New Frederico Nilsen - FLAVIA.pdf · 2012. 6. 2. · Av. Fernando Ferrari, 1358 – Goiabeiras – Vitória, ES, Brazil – 29075-010 [email protected] ou [email protected] Frederico

An analysis of the impact of non-accrued contingent liability on the equity market value of Brazilian companies

Amaury José Rezende

Ph.D. Professor – University of São Paulo – Ribeirão Preto, Brazil Av. Bandeirantes, 3900 – Monte Alegre – Ribeirão Preto, SP, Brazil – 14040-900

[email protected] ou [email protected]

Flávia Zóboli Dalmácio

Ph.D. Professor – FUCAPE Business School Av. Fernando Ferrari, 1358 – Goiabeiras – Vitória, ES, Brazil – 29075-010

[email protected] ou [email protected]

Frederico Nilsen

Undergraduate Student – University of São Paulo – Ribeirão Preto, Brazil Av. Bandeirantes, 3900 – Monte Alegre – Ribeirão Preto, SP, Brazil – 14040-900

[email protected]

May, 2010.

2

An analysis of the impact of non-accrued contingent liability on the

equity market value of Brazilian companies

Abstract: This study investigates the relation between equity market value and non-

accrued contingent liability that is disclosed in the footnotes but it is not recognized in

net income, under Deliberation Nº 489/05 of the Brazilian Securities and Exchange

Commission (CVM) and Statement NPC Nº 22 of the Brazilian Institute of Auditors

(IBRACON). Our analyses focus on whether the prices of shares in the stock market

and the return’s expectation of market react by the disclosure of non-accrued contingent

liability in footnotes. We analyze 1.745 footnotes from periods 2006 to 2008 of 159

companies of different sectors, in twelve quarterlies. These companies are listed in

different corporate governance’s levels of the São Paulo Stock Exchange (BOVESPA) –

Level 1, Level 2 and New Market. This study find that the disclosure of non-accrued

contingent liability in the footnotes understate CVM’s Deliberation Nº 489/05 is

perceived by share price. In particular, our findings are consistent with the results of

Aboody et at (2004) that found that stock options are viewed as expenses and they are

negatively associated with share price. Our findings indicate that investors view the

disclosure of non-accrued contingent liability as expense of the firms that reflect in their

valuation assessments. This suggests that managers believe that even though non-

accrued contingent liability to be disclosed in the footnotes and not recognized as

expense, it is relevant to financial statement users.

Keywords: Non-accrued Contingent Liability, Deliberation Nº 489/05 of the Brazilian

Securities and Exchange Commission (CVM), Timeliness of Share Price.

3

I. INTRODUCTION

The objective of this study is to understand to relation between equity market

value and non-accrued contingent liability that is disclosed in footnotes but not

recognized in net income, under Deliberation Nº 489/05 of the Brazilian Securities and

Exchange Commission (CVM) and Statement NPC Nº 22 of the Brazilian Institute of

Auditors (IBRACON). Accounting for contingent liability, specifically the provision

contingent liability, has been controversial financial reports issues, because debate

among accounting, managers, auditors, lawyers, regulators and accounting standard

setters. The focus of discussion on issues whether the estimation of the value of

provision of contingent liabilities represent the level of risk of the transactions realized

by enterprises. The lawyers and auditors have discusses whether it can be measured

reliably enough as an expense of the firm, same than disclosed in off balance sheet (in

the footnotes).

Specifically, our analyses focus on whether the prices of shares in the stock

market and the return’s expectation of market react by the disclosure of non-accrued

contingent liability in footnotes. It is based on estimates of realization, which depend on

expectations about the future.

Our findings indicate that investors view the disclosure of non-accrued

contingent liability as expense of the firm that reflected in their valuation assessments.

We find a significant negative relation between (Panel 1, 2, and 3) non-accrued

contingent liability variable and share price. These results can to indicate that the

investors view and understand this type information as a risk potential that can be

4

realized in the future. These results are adherents as prior researches which analyze the

effects of disclosed expenses in the footnotes, for example Gopalakrishnan (1994), Rees

and Stott (1998), Li (2002); Bell et al (2002) and Belzile et al (2006).

The criteria adopted by firms to estimate the provision represents a

recommendation of International Accounting Standards Boards (IASB). It argues that “a

provision shall be recognized when: (a) An entity has a present obligation (legal or

constructive) as a result of a past event; (b) It is probable that an outflow of resources

embodying economic benefits will be required to settle the obligation; and (c) A reliable

estimate can be made of the amount of the obligation. If these conditions are not met, no

provision shall be recognized. If these conditions are not met, no provision shall be

recognized” (IAS 37, par. 14).

Further, Brazilian accounting law requires firms that participate in the capital

market to adopt the recommendations of the CVM and IASB.

IASB defines a contingent liability as “ (a) a possible obligation that arises from

past events and whose existence will be confirmed only by the occurrence or non-

occurrence of one or more uncertain future events not wholly within the control of the

entity; or (b) a present obligation that arises from past events but is not recognized

because: (i) it is not probable that an outflow of resources embodying economic benefits

will be required to settle the obligation; or (ii) the amount of the obligation cannot be

measured with sufficient reliability” (IAS 37, par. 10).

This paper proceeds as follows. The next section summarizes discussion about

disclosure of accounts off balance Sheet and it discusses the possible effects of

understate CVM’s Deliberation Nº 489/05 to classify a contingent liability and related

5

research. The section 3 outlines the research design. The section 4 describes the data

and descriptive statistics. The sections 5 and 6 present our findings and the section 6

concludes.

II. BACKGROUND

Disclosure Accounts off Balance Sheet

In some cases, companies should disclose information that is up the net income,

but that in the future will impact on their results. This obligation imposed by law is a

way to reduce information asymmetry, which, according to Hendriksen and Breda

(1992), occurs when one of the parties of a transaction has more information than

another.

As an example, of this situation, we can cite the treatment of contingent

liabilities in Brazil and the stock option-based compensation in the USA. According to

Aboody et al (2006), accounting for stock option-based compensation is specified in

Accounting Principles Board Opinion (APB) No. 25 (APB, 1973)1 and SFAS 1232. If a

firm measures the expense under APB 25, SFAS 123 requires disclosure of pro forma

net income, which is what net income would have been had SFAS 123 expense been

recognized.

1 Under APB 25, stock option-based compensation expense is based on the difference at the measurement date between the stock price and option exercise price. Because for most fixed option grants the exercise price equals the stock price at the date of grant, the expense under APB 25 typically equals zero. 2 Under SFAS 123, the expense is calculated based on the option’s fair value at grant date, and is not adjusted for subsequent changes in value. SFAS 123 expense is grant-date option value multiplied by the number of granted options, amortized over the vesting period.

6

In this context, the Brazilian case, depending on the probability of occurrence of

contingent liabilities, they are accrued or only disclosed in footnotes (non-accrued). The

effect of non-accrued contingent liabilities, in relation to the Brazilian capital market,

needs study. The Brazilian Accounting’ rules recommend that determined transactions

those envelopment high levels of risk should be recognized and disclosed in the off

balance sheet, since that there are high levels probability’s realization.

The enterprises in the Brazil should to follow the rules of CVM’s Deliberation

Nº 489/05 to classify a liability contingent. The CVM’s Deliberation Nº 489/05 divides

the contingent liabilities in three groups:

• Probable – the companies need to disclose the provision account in the

balance sheet and the results of periods.

• Possible – the companies do not need to disclose the provision account

neither balance sheet or results of periods. However, the companies are

required to report an estimate of possible loss in footnotes.

• Remote - the companies are not required to provision neither report it.

Similarly as research of Aboody et al (2004), the objective of this study is to

determine whether the measuring reliability problems are sufficient to render non-

accrued contingent liability as expense valuation relevant to disclose it. In this context,

Williams and Gonçalves (2007) argue that there is a criteria’s lack to define and to

analyze of contingent expenses. They affirm that this facts have been classified of away

equivocated concern to of risks judgment to companies and are resultant in estimative

higher than that the prior.

7

Related Research

In this study, we examine the value relevance of the disclosed pro forma

expense, more specifically, the potential non-accrued contingent liability. The evidence

from prior studies generally suggests that the disclosures in the footnotes are perceived

as expenses and they are value relevant and incorporated into share prices.

Gopalakrishnan (1994) investigate if investors, in determining security prices,

differentiate between information recognized in the balance sheet and that disclosed in

the footnotes. The findings indicate that investors appear to consider pension

information disclosed in the footnotes as value-relevant, given that an accrued (prepaid)

pension liability (asset) is also recognized in the balance sheet. Second, investors attach

equal importance to both sorts of pension information.

Rees and Stott (1998) employ pro-forma company footnote disclosures to assess

the value-relevance of employee stock option compensation expense using the fair value

method as stipulated by Statement of Financial Accounting Standard No. 123. They find

a significant association between the disclosed compensation expense using the fair

value method and firm value. This result implies that the disclosed employee stock

option expense is a value-relevant measure and the incentives derived from employee

stock option plans provide value-increasing benefits to the firm.

Li (2002) provides a theoretical analysis and empirical investigation of the

valuation implications of employee stock options. The empirical results indicate the

existence of a cross-sectional negative association between share prices and both

outstanding employee stock options and expected stock option expense. Li (2002) find

8

that SFAS 123 footnote disclosures at 10-K filings communicate useful information

about employee stock options to investors.

Bell et al (2002) investigate the relation between share price and SFAS No. 1233

expense and they find an insignificant relation between share price and their stock-based

compensation expense variable for a sample of profitable computer software firms. But,

according to Aboody et al (2004) these results for computer software firms could not be

extend to other industries.

Aboody et al (2004) investigate the relation between share price and stock-based

compensation expense that is disclosed but not recognized under Statement of Financial

Accounting Standards (SFAS) No. 123, controlling for net income, equity book value

and expected earnings growth. They find stock options are viewed as an expense and

are negatively associated with share price. They find that investors view SFAS No. 123

(stock-based compensation) expense as an expense of the firm, and as sufficiently

reliable to be reflected in their valuation assessments.

Belzile et al (2006) examine whether changes in the way stock option

compensation is reported (recognition as an expense in the income statement or note

disclosure of pro forma net income and earnings per share) affect financial statement

users' judgments and investment decisions. The results indicate that the reporting

method does indeed significantly influence subjects' judgment of the expected stock

price direction, but has no material influence on their investment decisions.

3 SFAS No. 123 requires firms to disclose (in footnotes to the financial statements) the pro forma effects on earnings of employee compensation expense attributable to amortizing the fair value of employee stock options at the grant date. However, SFAS No. 123 does not generally require firms to recognize this employee stock options related compensation expense in the income statement, although it encourages firms to do so (Bell et al, 2002).

9

Laux and N’Dir (2007) investigate market reaction to SFAS 123 Revised, which

requires companies to recognize the fair value of employee stock options as expense on

the income statement. Using a sample of 128 firms for the 2004 and 2005 periods, they

find that markets have efficiently incorporated information formerly disclosed only in

footnotes to the financial statements, effectively nullifying the argument that formally

recognizing the expense would have a deleterious effect on stock prices of firms

offering this type of compensation. The results indicate that the market values stock-

based compensation expense efficiently whether disclosed or recognized formally in the

financial statements.

Niu and Xu (2009) examine the market valuation of employee stock option

expenses recognized by using the fair value approach under the Canadian Institute of

Chartered Accounts Handbook section (CICA HB) 3870. Based on a sample of

Canadian public firms traded on the Toronto Stock Exchange (TSX), they find that

investors value employee stock option expenses differently prior to and after the

implementation of the new standard. Specifically, pro forma compensation expenses

disclosed prior to the new accounting regulation are negatively associated with annual

stock returns, suggesting that the market interprets these expenses to have negative

valuation consequences. In contrast, recognized stock option expenses from using the

fair value approach mandated by the CICA HB 3870 are positively associated with

stock returns, indicating that the market now interprets these expenses as a type of

“asset” that contributes positively to firm valuation.

Hitz (2009) investigates the incidence and motives for disclosure of so-called

‘pro forma earnings’ – voluntarily disclosed earnings metrics that modify the bottom-

10

line number of the income statement to arrive at a (more) representative income

presentation. The results indicate that firms make extensive use of so-called earnings

before-metrics and, more importantly, of pure non-GAAP performance measures, both

in terms of frequency and reporting emphasis. This is accompanied by a low average

level of transparency for non-GAAP adjustments. Multivariate analysis of determinants

yields that both informative and strategic motives drive pro forma earnings disclosure

decisions, and that recent regulatory recommendations have had no discernible impact.

III. RESEARCH DESIGN

Our first set of tests focuses on determining whether expenses disclosed in

footnotes under CVM’s Deliberation Nº 489/05 and IBRACON’s Statement NPC Nº 22

explain disclosed option value estimates of the firms. In particular, we base our

inferences on the following equation:

ticg

itcgs

itsa

ita

itititititit

eCGSQ

ASSETNACLMEFNIBVP

,

2

2

19

2

11

2

54321

++++

+++++=

∑∑∑===

ξγδ

ααααα

(A)

Pit is share price, BVit is equity book value, NIit is net income, MEFit is mean

earnings forecast that is represented by mean year-end analyst and quarterly-end analyst

(Thomson ONE Analytics® Database), NACLit is non-accrued contingent liability4,

4 The liabilities analyzed in this study represent just the liability estimate and disclosed in the footnotes, because the CVM’s Deliberation Nº 489 and the IBRACON’s Statement NPC Nº 22 recommend which to account the contingent liabilities only whether the levels of risk is considered possible.

11

ASSET is the asses value. All variables were measured at quarterly-end and fiscal year-

end i and t denote firm and period.

We input three control’s variables: periods (Q); industry or sector (S); and

corporate governance’s level (CG). Q is period dummy variable; S is sector or industry

dummy variable; and CG is corporate governance’s level dummy variable which equals

1 (one) if firm i is in the Level 1, Level 2 or New Market of corporate governance’s

levels of the BOVESPA and zero otherwise. Subscripts i and t denote firms and period.

Thus, our estimating (Equation A) permits us to test whether the effects of

NACLit (non-accrued contingent liability) are associated with the variation of the Pit

(share price) of firms. Besides we include three vectors of control variables, which also

are described.

Our first analyses were based on a model deflated by number of shares

outstanding and at the end of the year which comprises following variables BV, NI,

MEF, NACL and ASSETS. We estimate all equation using robust regression.

The model (A) is consistent with Ohlson (1995). We include the expected future

abnormal earning is reflected in equity price before they are reflected in equity book

value and net income (Liu and Ohlson, 2000; Ohlson, 2001).

The operational earning of firm does not include the expense of NACL. Thus,

we hope which MEF can capture the impacts associated disclosed NACL in the

footnotes (off balance sheet) no reflected in net income.

We believe that expense as NACL can is likely related with expectative of

growth operational income; omitting MEF from (1) could affect our inferences about

12

relations Price and NACL. Based on prior research, we predict too α1, α2, and α3 are

positive. Thus, if investors view NACL as a future expense of the firms, and to the

extent that any effect associated or reflected on net income, equity book value, and

expected earnings grow, thus we predict α4 with signal negative.

To investigate whether disclosed NACL in the footnotes (off balance sheet) on a

timely base change in investor-perceive costs-perceptive cost associate with expense

NACL, we estimate:

ticg

itcgs

itsa

ita

itititititit

eCGSQ

AssetsNACLEGFNINIRET

,

2

2

19

2

11

2

54321

++++

++∆++∆+=

∑∑∑===

ξγδ

ααααα

(B)

RETit is annual share return and/or quarterly, and ∆ denotes annual change, ∆

NIit is net income, EGFit is earnings grow forecast that is represented by mean year-end

analyst and/or quarterly-end analyst (Thomson ONE Analytics® Database), and ∆

NACLit is non-accrued contingent liability5. All variables were deflated by share price

at the beginning of quarterly or year i and t denote firm and period.

IV. DATA AND DESCRIPTIVE STATISTICS

We collect the data for firms in the Economatica® Database and the firms listed

in the Securities and Exchange Commission of Brazil (CVM) and in the São Paulo

5 The liabilities analyzed in this study represent just the liability estimate and disclosed in the footnotes, because the CVM’s Deliberation Nº 489/05 and the IBRACON’s Statement NPC n º 22, recommended to account the contingent liabilities only whether the levels of risk is considered possible.

13

Stock Exchange (BOVESPA). We rank firms by BOVESPA’s corporate governance

levels. According to these levels of the corporate governance, the firms are divided into

three categories: Level 1, Level 2 and New Market.

To identify firms that published non-accrued contingent liability (NACL), we

have analyzed 1.745 footnotes published off balance sheet. It was selected a sample

between 96 to 159 firms with a total of 1.908 observations, during the periods from

2006 to 2008, that comprised 12 quarterly. We identify that in average 67,15% of firms

disclosed non-accrued contingent liability in the footnotes and 91,24% of firms had

accrued contingent liability (Table 13).

The data were collect from financial statements footnotes of firms. We identified

the nature of contingent liabilities, for example: 27,07% civil, 29,45% fiscal, 23,70%

and 19,77% other types6.

We use panel data, according to Pindyck and Rubinfeld (2004), Gujarati (2006)

e Fávero et al (2009), allows, among other features: increase considerably the sample

size; increased number of observations, increasing the degrees of freedom and

efficiency of the data; reduction of problems of multicollinearity of explanatory

variables; intertemporal dynamic, represented by the mix between the cross-sections

and time series; data more informative, and study more complex behavioral models. The

sample of work can be considered an unbalanced panel because it has different numbers

of observations of court for each year.

6 Others type represents the contingent liabilities as: environment, customers and suppliers.

14

V. PRIMARY FINDINGS

In the Table 1 and in the Table 2, we summary statistics from the Equation 1 and

the Equation 2.

tis

itsitititititit eSASSETMEFNACLNIBVP ,

19

254321 ++++++= ∑

=

γααααα (1)

ticg

itcgs

itsa

ita

itititititit

eCGSQ

ASSETMEFNACLNIBVP

,

2

2

19

2

11

2

54321

++++

+++++=

∑∑∑===

ξγδ

ααααα

(2)

In the Equation 1, we insert a dummy variable for industries (S) with variable

control. We find a relation significant between share price and NACL. As predicted, the

coefficient on BV and ASSET are significantly positive and the variable NACL is

significantly negative, but NI and MEF are not significant.

In the Equation 2, we insert more two dummy variables as control variables:

periods (Q) and corporate governance’s levels (CG). Therefore, when we add others

variable of control the results present better performance.

The results of the Equations 1 and 2 (Panel 1 and Panel 2) show that the model

based in the data deflated by number of share outstanding is able to capture the effect

15

that the market give to non-accrued contingent liability in the footnotes. Thus, it seems

that the market has absorbed this potential risk (information) and it has demanded a

premium for disclosed risk by the company.

In the Table 3, we summary statistics from the Equation 3.

tis

itsa

ita

itititititit

eSQ

ASSETMEFNACLNIBVP

,

19

2

11

2

54321

+++

+++++=

∑∑==

γδ

ααααα

(3)

The results of the Equation 3 (Panel 3) indicate that the model deflated7 too is

able to capture the effect that the market gives to non-accrued contingent liability in the

footnotes. However, the coefficient ASSET is not significant.


the Equation 5.

ticg

itcgs

itsa

ita

itititititit

eCGSQ

ASSETMEFNACLNIBVRET

,

2

2

19

2

11

2

54321

++++

+++++=

∑∑∑===

ξγδ

ααααα

(4 and 5)

7 All variables were deflated by share price at the beginning of quarterly or year.

16

We estimate the Equation 4 (Panel 4) to verify the relationship between the BV,

NI, NACL and ASSET variables and the return of share. Our results indicate that only

NACL and NI are significant. However, the values of estimated coefficients are very

small.

The obtained sign of the coefficient NACL (Panel 4) is contrary to expected, so

we recommend that results are analyzed with carefully. Nevertheless, the return of

market in the periods from 2006 to 2008 é influenced by the crisis in global capital

market. We observe that the majority of the shares of Brazilian companies have

negative returns, in almost all analyzed periods.

In the Equation 5 (Panel 5)8, we evaluate the adherence of accounting variables

on return of share. Therefore, the results indicate that there are not significant

differences between in the predictive of models with data deflated by number of shares

outstanding or with share price at the beginning of quarterly or year.

To investigate whether non-accrued contingent liability reflects on a timely basis

in investor perceived associated with this risk. We use the proposed model by Bushman

et al. (2004). Additionally, three control variables were included: industries, periods,

and corporate governance’s levels. Thus, we estimate:

8 The differences between the panel 4 and 5 is in the method of deflated. The data of Panel 4 were deflated just by number of shares outstanding and at the end of the year, while the data panel 5 was deflated by share price at the beginning of quarterly or year.

17

ticg

itcgs

itsa

itait

itititititit

eCGSQASSET

EGFACLNACLNINIRET

,

2

2

19

2

11

26

54321

+++++

+∆+∆+∆+∆+=

∑∑∑===

ξγδα

ααααα

(6 and 7)

Where RET is quarterly return of share price, and ∆ denotes quarterly change.


the Equation 7.

In the panel 6 and 79, we estimate if return of share price has captured the effects

that the market gives to non-accrued contingent liability disclosed in the footnotes. The

results do not allow concluding that there is significant evidences of the usefulness and

timeliness of variation of NACL and ACL. Only the NI and ASSET variables show

significant, in both models.

VI. ADDITIONAL ANALYSES

It was done some additional analyzes that has as objective to measure the effects

of potential risks of contingent liability disclosed in the footnotes about Earning Grow

9 The model used was proposed by Bushman et al (2004).

18

Forecasts (EGF) and Mean Earning Forecasts (MEF) elaborated by annalists of capital

market.


the Equation 9.

ticg

itcgs

itsa

ita

itititititit

eCGSQ

ASSETACLNACLNIBVMEF

,

2

2

19

2

11

2

54321

++++

+++++=

∑∑∑===

ξγδ

ααααα

(8 and 9)

The results of the Panel 8 indicate that the variable ACL (accrued contingent

liability) is statistically significant. Further, the coefficient of variable NACL present

negative signal but it is not statistically significant. Only the NI and ASSET variables

are statistically significant at levels of 5% and 10%.

In the Panel 9, all variables are deflated by share price at the beginning of

quarterly or year, but only the NI and ASSET variables are statistically significant. Thus

we recommend which conclusions are done with carefully.

In the Equation 10 and 11, we estimate if non-accrued contingent liability

(NACL) disclosed in the footnotes has impact on the Earning Grow Forecasts (EGF). In

this equation we use basically the variation of accounts between periods, according to

the equation bellow.

19

ticg

itcgs

itsa

itaitit

ititititititit

eCGSQASSETACL

ACLNACLNACLNINIBVEGF

,

2

2

19

2

11

287

654321

++++∆+∆+

+++∆++∆++=

∑∑∑===

ξγδαα

αααααα

(10 and 11)

In the Table 10 and in the Table 11, we summary statistics from the Equation 10

and the Equation 11.

According to the Table 10 (Panel 10), the ∆NACL variable presents a relation

statistically significant with EGF. However, the signal of the variable is contrary as

expected. These finding do not support the conjecture theoretical. Since the

announcement of the risk should have a negative impact on the forecasts of growth of

profits.

The findings of the Panel 11 show that only NI variable is statistically

significant. Unfortunately, we can not to identify evidences of that the disclosure of

non-accrued contingent liability in the footnotes has impact on EGF (earning grow

forecasts) variable. Thus, we believe that our results of the Panel 10 and the Panel 11

are inconclusive to determine any relation between the analyzed variables.

20

VII. SUMMARY AND CONCLUIDING REMARKS

This study investigates the relation between share price and non-accrued

contingent liability not recognizes in net income under CVM’s Deliberation Nº 489/05,

but disclosed in the footnotes. We use two models10 to evaluate the impact of the NACL

(non-accrued contingent liability) on share price and on return of share price. We use

with asset, industry, period of published and corporate governance’s levels with control

variables.

We estimate seven models, where: (i) three estimative relate share price to the

NACL variable; and (ii) four estimative examine the relationship between return of

share price with the NACL variable. Additionally, we analyze the relation between the

mean earning forecast (MEF) variable and the earning grows forecast (EGF) variable

with the non-accrued contingent liability (NACL) variable.

Our expectations were that there was a negative relation between the non-

accrued contingent liability (NACL) variable and share price and between the non-

accrued contingent liability (NACL) variable and return of share price. We find a

significant statistically relationship just to the share price.

In the panel 4 and 5, we find a significant statistically relationship between the

RET11 and the NACL variables. But, the effects find are not of according to expect. The

coefficients are significant statistically, but the signals are contrary to we expected.

10 (i) The model share price is based on model proposed by Ohlson (1995); Liu and Ohlson (2000); Ohlson (2001); and (ii) The model share return was proposed by Bushman et al. (2004). These models were used by Aboody, Barth and Kasznik (2004). 11 In this analyze all variables accounts were deflated just by number of shares outstanding and at the end-year.

21

Additionally, analyzing the relationship between the MEF and the NACL variables; and

the relationship between the EGF and the NACL variables, we do not find conclusive

results.

The primary results of this study find that the non-accrued contingent liability

disclosed in the footnotes understate CVM’s Deliberation Nº 489/05 is perceived by

share price. In particular, our findings are consistent with the results of Aboody et al

(2004) that found that stock options are viewed as an expense and they are negatively

associated with share price. Aboody et al (2004) found that investors view SFAS Nº 123

(stock-based compensation) expense as an expense of the firm, and as sufficiently

reliable to be reflected in their valuation assessments.

This suggests that managers believe that even though non-accrued contingent

liability to be disclosed in the footnotes and not recognized as expense, it is relevant to

financial statement users. More importantly, our findings suggest that some concerns

about the overall reliability of non-accrued contingent liability disclosed in the footnotes

are not warranted. We leave it to standard setters to determine whether the effects of

discretion on reliability are sufficient to cause them concern and, if so, how such effects

can be mitigated.

Our study is silent on the potential implications of changing the accounting

treatment of non-accrued contingent liability instituted by the CVM’s Deliberation Nº

489/05 and the IBRACON’s Statement NPC Nº 22. Although expense do not have seen

recognition would likely provide managers with greater incentives to understate the

expense, it would also likely increase costs related to audit, regulatory enforcement, and

scrutiny by investor groups associated with doing so.

22

We identified three limitations: (i) the composition of the sample because

include only companies listed in one of three corporate governance’s levels of the São

Paulo Stock Exchange (BOVESPA) – Level 1, Level 2 and New Market; (ii) The

number of companies surveyed, and (iii) Many companies do not report the values of

contingency liabilities in their footnotes, even being required to publish.

For future research, we suggest examining the type of provision represents an

important information for the market or if the relationship is restricted only to aggregate

information. Besides the increase time analyzed and number of companies.

23

REFERENCES

Aboody, D. 1996. Market Valuation of Employee Stock Options. Journal of Accounting and Economics 22 (Aug.-Dec.): 357-391.

Aboody, D., Barth, M. E., Kasznik, R. 2001. SFAS 123 Stock-Based Compensation Expense and Equity Market Values. Working Paper, Social Science Research Network (www.ssrn.com).

Aboody, David; Barth, Mary E; Kasznik, Ron. 2004. SFAS No. 123 Stock-Based Compensation Expense and Equity Market Values. The Accounting Review; Apr 2004; 79, 2; ABI/INFORM Global, p. 251-275.

Aboody, David; Barth, Mary E; Kasznik, Ron. 2006. Do Firms Understate Stock Option-Based Compensation Expense Disclosed under SFAS 123? Forthcoming, Review of Accounting Studies.

Balera, Wagner; Raeffray, Ana Paula Oriola de. Os Advogados e as provisões contábeis. Jornal Valor Econômico, São Paulo, 03/05/2007. Disponível em .

Bell, T. B.; Landsman, W. R; Miller, B. L.; Yeh, S. 2002. The valuation implications of employee stock option accounting for profitable computer software firms. The Accounting Review, 77, p. 971-996.

Belzile, Réjean; Fortin Anne; Viger Chantal. 2006. Recognition versus Disclosure of Stock Option Compensation: An Analysis of Judgements and Decisions of Nonprofessional Investors. Canadian Accounting Perspectives, v. 5, n. 2, p. 147-179.

BOVESPA. Bolsa de Valores de São Paulo. 2008. Disponível em: .

Bushman, R.; Chen, Q.; Engel, E.; Smith, A. 2000. The sensitivity of corporate governance systems to the timeliness of accounting earnings. Working Paper. Chapell Hill, University of North Carolina, 63 p.

24

Bushman, R.; Chen, Q.; Engel, E.; Smith, A. 2004. Financial accounting information, organizational complexity and corporate governance systems. Journal of Accounting and Economics, 37:167-201.

Bushman, R.; Smith, A. 2001. Financial accounting information and corporate governance. Journal of Accounting and Economics, 32:237-333.

CVM – Comissão de Valores Mobiliários. Estudo sobre as práticas contábeis brasileiras e as normas internacionais de contabilidade (IFRS). Disponível em acesso em 01/10/2008.

CVM – Comissão de Valores Mobiliários. Deliberação CVM Nº 489/05. Disponível em acesso em 01/10/2008.

CVM – Comissão de Valores Mobiliários. Instrução CVM Nº 202. Disponível em .

Fávero, L. P. L.; Belfiore, P.; Silva, F. L. ; Chan, B. L. . Análise de Dados: Modelagem Multivariada para Tomada de Decisões. Rio de Janeiro: Campus Elsevier, 2009. 646 p.

Gopalakrishnan, V. 1994. The effect of recognition vs. disclosure on investor valuation: The case of pension accounting. Review of Quantitative Finance and Accounting, v. 4, n. 4, p. 383-396.

Gujarati, D. 2006. Econometria básica. Rio de Janeiro, Campus, 812 p.

Hendriksen, Eldon S.; Van Breda, Michel F. Tradução de Antonio Zoratto Sanvicente. Teoria da Contabilidade. São Paulo: Atlas, 1999.

Hits, Joerg-Markus. 2009. Information Versus Strategic Reporting: Disclosure of Pro Forma Earnings by Large German Corporations. In European Accounting Association Annual Meeting, Tampere, Finland.

IAS 37. International Accounting Standards 37. 2002. In: International Accounting Standards. IASB: Londres.

25

IBRACON. Instituto Brasileiro dos Auditores Independentes. Norma e Procedimento de Contabilidade nº 22. Disponível em . acesso em 01/10/2008.

Laux, Judy A.; N’Dir Abdou. 2007. Employee Stock Options And Market Efficiency. Journal of Applied Business Research, v. 23, n. 2, p. 1-6.

Li, Haidan. 2002. Employee Stock Options, Residual Income Valuation and Stock Price Reaction to SFAS 123 Footnote Disclosures. Available at SSRN: http://ssrn.com/abstract=376580 or doi:10.2139/ssrn.376580.

Niu, Flora; Xu, Bixia. 2009. Does Recognition Versus Disclosure Really Matter? Evidence from the Market Valuation of Recognition of Employee Stock Option Expenses. Asia-Pacific Journal of Accounting & Economics, 16, p. 215–234.

Pindyck, Robert; Rubinfeld, Daniel L. Microeconomia. 5. ed. São Paulo: Prentice Hall, 2002.

Rees, Lynn L.; Stott, David M. 1998. The Value-Relevance of Stock-Based Employee Compensation Disclosures. Available at SSRN: http://ssrn.com/abstract=145929 or doi:10.2139/ssrn.145929.

Watanabe, Marta; Tributos não-cumulativos causam mais provisões – Jornal Valor Econômico, São Paulo, 03/02/2009. Disponível em .

Williams, Robert Ellis; Gonçalves, Luís Guilherme B. Classificação contábil X classificação jurídica. Jornal Valor Econômico, São Paulo, 16/08/2007. Disponível em .

26

APPENDIX

TABLE 1 Summary Statistics from Regression Equation 1

(Panel 1)


(Panel 2) PRICE Coef. Std. Err. t-statistic P>|t| PRICE Coef. Std. Err. t-statistic P>|t|

BV 0.052 0.030 1.72 0.086*** BV 0.061 0.031 1.99 0.047** NI -1.608 1.882 -0.85 0.393 NI -1.981 1.871 -1.06 0.290

NACL -0.066 0.021 -3.07 0.002* NACL -0.068 0.020 -3.41 0.001* MEF 2.997 1.885 1.59 0.112 MEF 4.304 1.952 2.20 0.028**

ASSET 0.021 0.005 4.07 0.000* ASSET 0.011 0.008 1.40 0.161 n. Obs. 1138 n. Obs. 1138

Statistic-F 179.98 Statistic-F 129.39 (Prob.) 0.000 (Prob.) 0.000

R-squared 0.7706 R-squared 0.7908 Notes: *Significance level 1%; **Significance level 5%; ***Significance level 10%. PRICE is share price, BV is equity book value, NI is net income, NACL non-accrued contingent liability, MEF is mean earnings forecast that is represented by mean year-end analyst and quarterly-end analyst (Thomson ONE Analytics® Database). In equation 1, we controlled the effect just of sectors that it was statistically significant. In equation 2, we controlled the effects: governance level, period and sectors, but only the period was statistically significant.

27


(Panel 3) PRICE Coef. Std. Err. t-statistic P>|t|

BV -0.347 0.104 -3.35 0.001* NI 2.930 1.645 1.78 0.075***

NACL -0.274 0.115 -2.38 0.017** MEF -6.354 3.646 -1.74 0.082***

ASSET 0.011 0.030 0.37 0.709 n. Obs. 1138

Statistic-F 127.95 (Prob.) 0.000

R-squared 0.7782 Notes: *Significance level 1%; **Significance level 5%; ***Significance level 10%. PRICE is share price, BV is equity book value, NI is net income, NACL non-accrued contingent liability, MEF is mean earnings forecast that is represented by mean year-end analyst and quarterly-end analyst (Thomson ONE Analytics® Database). In the equation 3, we controlled the effects: period and sectors, but it being that for most periods and almost all sectors were statistically significant.

28


(Panel 4)


(Panel 5) RET Coef. Std. Err. t-statistic P>|t| RET Coef. Std. Err. t-statistic P>|t| BV -0.000 0.000 -0.1 0.920 BV 0.000 0.005 0.09 0.931 NI 0.014 0.006 2.15 0.032** NI 0.060 0.049 1.230 0.218

NACL 0.001 0.000 4.4 0.000* NACL 0.012 0.003 3.98 0.000* MEF -0.006 0.009 -0.67 0.502 MEF 0.209 0.134 1.57 0.117

ASSET -0.000 0.000 -0.27 0.785 ASSET 0.001 0.001 0.37 0.715 n. Obs. 1134 n. Obs. 1134


R-squared 0.426 R-squared 0.4362 Notes: *Significance level 1%; **Significance level 5%; ***Significance level 10%. RET is quarterly return of share price, BV is equity book value, NI is net income, NACL non-accrued contingent liability, MEF is mean earnings forecast that is represented by mean year-end analyst and quarterly-end analyst (Thomson ONE Analytics® Database). In the equation 4 and 5, we controlled the effects of: governance level, period and sectors, but only the period was statistically significant.

29


(Panel 6)


(Panel 7) RET Coef. Std. Err. t-statistic P>|t| RET Coef. Std. Err. t-statistic P>|t| NI 0.017 0.009 1.86 0.064*** NI 0.279 0.120 2.33 0.020**

∆NI - 0.001 0.003 -0.26 0.797 ∆NI -0.001 0.001 -0.39 0.697 ∆NACL - 0.000 0.000 -0.08 0.933 ∆NACL 0.000 0.000 0.03 0.973 ∆ACL - 0.000 0.000 -0.46 0.646 ∆ACL - 0.000 0.000 -0.84 0.401 EGF -0.006 0.005 -1.25 0.211 EGF 0.002 0.006 0.28 0.782

ASSET -0.000 0.000 -0.02 0.983 ASSET 0.013 0.006 2.08 0.038** n. Obs. 482 n. Obs. 468


R-squared 0.4682 R-squared 0.5151 Notes: *Significance level 1%; **Significance level 5%; ***Significance level 10%. RET is quarterly return of share price, NI is net income, NACL non-accrued contingent liability, ACL is accrued contingent liability, EGF is earnings grow forecast that is represented by mean year-end analyst and/or quarterly-end analyst (Thomson ONE Analytics® Database), ∆ denotes annual change. In the equation 6, we controlled the effects of: governance level, period and sectors, but only the period was statistically significant. In the equation 7, we controlled the effects of: governance level, period and sectors, but only the governance level was statistically significant.

30


Panel 8


Panel 9 MEF Coef. Std. Err. t-statistic P>|t| MEF Coef. Std. Err. t-statistic P>|t| BV 0.002 0.002 0.93 0.350 BV 0.004 0.003 1.17 0.241 NI 0.179 0.081 2.21 0.027** NI 0.319 0.104 3.07 0.002*

NACL -0.002 0.002 -0.82 0.413 NACL -0.003 0.002 -1.2 0.229 ACL 0.043 0.018 2.45 0.015** ACL -0.035 0.032 -1.09 0.275

ASSET -0.000 0.000 -1.76 0.078*** ASSET 0.001 0.000 2.38 0.017** n. Obs. 1274 n. Obs. 1138


R-squared 0.5716 R-squared 0.5682 Notes: *Significance level 1%; **Significance level 5%; ***Significance level 10%. MEF is mean earnings forecast that is represented by mean year-end analyst and quarterly-end analyst (Thomson ONE Analytics® Database), BV is equity book value, NI is net income, NACL non-accrued contingent liability, ACL is accrued contingent liability. In equation 8, we controlled the effects of: governance level, period and sectors, but they were not statistically significant. In equation 9, we controlled the effects of: governance level, period and sectors, but only the period was statistically significant.

31


(Panel 10)


(Panel 11) EGF Coef. Std. Err. t-statistic P>|t| EGF Coef. Std. Err. t-statistic P>|t| BV -0.005 0.003 -1.44 0.151 BV 0.148 0.120 1.23 0.218 NI 0.213 0.118 1.81 0.072*** NI 1.374 0.580 2.37 0.018**

∆NI 0.006 0.021 0.29 0.775 ∆NI -0.016 0.018 -0.88 0.381 NACL 0.064 0.089 0.73 0.469 NACL 0.097 2.706 0.04 0.971

∆NACL 0.001 0.001 2.34 0.020** ∆NACL 0.001 0.001 1.51 0.131 ACL 0.013 0.031 0.42 0.677 ACL 0.137 1.079 0.13 0.899

∆ACL 0.000 0.000 0.35 0.727 ∆ACL 0.001 0.001 1.05 0.295 ASSET 0.000 0.001 0.18 0.854 ASSET -0.020 0.045 -0.44 0.661 n. Obs. 498 n. Obs. 468


R-squared 0.5994 R-squared 0.5264 Notes: Notes: *Significance level 1%; **Significance level 5%; ***Significance level 10%. EGF is earnings grow forecast that is represented by mean year-end analyst and/or quarterly-end analyst (Thomson ONE Analytics® Database), BV is equity book value, NI is net income, NACL non-accrued contingent liability, ACL is accrued contingent liability, ∆ denotes annual change. In the equations 10 and 11, we controlled the effects: governance level, period and sectors, but in both equations, the variables were not statistically significant.

32

TABLE 12 Pearson Correlation

RET PRICE BV NI NACL ACL

PRICE -0.026

Sig 0.337

BV 0.024 0.154

Sig 0.364 0.000

NI 0.155 0.081 0.035

Sig 0.000 0.005 0.193

NACL 0.026 0.000 0.0211 0.006

Sig 0.331 0.991 0.387 0.834

ACL 0.058 0.082 0.409 0.118 0.002

Sig 0.031 0.002 0.000 0.000 0.933

ASSET 0.048 0.435 0.853 0.034 0.017 0.996

Sig 0.085 0.000 0.000 0.230 0.516 0.000

PRICE is share price, BV is equity book value, NI is net income, NACL non-accrued contingent liability, ACL is accrued contingent liability.

33

TABLE 13 Relations of Surveyed Firms

Periods Total of Firms Nº of Firms - Account Accrued

Contingent Liability % firms

Nº of Firms - Non-accrued

Contingent Liability

% Firms

Mar/06 96 84 87.50% 71 73.96%

Jun/06 102 89 87.25% 48 47.06%

Set/06 104 91 87.50% 49 47.12%

Dec/06 135 132 97.78% 94 69.63%

Mar/07 144 126 87.50% 86 59.72%

Jun/07 148 133 89.86% 93 62.84%

Set/07 149 134 89.93% 93 62.42%

Dec/07 155 144 92.90% 115 74.19%

Mar/08 151 138 91.39% 109 72.19%

Jun/08 152 140 92.11% 111 73.03%

Set/08 152 141 92.76% 112 73.68%

Dec/08 158 147 93.04% 125 79.11%

Average 137 125 91.24% 92 67.15%

34

TABLE 14 Sectors of Companies

Sectors Number % Agriculture 3 1.89

Food and beverages 9 5.66 Trade 7 4.40

Construction 22 13.84 Electronics 2 1.26 Electricity 14 8.81

Finance and Insurance 19 11.95 Industrial Machinery 3 1.89

Mining 2 1.26 Minerals Metallurgy 2 1.26

Other 31 19.50 Paper 3 1.89

Oil and Gas 1 0.63 Chemistry 5 3.14

Steel & Metallurgy 9 5.66 Software and Data 2 1.26 Communications 3 1.89

Textiles 6 3.77 Transport and Services 11 6.92

Vehicles and parts 5 3.14 TOTAL 159 100.00

Documents

New Frederico Nilsen - FLAVIA.pdf · 2012. 6. 2. · Av. Fernando Ferrari, 1358 – Goiabeiras – Vitória, ES, Brazil – 29075-010 [email protected] ou [email protected] Frederico