Economics - Issue 5 (2009)*

As shown in Table (5), Adj-R2 increased in Model (3) to (9.1%) and discretionary accruals

factor is significant, which means that discretionary accruals has incremental information content to

current profitability and stock prices when predicting future profitability.

This result indicates that discretionary accruals tools help in predicting future profitability,

which means that managers of Jordanian companies try to control future profitability by using

discretionary accruals.

Table (6) reports the results of regressing systematic and insolvency risks on the discretionary

accruals.

**Table 6:**

Relationship between Risk and discretionary accruals

**Model #**

**Constant**

**DISCit-1 ASSETit-1 F-statistic**

**Adj**

**R2**

4 -0.48

0.61

0.14

(Systematic risk)

(-0.95)

(1.75)*

(1.92)*

(3.44)***

*0.014*

5 5.16

28.72

2.53

(2.12)

(Insolvency risk)

(0.25)

(1.90)*

(0.87)

*0.013*

* The Factor is significant at the 0. 1 level.

** The Factor is significant at the 0.05 level.

*** The Factor is significant at the 0.01 level.

**Hypothesis number (4):-**Discretionary accruals are not associated with the level of systematic risks.

**Model number (4):**SESRISKit= v0+ v1 DISCit-1+v2ASSETit-1+ e4it

In this model, the relationship between company's systematic risk and the managers' tools to

manipulate earnings (discretionary accruals) has been studied.

Table (6) reports that discretionary accruals coefficient is statistically significant and Adj-R2

factor equal (1.4%), which means that discretionary accruals can predict the company's systematic risk,

so the null hypothesis will be rejected, thus, discretionary accruals are associated with the level of

systematic risks.

**Hypothesis number (5):**Discretionary accruals are not associated with the level of insolvency risks.

**Model number (5):**Zit= q0+ q1 DISCit-1+q2ASSETit-1+ e5it

We investigate in this model the relationship between discretionary accruals and bankruptcy or

insolvency risks.

As shown from Table (6), discretionary accruals coefficient is significant and Adj-R2 factor

equals (1.3%) in all year regression.

Therefore, the result is consistent with the naïve hypothesis that investors will misinterpret high

reported earnings as being favorable information about company health.

This result and the result of Model (4) agree with (Yasuda, et al. 2004), which means that

Jordanian investors are not rational and cannot assess the surrounding risks.

**Research Results**

1. Stock prices can't predict future profitability in Jordanian companies, this result happens

because of the weak relationship between stock price and profitability of the firm.

2. Current profitability has incremental information relative to stock prices when predicting future

profitability. This normal result asserts that the best predictor to the future profitability is the

lagged one which agrees with the literature review and previous studies.

3. Discretionary accruals have incremental information content to current profitability and stock

prices when predicting future profitability.

These findings mean that Jordanian companies actively use Discretionary accruals to achieve a

target level in earnings.

1. Discretionary accruals are associated with the level of systematic risks. This means that

Jordanian investors are mislead by the high earnings figures reported by the managers.