EARNINGS AND BALANCE SHEET CONSERVATISM IN MALAYSIA : THE EFFECT OF MALAYSIA ' S CONVERGENCE TO INTERNATIONAL FINANCIAL REPORTING STANDARDS ( IFRS )

The objective of this paper is to examine the impact of International Financial Reporting Standards (IFRS) on both earnings and balance sheet conservatism in Malaysia. Earnings conservatism has been used extensively in common-law countries such as the United States (US) and the United Kingdom (UK), which are known to have good financial reporting quality. In contrast balance sheet conservatism is more evident in code-law countries such as Japan and France, where accounting practice is highly influenced by the institutional culture similar to Malaysia such as political connection and family-firms. Since Malaysia uses international accounting standards adopted from common-law countries, we hypothesise that earnings conservatism will increase whereas balance sheet conservatism will decrease after the incorporation of IFRS in the financial reporting environment. Consistent with the hypothesis, we find that earnings conservatism and balance sheet conservatism increase and decrease respectively after Malaysia's convergence to IFRS. The results indicate that Malaysia has successfully adopted IFRS from common-law countries as earning conservatism increase after IFRS convergence.


INTRODUCTION
The issue of financial reporting quality has become more prominent due to recent corporate failures that occurred in developed countries, such as the United States of America, where rules and regulations are more stringent and rigid compared to other countries.The collapse of Enron and WorldCom have drawn the attention of regulators, practitioners, auditors and researchers, both in developed and developing countries, to the issue of financial reporting transparency.These cases have shattered the whole nation, particularly the companies' shareholders, as the financial reporting of the companies were 'decorated' with promising profits, lower debts and high revenues.In financial reporting, the issue of corporate transparency is more important as information that is accurate and relevant is needed by investors in their investment decision making and at the same time to ensure the security of their investments.Regulators are also demanding corporate transparency in order to exert more control on company operations and the behaviour of managers.Levitt (1998) defines quality as financial reporting transparency that signifies the basis of any business. 1He further states that success of capital markets is explicitly dependent on the accounting and disclosure system where high-quality accounting standards will produce financial statements that report events in realtime.High-quality financial reporting ensures no added reserves and deferral of losses, and actual volatility is not smoothed away to portray a false representation of steady and consistent growth.Following the definition of quality as proposed by Levitt (1998), conservatism is regarded as the best measure for financial reporting transparency as this concept requires a higher level of verification in distinguishing between the good and the bad news.In this concept bad news is recognised earlier than good news and therefore earnings reflect bad news more quickly than good news.This means bad news is identified sooner than good news, thus, earnings reflects bad news faster than good ones.Francis, Lafond, Olsson, and Schipper (2004) and Vichitsarawong, Eng and Meek (2010) use conservatism as a measure of financial reporting transparency since conservatism is regarded as a desirable attribute of accounting earnings that can be used in order to avoid unexpected economic downturn and corporate failure.
There are two types of conservatism: Earnings and balance sheet conservatism.Earnings conservatism is defined as the asymmetric response of earnings to economic gains and losses as such losses are recognised immediately in the current period, whereas profit will be reflected in the financial statement gradually over a number of years (Basu, 1997).Meanwhile, balance sheet conservatism is defined as the understatement of book values of net assets and overstatement of liabilities in the balance sheet with respect to market value of the firm (Feltham & Ohlson, 1995).Both types of conservatisms lead to understatement of assets or profit and overstatement of liabilities or losses.However, what makes these two differ is the time of recognition as earnings conservatism is applied when there is news or indication from the stock prices whereas balance sheet conservatism is persistently applied in the balance sheet when it is compared to market value of the firm without any basis or indication.
Due to the basis of recognition, earnings conservatism has been extensively used as a proxy for financial reporting quality compared to balance sheet conservatism and this type of conservatism has been tested on various applications such as litigation, contracting, corporate governance and other regulations.Most of the results provided by the previous studies (Ball, Kothari, & Robin, 2000;Ball, Robin, & Shuang Wu, 2003;Basu, 1997;Bushman & Piotroski, 2006) support that earnings conservatism play an important role in alleviating agency problems and limit the losses from poor investment decisions.In contrast, balance sheet conservatism has been criticised as it provides understatement of book value without any basis or indication and hence facilitates the creation of "cookie-jar reserves" which may lead to earnings management 2 .
Investigating the effects of regulation such as IFRS on both types of conservatism is interesting as earnings conservatism is more apparent in common-law countries such as U.K which are known as having high financial reporting quality whereas balance sheet conservatism is more evident in code-law countries such as Germany and Japan where the main sources of financing are from a system of relationship-oriented capital, ownership structure which is more concentrated, less shareholder activism and a taxation policy that relies more on the lower value of assets (Black & White, 2003;Gassen, Fuelbier, & Sellhorn, 2006;Joos & Lang, 1994;Lara & Mora, 2004).Lara and Mora (2004) argue that in the so-called common-law-based countries such as UK, the degree of balance sheet conservatism will be less pronounced than in the so-called code-law-based countries.Higher financial reporting quality in common law country is recognised as timely loss recognition as compared to code law countries (Ball et al., 2003).The loss recognition is considered as more important as the nature of companies tend to disclose good news as compared to bad news.Malaysia is one of the countries that decided to adopt IFRS by the year 2012.Despite the adoption, Malaysia provides a unique institutional setting.Principally, Malaysia uses International Accounting Standards (IAS), adopted by common-law countries such as the US and UK which have been known as having high quality of financial reporting (Ball et al., 2003).Nevertheless, primarily, accounting practices in Malaysia are highly influenced by its institutional culture similar to the code-law countries' such as political connection and family firms (Ball et al., 2000, Muniandy & Ali, 2012).The CLSA (Credit Lyonnais Securities Asia) Asia Pacific Markets report on the overall of Malaysia's performance in Corporate Governance Watch 2007 reported that there is little confidence in the Malaysian market that independent directors are genuinely independent.Although public enforcement efforts have improved, the report blames the regulators for not having a reputation for treating companies and individuals equally.The CLSA committees make consensus that it is politics that hampers the ability of regulators to do their job properly (CLSA Asia Pacific Markets, 2007).The CLSA report in 2010 once again highlighted this issue as they pointed out that the main area of weaknesses in Malaysian corporate governance is on accountability and the number of independent audit committees.Their report highlight that Malaysia's financial reporting quality is significantly affected by culture (CLSA Asia Pacific Markets, 2010) as Malaysia's score for corporate governance culture dropped 1% from 33% in 2007 to 32% in 2010. 3 When comparing the governance score between political and regulatory and CG culture of 11 capital markets in Asia, the gap between these two scores is greatest in Malaysia which shows that CG culture still poses a threat to the corporate governance practices here even though there is improvement in the enforcement and regulatory system.
Due to unique institutional setting of Malaysia, investigating the effects of regulation such as IFRS on both types of conservatism is appealing.In addition, evidence on the impact of IFRS in Malaysia is rather limited.Wan Ismail, Kamarudin, Van Zijl and Dunstan (2013) provide initial evidence on the effects of IFRS on earnings quality in Malaysia.They found that, based on a proxy for earnings management, IFRS adoption increases the earnings quality in Malaysia.We opted for a similar approach to Wan Ismail et al. (2013).However, instead of examining the effects on accruals quality, we investigate the effects of IFRS on both earnings and balance sheet conservatism.In addition to extending an outlook on earnings quality in Malaysia, our choice of measures of conservatism will provide a support on the role of institutional settings in Malaysia.
We offer the following motivations.First, to our humble knowledge, we initiate the first investigation on balance sheet conservatism in Malaysia.We view balance sheet conservatism as a complimentary tool on earnings conservatism and as proxies for earnings quality.Evidence on conservatism (either earnings or balance sheet) is limited in Malaysia.Mohamed Yunos, Ismail and Smith (2012) provide some evidence on the relationship of various ethnic groups and earnings conservatism and found mixed evidence.The use of a single type of conservatism (i.e.earnings conservatism) may not justify the insignificant findings between ethnicity and conservatism in their study.Therefore, an investigation on the effect of IFRS on different types of conservatism may provide a different picture on Malaysia's institutional setting and its relationship with earnings quality.
We analyse earnings and balance sheet conservatism before and after IFRS.For earnings conservatism we used the model by Basu (1997) whereas for balance sheet conservatism we opted the model by Garcia- Lara and Mora (2004).This study is motivated by Malaysia's convergence to IFRS in 2006.Malaysia has announced its full convergence to IFRS on 1 August 2008 by 2012 but the effects of IFRS convergence with local standards is still largely unexplored.This study differs from other studies on conservatism and IFRS in several ways.First, this study is conducted specifically on Malaysia's institutional settings which could provide better and compelling evidence.Recent studies conducted utilised multiple countries by comparing the effect of IFRS adoption in code and common law countries (Callao, García, Jarne, & Gadea, 2010;Lara, Torres, & Viera, 2008) or in developing countries which have different legal systems and culture from emerging countries like Malaysia (Callao, Jarne, & Laínez, 2007;Karampinis & Hevas, 2011;Piot, Janin, & Dumontier, 2011).Second, this study investigates the effects of IFRS convergence on both types of conservatism; earnings and balance sheet conservatism.The investigation of both types of conservatism will help us to understand the uniqueness of Malaysia's institutional setting and its relation to financial reporting quality.
Our results suggest that earnings conservatism is enhanced after IFRS.We also find that balance sheet conservatism is lower after the convergence.Therefore we provide strong evidence that IFRS is one of the main determinants of conservatism supporting the argument by Watts (2003aWatts ( , 2003b)).In 2005, as support for the IFRS issued by IASB, MASB rename the MASB standards to Financial Reporting Standards (FRS) meant to make it in line with standards issued by IASB except for some minor modifications.In January 2006, all Malaysian firms are required to converge the local reporting standards with IFRS by preparing financial statements according to IFRS.Since this year, Malaysian IFRS standards have been identical to the respective IFRS and IAS.The difference lies in the standards that Malaysia has not adopted.With the convergence, modifications of the standards are made if necessary.Instead of adopting the whole standards as a required standard to prepare financial statements, Malaysia introduced two-tier financial reporting framework whereby the IFRS framework is made mandatory for non-private entities while the private entities can continue using the old MASB standards known as PERS (Private Entity Reporting Standards) framework.Figure 1 shows the milestones for Malaysia's convergence to IFRS.IFRS still provide various mechanisms which indirectly ensure the application of conditional conservatism.For example, recognition of probable liabilities and non-recognition of contingent asset in IAS 37, lower cost or net realisable value for inventories (IAS 2) and impairment of asset in IAS 36.Comparing IFRS with the local GAAP, IFRS introduced more stringent and systematic impairment testing rather than amortisation based on the judgement.The differences of financial reporting standards in the IAS and IFRS that would lead to more conservatism using IFRS are shown in Appendix.

Malaysia's Political Economy and Ownership Structure
Malaysia's political economy is highly influenced by family ownership and political connection (Claessens & Fan, 2002;Jaggi, Leung, & Gul, 2009).This type of political economy has enabled corporate entities in Malaysia to seek capital fund from "insiders" rather than from capital market.Faccio (2007) reports that connected firms enjoy easier access to debt financing from state-controlled banks, even though their situations do not justify additional credit.Johnson and Mitton (2003) found that politically connected public firms in Malaysia had significantly better returns even under the situation of capital control.Therefore, there is less demand of informative financial statements by the public.In addition, being politically connected facilitates firms to get private information and, hence, any information asymmetry between shareholders and managers are settled through "insider communication" rather than through "public disclosure".While compliance with approved accounting standards is mandatory, companies may send voluntary signals about their activities and performance using private information (Mohd Saleh, Che Abdul Rahman, & Hassan, 2009).This situation creates less incentive for the preparers of financial statement to enhance disclosure by adopting IAS.
The influence and dominance of family presence and ownership in Malaysia has been well documented (Claessens & Fan, 2002;Jaggi et al., 2009).According to statistics, which has been presented by South China Morning Post (SCMP), Malaysia has the second highest percentage of family ownership of listed companies in the region after Indonesia (Jaggi et al., 2009).Claessens, Djankov and Lang (2000) found that the presence of family dominance has enabled them to control firms and this represents a large percentage of stock market capitalisation in nine East Asian countries including Malaysia.This situation has led to Type II agency problem, which is the conflict between majority and minority shareholders in which minority shareholders have less power to voice out their dissatisfaction with a firm's poor governance.
Institutional background of Malaysia, which emphasises the Bumiputras privileges, has also leaved it hostile to ethnic Chinese Malaysians. 4 In order to establish their position among the Malays, Chinese businesses have responded in several ways.Chinese Malaysians with mobile capital investments have move away from active politics, engaging in high-risk, short-term speculative ventures and diversifying overseas to strengthen the concession in Malaysia's political economy.Malaysia's richest man, Robert Kuok Hock Nien, for example, has diversified his interests across Asia without any political favouritism.Some of them, for example Vincent Tan Chee Yioun from property and gambling conglomerate Berjaya Group, seeks direct ties with a Malay political party through its official Chinese party in the UMNO coalition, the Malaysian Chinese Association (MCA) (Pepinsky, 2008). 5  Malaysia's convergence to IFRS is seen as a step forward to produce accounting earnings characteristics closer to common-law countries such as the U.K. It is seen as good effort by the regulators to enhance transparency and give awareness to public listed companies and stakeholders on the importance of faithful and truthful financial statements.Dato' Zainal Abidin Putih, the chairman of MASB, in one of his speech on Malaysia's full convergence to IFRS in 2012, commented that since Malaysia has included the provisions of the international standards into its local accounting standards beginning 1978 and these standards have also been used by more than a hundred countries worldwide, it is believed that IFRS will facilitate comparability and increase transparency (PricewaterhouseCoopers, 2009) and thus is expected to prevail over the code law countries' characteristics.

EMPIRICAL PREDICTIONS
The concept of conservatism historically has been one of the most influential accounting principles (Sterling, 1967) and has dominated other accounting principles such as historical cost and realisation concept for centuries (Basu, 1997).Various definitions of conservatism have been proposed by the accounting authority bodies such as IASB and FASB and these definitions have also led to debate on the applications of conservatism as a proxy for financial reporting quality.
In the Statement of Standard Accounting Practice No. 2 (SSAP2): "Disclosure of Accounting Policies" conservatism is described as understatement of assets or profits and overstatement of liabilities or losses (Accounting Standards Board, 1971).Nevertheless, the general authoritative definition to some extent does not properly discuss on what is the appropriate level of conservatism and how it is determined (Hellman, 2008).Hellman (2008) illustrates that previous researchers (such as Schmalenbach [1959] and Sillen and Vasthagen [1962]) are also unclear on how far conservatism should be applied.To some extent, they argue that conservatism can play its role to protect the stakeholders, whereas in other extent conservatism may lead to manipulation and distortion of information.Basu (1997) proposes an operational definition of conservatism, which provides a clearer base of assets or profits and liabilities or losses valuation, compared to the theoretical definition provided by the authority bodies.Basu (1997) defines conservatism as reflecting bad news more quickly than good news.He states that conservatism is applied when there is news by requiring higher degree of verification for recognising good news compared to bad news.Chi, Liu and Wang (2009) state that there are two important reporting features of earnings conservatism, which are asymmetric timeliness in recognition of gains versus losses and, a systematic understatement of net assets.The term "systematic" provides an explanation that the earnings conservatism based on valuation needs a formal indication on when to recognise and how much to recognise.

Balance Sheet Conservatism
Balance sheet conservatism is consistent with unconditional or ex-ante conservatism as this type of conservatism is not dependent on the news and it is recognised before or without receiving any news (Pae, Thornton, & Welker, 2005).Givoly and Hayn (2000) use market-to-book ratio as a proxy for the degree of balance sheet conservatism.This measurement is based on the balance-sheet-oriented definition of conservative accounting suggested by the theoretical framework developed by Feltham and Ohlson (1995).Feltham and Ohlson (1995) define balance sheet conservatism as "the existence of a persistent understatement of the book value figure with respect to market's valuation of the firm".The market-to-book ratio, that is greater than one, indicates conservative accounting and, other things being equal, an increase in the ratio over time suggests an increase in the degree of reporting balance sheet conservatism.

Conservatism and IFRS
Most of prior studies have documented positive effects of IFRS.There are evidence that IFRS can reduce analysts' absolute forecast error (Ashbaugh & Pincus, 2001), have more persistent and more conditionally conservative earnings (Gassen et al., 2006), can cause an increase in market liquidity (Daske, Hail, Leuz, & Verdi, 2008), and a decrease in earnings management (Barth, Landsman, & Lang, 2008;Zéghal, Chtourou, & Sellami, 2011) and an increase in stock price informativeness (Beuselinck, Joos, Khurana, & Meulen, 2009).There are also evidences that IFRS can enhance conservatism.Lara et al. (2008) show in their results that earnings conservatism is more pronounced in common-law-based developed economies and the voluntary use of IASB standards in Europe (prior to 2005) has significantly increased the measures of earnings conservatism in adopting firms.
Previous studies (e.g.Barth et al., 2008;Daske et al., 2008) offer three arguments on why the adoption of IFRS can yield significant capital market benefits.Firstly, its reporting promotes transparency and this improves the quality of financial reporting since the standards are more capital-market-oriented and more comprehensive than local GAAP, particularly when involving disclosures.Secondly, this practice also lowers the amount of reporting discretion in relation to local GAAPs.Thirdly, IFRS reporting has been found to improve financial statement comparisons across firms.Additionally, according to Ewert and Wagenhofer (2005), the tightening accounting standards can reduce the level of earnings management and hence improve financial reporting quality.
IFRS emphasises on fair value principle which involves a process of asset or liability recognition, initial measurement at fair value, re-measurement (again, largely, at fair value) and de-recognition.Based on this principle, losses will be valued at market value, thus, increases conservatism among managers. 6Since the natural behaviour of humans is to be more inclined to disclose good news rather than bad news, recognition of losses is considered as more important in increasing transparency among managers.Hellman (2008) proposes that the consistent understatement of assets characterised by the balance sheet approach is overcame in the IFRS as the new standards temporarily lower down the value of assets and will be reversed later due to change in accounting estimates.Hellman (2008) stresses that conservatism applied in IFRS can reduce the creation of hidden (off-balance) reserves or excessive provisions created by consistent undervaluation of asset in balance sheet conservatism.
It is believed that rules and regulations such as IFRS is one of the determinants of earnings conservatism that can induce firms to protect their shareholders and debt holders.Malaysia's convergence of local standards to IFRS is seen as a step forward by the regulators to enhance transparency and put awareness on public listed companies and the stakeholders on the importance of faithful and truthful of financial statements.Therefore, we hypothesise that: H1: Earnings conservatism increases after IFRS convergence Ball and Shivakumar (2005) state that there are three reasons earnings conservatism is preferable than balance sheet conservatism.First, the timely loss recognition of earnings conservatism provided by Basu's (1997) model provides an important role of conservatism in governance.Secondly, by understating the book value of assets, balance sheet conservatism reduces the opportunities of firms to account value in a conditionally conservative fashion.For example, if firms take into account balance sheet conservatism by immediate expensing assets, it eliminates their opportunity to impair the asset when the market value provides evidence that the value has been reduced.Thirdly, balance sheet conservatism can reduce contracting efficiency as it introduces randomness in decisions based on financial information.
Due to these reasons, this type of conservatism has been subjected to criticisms and is associated with lower financial reporting quality.Ball and Shivakumar (2005) argue that conservatism "seems at best neutral if the bias of understatement is known".If the bias is unknown, then it may lead to inefficiency.Therefore, even though both of the definition of balance sheet and earnings conservatism is understood as lead to bias (higher verification of revenue compared to losses), at least the bias in earnings conservatism is known.Due to these arguments, we hypothesise that: H2: Balance sheet conservatism decreases after IFRS convergence Empirical research have been done by researchers (e.g.Pae et al., 2005;Beaver & Ryan, 2005;García-Lara & Mora, 2004;Givoly, Hayn, & Natarajan, 2007;Roychowdhury & Watts, 2006;Pope & Walker, 2003) to test the relationship between earnings and balance sheet conservatism and they support that there is a negative relationship between earnings and balance sheet conservatism as such balance sheet conservatism pre-empt earnings conservatism.

Sample Selection and Data Collection
Our initial sample consists of 857 firms from Bursa Malaysia listed firms over the period 2004-2009.Financial institutions, insurance and real estate companies are excluded due to different regulatory framework.PN4 companies, companies that change their financial year end and companies with missing data are also excluded from the sample yielding a final sample of 3274 observations for earnings conservatism and 4018 observations for balance sheet conservatism.The differences between the sample of earnings and balance sheet conservatism is mainly attributable to the need of one year observation of price adjusted on pre-sample period to calculate return and also to the fact that excludes all missing values for the price adjusted.The sample is divided into two periods which are pre- IRS (2004 and2005) and post-IFRS (2006 till 2009).The year selection is based on effective date for preparing financial statements based on IFRS for all Malaysian firms which is on 1 January 2006.Data are extracted from Compustat Global.The minimum data required for each firm-year observation are the current year's earnings, the previous fiscal year-end stock price, book value of assets and equity and returns data (Basu, 1997).Following Vichitsarawong et al. (2010) and Ball et al. (2003), accounting variables are deflated by the beginning of period price to control for heteroscedasticity.In addition, serial correlation of period SUR is reported for regressions to correct for heteroscedasticity and general correlation of observations within a cross-section.The 1st and 100th percentiles of each variable are excluded to reduce the effect of outliers.Finally, each firm-year observation with a missing value for any of the variables is excluded.Table 2 shows the definition of variables used in this study.

Model Specification
Earnings conservatism was tested using Basu's (1997) model specification.This model has been tested in Malaysia by Ball et al. (2003) and Vichitsarawong (2007).Basu (1997) uses reverse regression of annual earnings on contemporaneous returns to investigate the relationship between economic income, as measured by stock returns, and accounting income.The model is shown as follows: Where N it is net income before extraordinary items per share of firm i, deflated by beginning of period share price; R it fiscal year continuously compounded return; and DR it is dummy variable, equalling one if R it is negative, and zero otherwise.The earnings variable N it is calculated, as X it /N it P it-1 where X it is net income before extraordinary items for firms i, N it is the adjusted number of shares and P it-1 is the share price.
In this model, stock returns are the independent variable, while earnings are the dependent variable.The coefficient on stock returns β 2 measures the sensitivity of accounting income to positive stock returns (a proxy for economic gains).The coefficient β 3 is the main measurement for earnings conservatism which measures the incremental sensitivity of accounting income to the incorporation of bad news as measured by negative stock returns (a proxy for economic losses).The total sensitivity of accounting income to negative stock returns is measured by (β 2 + β 3 ).
Previously, Givoly and Hayn (2000) used market-to-book ratio to examine the existence of balance sheet conservatism.Nevertheless, this measurement has been criticized as market-to-book ratio can be driven by many other factors (growth options, the possibility of monopoly rents and/or synergies, inflation, etc.).Consequently, Lara et al. (2004) investigate the effects of balance sheet conservatism using a valuation framework provided by Ohlson (1995), which expresses price as a function of both earnings and book value of equity.
Ohlson model (1995) has been widely used in value-relevance study.Despite of that, the model has also been used to capture different properties of accounting such as conservatism.The model illustrates the effect of conservative accounting on the relation between equity value, accounting book value and future earnings (Lundholm, 2010).
This model has been utilised by Garcia-Lara et al. (2004) who investigates the effects of balance sheet conservatism using a valuation framework provided by Feltham and Ohlson (1995), which expresses price as a function of both earnings and book value of equity: Where P it is the share price at the balance sheet date, NI it is the earnings after extraordinary items per share, BV it is the period-end book value of shareholders' equity per share.In the particular model above, the intercept of the undeflated regression of price on earnings and book value is significantly larger, while the book value coefficient is significantly positive if the balance sheet conservatism exists and practiced in Malaysia.The coefficient B 2 is the main measurement for balance sheet conservatism.
In order to test the relationship between the adoption of IFRS and conservatism, both of the conservatism models are employed using dummy year variable for period pre (year 2004 and 2005) and post of IFRS (2006IFRS ( till 2009)).
where all variables as described above, D_IFRS it is dummy variable which takes the value of 1 for the period after IFRS and 0 otherwise.
In model (3), dummy variables DR it capture the intercept and slope effects for the negative return sample.Our focus variable is the coefficient β 7 which measures the relationship between the level of asymmetric timeliness of conservatism and IFRS.It is expected that the coefficient to be positive and significant, showing that earnings conservatism is increased after the adoption of IFRS.
Coefficient β 2 measures the responsiveness of earnings to good news before the IFRS adoption, while the sum of β 2 + β 6 is measuring the responsiveness of earnings to good news after the IFRS adoption.A positive significant coefficient β 6 implies incremental responsiveness of earnings to good news after IFRS adoption.We offer no prediction for the intercept and incremental intercept coefficients β 0 , β 1 DR it , β 4 D_IFRS and β 5 D_IFRSxDR it in measuring earnings conservatism.
In model ( 4), the relationship between balance sheet conservatism is determined by the coefficient β 5 which is the coefficient of the interaction between book value and the dummy year of IFRS.It is expected that the coefficient will be significantly negative, leading to the positive effect of IFRS as it leads to a decrease in the understatement of assets relative to market's valuation.

Descriptive Statistics
Table 3 reports descriptive statistics for all the variables within the periods.For earnings conservatism presented in Panel A of Table 3, the mean (median) value for the dependent variable, which is earnings, and independent variable which is return are 4.427 (0.000) and 1.174 (0.990) respectively.The mean value shows that throughout the period earnings are higher than return which indicates that overall the sample do not shows earnings conservatism.Nevertheless, in term of median, earnings are lower than return which indicates the existence of earnings conservatism.Standard deviation for earnings are also far larger than the standard deviation of return which indicates that throughout the period earnings is not really a function of return.For balance sheet conservatism tabulated in Panel B of Table 3, the mean (median) for book value of shares (BV) is 0.808 (1.000), which is slightly lower than share price (P), which shows a mean of 1.742 (0.910).The result for mean indicates that balance sheet exist among the firms in the sample as the book value is lower than the share price.Nevertheless, the result for median indicates that book value is slightly higher than the share price which indicates that firms are not conservative.

Univariate
We run univariate analysis to test the effect of IFRS on both types of conservatism.Table 4 presents differences in the mean and median values of our measures of measures of conservatism for the periods before and after IFRS.The table shows a significant improvement in some of conservatism variables.The average of return R has significantly increased from 0.981 to 1.239 after IFRS.Nevertheless, the value of earnings is not significantly difference from pre to post IFRS.The result indicates that even though return is significantly increasing, the value of earnings has no significant difference.This result indicates early evidence that IFRS has some effect on earnings conservatism.The average of share price P decreased from 1.895 to 1.691 whereas the average of net income NI shows a slight decrease from 0.205 to 0.175 after IFRS.The average of book value of shareholder's equity per share also decreases after IFRS from 0.879 to 0.784.The average of share price P decreased from 1.895 to 1.691 and it is only significant using Mann-Whitney test.The average book value of shareholder's equity per share (BV) show slight decreases after IFRS which is from 0.879 to 0.784 respectively.As presented in the table, the mean values for the share price are greater than the values for book value of shareholders' equity per share.This provides evidence of the existence of balance sheet conservatism in this country throughout the sample period.The decrease in value is even more after IFRS, providing more evidence that balance sheet conservatism is decreasing after IFRS.
on earnings conservatism is more pronounced in common-law countries compared to code law countries.The results also support previous research (Barth et al., 2008;Daske et al., 2008) which argue that IFRS reporting increases transparency, reduces the amount of reporting discretion and improves comparisons across firms.Note: N = net income before extraordinary items per share of firm i, deflated by beginning of period share price; R = fiscal year continuously compounded return; DR = dummy variable, equalling one if R is negative, and '0' otherwise; D_IFRS = dummy variable which takes the value of 1 for period after IFRS and '0' otherwise.*, ** and ***denote significance at the 10%, 5% and 1% level respectively.Joos (2008) explains that IFRS differs from a country's individual accounting principles in terms of measurement and disclosure of accounting standards.IFRS emphasises on the fair value principle which involves a process of asset or liability recognition, initial measurement at fair value, re-measurement (again, largely, at fair value) and de-recognition.Based on this principle, losses will be valued at market value and hence increase conservatism among managers.In terms of disclosure, IFRS emphasizes on detailed footnote disclosures.It sheds light on off-balance sheet information such as pension assets and liabilities and relevant business information such as segment information, litigation, share-based compensation, related-party transactions.relative to market's valuation is still practised in Malaysia throughout the period.Nevertheless, the coefficient of net income is also consistently positive throughout the model indicating that earnings remain persist.The result indicates that both types of conservatism exist among the firms in the sample.Nevertheless when we interact the book value coefficient with IFRS (BV*D_IFRS), the result indicates that balance sheet conservatism decrease after IFRS and it is significant at 5% level.Supporting H 2 , the results indicate that the consistent understatement of assets characterised by the balance sheet approach is overcame in the IFRS as the new standards temporarily lower down the value of assets and will be reversed later due to changes in accounting estimates (Hellman, 2008).The negative relationship between balance conservatism and IFRS has been a sign that Malaysia has successfully adopted IFRS from common-law countries model even though their institutional structures can still be characterized as a variant of the code law model (Ball et al., 2003).Malaysia is one of the East Asian countries that has strong common-law influence especially in terms of its accounting standards as this country has adopted accounting standards from early British colonial influence to the more recent influence of IFRS.Clarkson, Hanna, Richardson and Thompson (2011) state that the valuation role of earnings is more pronounced in shareholder oriented financial reporting compared to stakeholder-based.Ball et al. (2000) and Soderstrom and Sun (2007) emphasise that legal systems either code or common law countries, play an important role in influencing accounting standards.This is because, in common law countries, accounting standards are set by private sector bodies and, therefore, the purpose of standard setting is to satisfy the informational needs of investors.Meanwhile, in code law countries, standards are influenced by governments and other important stakeholders.Therefore, the role of accounting is to divide the interest between those parties.Bae, Tan and Welker (2008) report that there is greater divergence between a country's individual accounting principle and IFRS in code and common law legal system.

Earnings conservatism
Because many other studies have criticised Basu's (1997) measure as having econometric biases (e.g., Dietrich, Muller, & Riedl, 2007;Patatoukas & Thomas, 2011), we measure conditional conservatism beyond Khan and Watts' (2009) modified conservatism measure based on Basu's (1997) work to ensure the robustness of our findings.The measure we use is based on the Ball and Shivakumar (2005) accruals-based loss recognition model.We modify this measure using the methodology that Khan and Watts (2009) adopted to estimate a firm-year measure of conservatism: where ACC it is the total accruals in year t, deflated by the year t-1 market value of equity; CFO it is the cash flow from operations in year t, deflated by the year t-1 market value of equity and DC it is a dummy variable that is equal to 1 if CFO it is negative and is 0 otherwise.The coefficient of the interaction term between DC and CFO (β 3 ) measures the conditional conservatism.If economic losses are recognised in a more timely manner than gains, then β 3 will be greater than 0. The result using Khan and Watts (2009) adopted measure of earnings conservatism in Table 7 is consistent with Basu's (1997) model.The interaction DC and CFO is significantly negative which indicates that firms in the sample are not conditionally conservative.Nevertheless, when we interact the conservatism with IFRS, the result is significantly positive indicating that IFRS enhance earnings conservatism.Previously, Givoly and Hayn (2000) used market-to-book ratio to examine the existence of balance sheet conservatism.Nevertheless, this measurement has been criticised as market-to-book ratio can be driven by many other factors (growth options, the possibility of monopoly rents and/or synergies, inflation, etc.).Despite of the criticism, using market-to-book ratio (MTB) as a proxy for the degree of balance sheet conservatism is sensible, though it is still subject to argument.Therefore we provide the result for balance sheet conservatism using MTB in Table 8 as a measure of balance sheet conservatism.Firms are conservative when MTB is larger as firms understate the book value in relevant to market value.The result produced in Table 8 is same with the previous one and in fact, it is more significantly negative which indicates that balance sheet conservatism decrease after IFRS.

CONCLUSION
Balance sheet conservatism has been used extensively in prior research parallel to earnings conservatism as one of the proxy for financial reporting quality due to its informational and behavioural aspects.Nevertheless this type of conservatism is more evident in code law country such as Germany and Japan where the main sources of financing are from system of relationship oriented capital, ownership structure which is more concentrated, less shareholder activism and taxation policy which relies more on the lower value of assets.In contrast, earnings conservatism is more apparent in common law countries where investors' protection is very high.
This study finds strong evidence that earnings conservatism is enhanced after IFRS, whereas balance sheet conservatism is reduced after the IFRS convergence.Therefore we provide strong evidence that regulation is one of the main determinants of conservatism supporting the argument by Watts (2003aWatts ( , 2003b)).The results indicate that Malaysia has successfully adopted IFRS from common-law countries models even though their institutional structures are still can be characterised as a variant of the code law model resulting in a decrease in balance sheet conservatism.The results, therefore, provide a motivation for the regulators in this country on the importance of regulation enforcement.It provides an empirical evidence that the role of regulators are very important to always review our accounting and corporate governance system in order to ensure that we can provide a safe and comfortable market to the investors.

NOTES
1. Arthur Levitt is the former Securities Exchange Commissions (SEC) in the US. 2. Cookie jar reserves are an accounting practice in which a company uses generous reserves from good years against losses that might be incurred in bad years.3. CLSA (Credit Lyonnais Securities Asia) is a brokerage and investment groups focused on institutional brokerage, investment banking and asset management for corporate and institutional clients around the world.4. In Article 153 in the Federal Constitution of Malaysia, Bumiputras is defined as "Malay" and "aborigine", "natives" of Sarawak and "natives" of Sabah. 5. Vincent Tan Chee Yioun is also one of the richest men in Malaysia.
6. Fair value principle is defined as "the amount for which an asset could be exchanged, or a liability settled, between knowledgeable, willing parties in an arm's length transaction" (Riahi-Belkoui, 2004).7. Consistent with extant literature on conservatism, a number of interaction terms are used in this study.These interaction terms could give rise to potential multicollinearity issues.In some cases, the variance inflation factors are found to be higher than the preferred threshold of 10.

IAS 36
For fixed asset, impairment is calculated by comparing undiscounted cash flows to the carrying amount of the asset group.Therefore if undiscounted cash flow is lower than the carrying amount, then there is impairment being charged.
For fixed asset, impairment is charged by comparing the carrying amount with recoverable amount.Under IFRS impairment is assessed on the basis of recoverable amount, which is calculated as either fair value less costs to sell or value in use (discounted cash flows).
If impairment is indicated, assets are written down to the higher recoverable amount.
Therefore, impairment charges will likely occur sooner than under IAS.For example, assume a store's undiscounted cash flow exceeds the asset-carrying amount but value in use (i.e., based on discounted cash flows) is less than the asset-carrying amount.No impairment charge would be recorded under IAS but impairment would be recorded under IFRS.

IAS 36
An item of property, plant and equipment that qualifies for recognition, as an asset shall be measured at its cost.The cost of an item of property, plant and equipment is the cash price equivalent at the recognition date.
Under IFRS, an entity may elect to value property, plant and equipment (PP&E) using either the cost or revaluation model.Under the revaluation model, an entire class of PP&E is revalued at fair value regularly.Revaluation increases are credited to equity.Revaluation losses are charged first against any revaluation surplus in equity related to the specific asset, and any excess charged to income.
Estimates of useful life and residual value and the method of depreciation are reviewed at least annually.

Table 1
Data distribution of sample firms

Table 3
Descriptive statistics Note: SD = Standard Deviation

Table 5
Earnings conservatism and IFRS

Table 6
reports the result of for the relationship between balance sheet conservatism and IFRS.The intercept is significantly positive throughout the models.Consistent with earnings conservatism model in Table5, we present model 1 till 6 to show one by one the effect of balance sheet conservatism measured by the book value coefficient.Model 1 in Table6indicates the basic model of balance sheet conservatism.The book value coefficient is significantly positive throughout the model indicating the existence of balance sheet conservatism among the firms in the sample.The result indicates that understatements of assets

Table 6
Balance sheet conservatism and IFRS

Table 7
(Ball & Shivakumar, 2005) IFRS(Ball & Shivakumar, 2005)ACC it = total accruals in year t, deflated by the year t-1 market value of equity; CFO it = cash flow from operations in year t, deflated by the year t-1 market value of equity; DC it = dummy variable that is equal to 1 if CFO it is negative and is 0 otherwise.

Table 8
Balance sheet conservatism and IFRS (Using MTB) Note: P = share price at the balance sheet date; NI = earnings after extraordinary items per share; MTB = marketto-book ratio; D_IFRS = dummy variable which takes the value of 1 for period after IFRS and '0' otherwise.*, ** and ***denote significance at the 10%, 5% and 1% level respectively.