The Effect of Deferred Taxes on Firm Market Value: Evidence from Hong Kong

Transcription

1 The Effect of Deferred Taxes on Firm Market Value: Evidence from Hong Kong BY GAO Fan Accounting Concentration JIANG Wei Accounting Concentration An Honors Degree Project Submitted to the School of Business in Partial Fulfillment of the Graduation Requirement for the Degree of Bachelor of Business Administration (Honors) Hong Kong Baptist University Hong Kong April 2012 Supervisor: Dr. H. K. Daniel Ho

2 Acknowledgement We would like to use this opportunity to express our deepest gratitude to our honors degree project supervisor, Dr. H. K. Daniel Ho for his valuable and insightful advice and guidance on our project. We are grateful to him for his time and patience and his care and kindness has encouraged and inspired us through our learning process. We would also like to thank our friends for their support and encouragement throughout our research process. We sincerely appreciate their kindness and care. We are faithfully gratitude to everyone who has helped and supported us with our project. Last but not least, we would like to thank our beloved School of Business for providing us with this opportunity to conduct an honors degree project. We not only obtained valuable experience of academic research, but also learned how to cooperate, coordinate and balance through the process of our project. GAO Fan JIANG Wei

3 Abstract This study examines empirically the effect of deferred taxes recognized under Hong Kong Accounting Standards (HKAS) 12 on firm market value in Hong Kong stock market. Our empirical model is based on Feltham and Ohlson (1995) Model. Our sample is based on 120 selected companies listed in the Hong Kong Stock Exchange over 2005 to We document a negative association between deferred tax liabilities (DTL) and firm market value, and a positive association between deferred tax assets (DTA) and firm market value. There is no significant relation between unrecognized deferred tax assets (UDTA) and market value of a firm. Key word: Deferred tax liabilities (DTL), Deferred tax assets (DTA), Unrecognized deferred tax assets (UDTA), HKAS Introduction Accounting for deferred taxes has always been a controversial issue in many nations around the world and it seems that Hong Kong is no exception. While the accounting treatment for current taxes - the unpaid income tax to the tax authorities in current and prior periods - is relatively straightforward, the recognition and measurement of deferred taxes are more subjective. Same with other assets and liabilities, inherently, deferred tax assets and liabilities are what the reporting entity expects to recover or settle in the future. In short, current tax is a past story, whereas deferred taxes deal with uncertainties about the future. The subjectiveness influences investors perception of deferred taxes, therefore, we are interested in evaluating how the market views deferred taxes. There are two approaches commonly used for accounting of deferred taxes: the income statement approach and the balance sheet approach. 1.1 Statement of Problems There are two components of International Financial Reporting Standards (IFRS): - International Financial Reporting Standards (IFRS) standards issued after International Accounting Standards (IAS) standards issued before

4 Starting in 2005, Hong Kong Financial Reporting Standards (HKFRS) are identical to International Financial Reporting Standards (IFRS). The IAS 12: Income Taxes sets guidelines for accounting of deferred taxes and the HKAS 12: Income Taxes 1 is the equivalent component of Hong Kong Accounting Standards (HKAS). The income statement approach focuses on the timing difference between the accounting profit and the taxable profit that originates in one period and reverses in later ones. The income statement approach was the approach taken internationally until the 1990s. The balance sheet approach focuses on the temporary difference between the carrying amount of an asset or a liability and its tax base. The tax base of an asset is the amount that will be deductible for tax purposes against any taxable economic benefits that will flow to an entity when it recovers the carrying amount of the asset. If those economic benefits will not be taxable, the tax base of the asset is equal to its carrying amount (HKAS 12 para. 7); the tax base of a liability is its carrying amount, less any amount that will be deductible for tax purposes in respect of that liability in future periods. In the case of revenue which is received in advance, the tax base of the resulting liability is its carrying amount, less any amount of the revenue that will not be taxable in future periods (HKAS 12 para. 8).The balance sheet approach is the approach currently required by the International Financial Reporting Standards (IFRS). The temporary difference is a more comprehensive concept than that of the timing difference since it includes all items defined under the timing difference concept and adds a number of additional items, e.g. the difference arising from revaluation of assets acquired in a business combination. Under the balance sheet approach under Hong Kong Accounting Standards 12 (HKAS 12) (Figure 1), the temporary difference between the carrying amount of an asset or a liability and its tax bases gives rise to a deferred tax 2. The temporary differences can be either taxable or deductible. Taxable temporary differences lead to deferred tax liabilities (DTL) (HKAS 12 para. 15) and deductible temporary differences lead to deferred tax assets (DTA) (HKAS 12 para. 27). 1 HKAS 12 takes effect from 1 January The most recent amendment is made on December HKAS 12 is based on International Accounting Standard 12 (IAS 12) issued by International Accounting Standard Board (IASB) without major textual differences (Ng, 2009). 2 Examples of temporary difference include accelerated tax depreciation of property, plant and equipment, revenue (expenses) taxed (relieved) on a cash basis, fair value adjustments in a business combination and revaluations (Clark, 1997). 3

5 Figure 1. Sources of Deferred Taxes Temporary difference Deferred tax liabilities Source of deferred taxes Deferred tax assets Tax loss/ tax credit Unrecognized deferred tax assets *If there is not probable future profit to utilize the temporary deductible temporary difference, it will fall into unrecognized deferred tax assets category. This relationship is not shown in the graph, because the major part of unrecognized deferred tax assets come from tax loss or tax credits. (According to our sample) Calculation: Carrying amounts of assets or liabilities Tax bases of assets or liabilities Taxable or deductible temporary differences Taxable or deductible temporary differences Tax rates Deferred tax liabilities or assets Besides deductible temporary differences, deferred tax assets also arise from unused tax losses that tax law allows to be carried forward. Calculation: Unused tax losses Tax rates Deferred tax assets 4

6 Recognizing deferred tax assets is probably the most complex and subjective area in financial reporting regulations (Petree et al., 1995). According to HKAS 12, para. 36, deferred tax assets can be recognized if there are probable future taxable profits to utilize the carry-forward tax benefits, otherwise the deferred tax assets will fall into the unrecognized category and disclosed in notes to financial statements. The recognition of deferred tax assets involves two subjective measures: the expectation of future taxable income and tax planning strategies (Visvanathan, 1998). Visvanathan (1998) and Phillips et al. (2003) pointed out the possible manipulation of the recognition criteria to manage earnings. Comiskey et al. (1994) bring to creditors the concern of realizability of deferred tax assets. Regarding the recognition criteria, the US GAAP (SFAS no. 109) allows full recognition but uses a valuation allowance to reduce asset to the amount that is more likely than not to be recognized. There is a considerable amount of literature on the evaluation of the valuation allowance in the US. HKAS 12 adopts a similar criterion with more likely than not : Amounts are recognized to the extent that it is probable they will be realized. Petree et al. (1995) suggests that it is easier for companies with a history of profits than those with a history of losses to justify its future profits. In this study, we examine the relation between firm market value and deferred taxes. Our research scope includes deferred tax liabilities (DTL), deferred tax assets (DTA) and unrecognized deferred tax assets (UDTA). Besides looking into the investors perceptions of these accounts, we also evaluate the usefulness of the accounting standards from a regulatory perspective. Significance of the relationship will indicate whether the reported deferred taxes provide useful information to investors. 3 Theoretically, the test is supported by Feltham and Ohlson (1995) which models the relation between a firm s market value and accounting data. We narrow the accounting data to deferred taxes because deferred tax assets and liabilities are unique assets and liabilities in investors perception, thus an interesting area to investigate. According to Chang et al. (2009), deferred tax assets and liabilities are different from other assets and liabilities because they are not always perceived as real financial assets or liabilities by the market. There are disagreements among 3 Usefulness of accounting information is defined in Framework issued by HKICPA as the extent to which financial information help existing and potential investors, lenders and other creditors make decisions about the reporting entity on the basis of information in its financial report. Fundamental qualitative characteristics of useful financial information are relevance and faithful representation. (Framework 2010) 5

7 previous studies about the perception of deferred taxes: some documented significant association between deferred taxes and firm market value [e.g. Givoly and Hayn (1992), Amir et al. (1997) and Anderson (1997)] while others got insignificant results [e.g. Chaney and Jeter (1988) and Deflieses (1991)]. 1.2 Statement of Objectives The major objective of the study is to determine how the Hong Kong stock market perceives deferred taxes: - Do deferred tax liabilities mean future tax payments and are negatively associated with firm market value? - Do deferred tax assets, whether recognized or not, indicate tax savings in the future and are positively associated with firm market value? - Do deferred taxes reported under Hong Kong Accounting Standards (HKAS 12) provide useful financial data to the investors in the market? After all, the study is worth doing for at least three related reasons. First, it evaluates the accounting policy concerning income taxes in Hong Kong and thus provides useful data for accounting policy makers. Second, the number of tax accounting literature on deferred taxes in Hong Kong is far from satisfactory. There are only some recent studies on tax compliance [e.g. Chan et al. (2000)] and policy formations [e.g. Schiffer (1991)]. Our study can contribute to the limited tax accounting literature in Hong Kong. Finally, few current empirical studies have researched into the balance sheet approach of income taxes required under IFRS. 4 Our study can also provide insights for evaluating the usefulness of incomes tax policies subscribed in IFRS by IASB. The remainder of this paper is organized as follows. The second section reviews prior theoretical literature and empirical studies. The hypothesis is developed in the third section. The forth section is about the methodology and the empirical model. The fifth section discusses the sample 4 There are several studies on the balance sheet approach under Statement of Financial Accounting Standards No. 109 (SFAS No. 109) in an American context [e.g. Amir (1999), Ayers (1998)]. 6

8 selection and data. The sixth section is composed of the discussions on the results. The seventh section is our conclusion. 7

9 2. Literature Review Prior research mainly focuses on three aspects to examine the perception and valuation of deferred taxes, including: (1) Deferred tax liabilities (DTL); (2) Deferred tax assets (DTA); and (3) Unrecognized deferred tax assets (UDTA). 2.1 Deferred Tax Liabilities According to HKAS 12 (revised in December 2010), a balance sheet method is used to determine deferred tax liabilities. Deferred tax liabilities (DTL) are the amounts of income taxes payable in future periods in respect of taxable temporary differences (HKAS 12 para. 5). Generally speaking, if the carrying amount of an asset is higher than its tax base, the temporary difference is a taxable temporary difference; if the carrying amount of a liability is lower than its tax base, the temporary difference is a taxable temporary difference. Accounting for assets: Carrying amounts Tax bases of of assets assets Accounting for liabilities: Carrying amounts Tax bases of of liabilities liabilities Taxable temporary differences Taxable temporary differences A lot of work has been done on research of the people s perception of deferred taxes and prior studies present different opinions on investors perception of deferred taxes liabilities in different markets. As argued by Lasman and Weil (1978), future payments may never be made for the liabilities recognized so there is little relationship between the deferred tax liabilities and future tax payment. For example, for the deferred taxes resulting from different depreciation bases, as long as a company is operating continuously or growing, the reversing differences are generally offset by other originating differences which are equal or even larger than previous ones. As a result, the timing differences will not be recurred and the recognized liabilities will not be settled in the foreseeable future. Others argued that deferred tax liabilities are insufficiently reliable and supported the findings above. Deflieses (1991), Chaney and Jeter (1988) and Chaney and Jeter (1989) claimed that the relationship between the deferred tax liabilities and future tax payments 8

10 is not apparent since there is uncertainty whether those liabilities will be paid back in the foreseeable future. Therefore, as a significant portion of the deferred tax liabilities are unlikely to be settled with future tax payments, deferred tax liabilities should not be considered as liabilities of a company or relevant in its valuation. On the other hand, there is empirical evidence from studies based on the US and Australian markets suggesting strong relationship between deferred tax liabilities and the valuation of a firm. Sansing (1998) showed that deferred tax liabilities have value and should be related to firm valuation. Amir et al. (2001) claimed that deferred tax liabilities are more accurate in valuation when they are discounted to net present value. Amir et al. (1997), Ayres (1998), Beaver and Dukes (1972) and Givoly and Hayn (1992) concluded that deferred tax liabilities are actually considered as financial liabilities by the investors of the US market. Besides, Beaver and Dukes (1972) and Givoly and Hayn (1992) documented that the deferred tax liabilities recognized under the income statement method are negatively associated with security returns. In addition, Chang et al. (2009) also found that the Australian market perceives deferred tax liabilities as liabilities, representing tax costs in the future. Therefore, it seems that deferred tax liabilities are perceived similarly to the other liabilities recognized on the balance sheet of a firm. 2.2 Deferred Tax Assets According to HKAS 12 (revised in December 2010), a balance sheet method is used to calculate deferred tax assets. Deferred tax assets (DTA) are the amounts of income taxes recoverable in future periods in respect of: (a) deductible temporary differences; (b) the carry-forward of unused tax losses; and (c) the carry-forward of unused tax credits 5. For the temporary differences, if the carrying amount of an asset is lower than its tax base, the temporary difference is a deductible temporary difference; if the carrying amount of a liability is higher than its tax base, the temporary difference is a deductible temporary difference. 5 A tax credit is a sum deducted from the total amount a taxpayer owes. A tax credit may be granted for various types of taxes, such as an income tax and property tax. It may be granted in recognition of taxes already paid, as a subsidy, or to encourage investment or other behaviors. 9

11 Accounting for assets: Carrying amounts Tax bases of of assets assets Accounting for liabilities: Carrying amounts Tax bases of of liabilities liabilities Deductible temporary differences Deductible temporary differences For the unused tax losses and unused tax credits, a deferred tax asset shall be recognized for the carry-forward of unused tax losses and unused tax credits to the extent that it is probable that future taxable profit will be available against which the unused tax losses and unused tax credits can be utilized (HKAS 12 para. 34). Prior studies have different views on deferred tax assets as they do on deferred tax liabilities. According to Amir et al. (1997), deferred tax assets arising from tax losses carried forward may be interpreted as signals of future tax losses rather than future tax savings because those firms with carry-forward tax losses are more likely to make losses in the future since they have already experienced losses before. They found evidence that investors value earnings and book values of firms with carry-forward tax losses less than earnings and book values of firms without carryforward tax losses. Therefore, the market seems to consider deferred tax assets as signals for future losses and value deferred tax assets negatively in respect of firm value. However, empirical evidence from different markets suggests a measurement perspective of deferred tax assets. De Waegenaere et al. (2003) argued that losses carried forward should be considered as assets because those losses can provide future tax savings to the company. Zeng (2003) claimed that deferred tax assets arising from tax losses carried forward can enhance value of firms in Canadian markets. Besides, a study of Amir et al. (1999) on the US markets found that there is a positive relation between deferred tax assets arising from losses carried forward and the share prices, suggesting that the investors value such deferred taxes as assets. Another Australian study of Chang et al. (2009) found that Australian investors valuate deferred tax assets as assets and there is a positive relation between firm value and deferred tax assets. Therefore, it seems that deferred tax assets are positively associated with firm valuation because they represent future tax savings. 10

12 2.3 Unrecognized Deferred Tax Assets The unrecognized deferred tax assets (UDTA) are the unused tax losses and tax credits in the notes of the financial statements of the company. The criteria for recognizing deferred tax assets arising from the carry-forward of unused tax losses and tax credits are the same as the criteria for recognizing deferred tax assets arising from deductible temporary differences (HKAS 12 para. 35). Prior studies made mixed comments on unrecognized deferred tax assets. On the one hand, the unused tax losses and credits can be carried forward with certain limitation and provide tax savings against future taxable incomes. So they should be evaluated positively. On the other hand, the existence of unused tax losses and credits is strong evidence that future taxable profit may not be available because when an entity has a history of recent losses, the entity can recognize a deferred tax asset arising from unused tax losses or tax credits only to the extent that the entity has sufficient taxable temporary differences or there is convincing evidence that sufficient taxable profit will be available against which the unused tax losses or unused tax credits can be utilized by the entity. The supporting evidence is required for disclosure in such circumstances (HKAS 12 para. 35). Guenther and Sansing (2004) predicted that unrecognized deferred taxes have intrinsic values and are relevant in valuation. Lynn et al. (2008) found evidence in support of those predictions and suggested that unrecognized deferred taxes provide incremental information to investors and are value-relevant. Another Australian study, Chang et al. (2009), documented a mix conclusion using different models on the relation between market value and unrecognized deferred tax assets disclosed in the notes of financial statements and a negative relation dominates in their conclusion. 11

13 3. Hypothesis Development This study examines the perceptions of investors in the Hong Kong stock market on deferred tax liabilities (DTL), deferred tax assets (DTA) and unrecognized deferred tax assets (UDTA). We want to know whether they are considered as real liabilities and assets by the market as other liabilities and assets appearing on the financial statements. In addition, this study examines the influence on market value of a firm by its corresponding deferred tax liabilities (DTL), deferred tax assets (DTA) and unrecognized deferred tax assets (UDTA) in ways investors evaluate these elements during their valuation process. As discussed above, prior studies propose different opinions on deferred tax liabilities. Some argued that there is little relationship between deferred tax liabilities and future tax payments (e.g. Lasman and Weil, 1978) while some supported the negative association between deferred tax liabilities and the market value of a firm by empirical evidence (e.g. Chang et al., 2009). In our study, an empirical test based on Hong Kong stock market will be conducted and Hong Kong stock market is mature as the stock market in the United States and Australia. So we prefer a similar perception of a market supported by empirical evidence from stock markets in United States and Australia and thus we hypothesize that investors in Hong Kong security markets consider deferred tax liabilities recognized under the balance sheet method according to HKAS 12 as real liabilities similar to other liabilities. Since these liabilities will decrease value of a firm, we predict that there is a negative relation between deferred tax liabilities and the firm market value. H1: Ceteris paribus, deferred tax liabilities are negatively associated with firm market value. As to the deferred tax assets, prior studies cannot come to an agreement on the perception of the market, either. Some claimed that deferred tax assets arising from tax losses carried forward can be interpreted as bad signals of possible loss in the future since losses have been made recently (Amir et al., 1997). However, some others thought those tax losses carried forward can be used together with deductible differences to offset taxable incomes and provide tax savings and they found some empirical evidence to support their opinions (Zeng, 2003). In our study, same as the consideration for deferred tax liabilities, we are going to conduct an empirical test on the perception of the Hong Kong stock market. So we prefer the perception supported by empirical 12

14 evidence and we hypothesize that investors in Hong Kong security markets will consider deferred tax assets recognized under the balance sheet method according to HKAS 12 as assets representing future tax savings. Since these assets will add value to a firm, we predict that there is a positive relation between deferred tax assets and the market value of firms. H2: Ceteris paribus, deferred tax assets are positively associated with firm market value. For the unrecognized deferred tax assets, prior studies provide mixed views on the perception of investors in the market. Some believed that the unused tax losses and tax credits can be used to offset future taxable incomes and provide tax savings while some others thought that the unused tax losses and tax credits put suspicion on the potential profit-making possibility of the company because a deferred tax asset shall be recognized for the carry-forward of unused tax losses and tax credits only to the extent that it is probable that future taxable profit will be available against which the unused tax losses and unused tax credits can be utilized (HKAS 12 para. 34). Since not all the tax losses and tax credits are used, the probable estimated profit is not desirable. In our study, we try to consider the nature of unrecognized deferred tax assets. From one point, the unrecognized portion of the deferred tax assets may be valued by the market as an asset of the firm. Most of these unrecognized deferred tax assets arise from tax losses which can be carried forward. Therefore, the market may view the tax losses to be an asset when it returns to profit. On the other hand, investors of the market may also view the unrecognized deferred tax assets as negative signal about the firm s ability to generate future profits. Therefore, a negative relation of unrecognized deferred tax assets and value of firms may exist. Given the uncertainty of the relation between unrecognized deferred tax assets and firm value drawn from prior studies, the perception of the market depends more on the recent performance of the firms listed in the market and the entire market sentiment. In other words, as long as the investors of the market expect the firms will generate enough profits for the unused tax losses and credits to offset, the unused tax losses and credits can add value to the firm and the investors in the market will percept them positively. Otherwise, these unused deferred tax assets will be considered as the limitation of the firm s probable profit and the investors will percept them negatively. So it depends on the recent performance of the firms and the market sentiment and we think either a positive association or a negative association may dominate. Therefore, we 13

15 predict that there is, non-directional, either a positive or a negative relation between unrecognized deferred tax assets and firm value. H3: Ceteris paribus, unrecognized deferred tax assets are either positively or negatively associated with firm market value. Summary of Hypothesis Development Table 1 Hypotheses Hypothesis 1: Deferred tax liabilities Hypothesis 2: Deferred tax assets Hypothesis 3: Unrecognized deferred tax assets Expected relation Discussion The market view deferred tax liabilities as financial liabilities, although the relation may be tempered by recurring items like depreciation. The market view recognized deferred tax assets as financial assets, although the recognition of tax loss may be viewed as a signal for future loss. Positive relation: Market views unrecognized deferred tax assets as future tax savings, though it does not meet recognition criteria at the moment. Negative relation: Market views unrecognized deferred tax assets as a signal for future losses. 14

16 4. Research Methodology The theoretical background of our study comes from Feltham and Ohlson s model, the F&O (1995) Model 6. Generally, the model relates a firm s market value with its accounting data (Feltham and Ohlson, 1995). The firm s market value is expressed as a function of the components of financial statements (i.e. incomes statement, balance sheet, and statement of changes in shareholder s equity) concerning operating and financial activities. To be specific, the model suggests that the market value of the firm equals the book value of shareholders equity plus unrecorded goodwill. Unrecorded goodwill captures the discrepancy between market value and book value. The discrepancy results from the measurements of operating accounting earnings and the use of accounting conventions for accruals (Feltham and Ohlson, 1995). Given unbiased accounting (versus conservative accounting), the unrecorded goodwill should be zero, i.e. market value equals book value. Furthermore, under the assumption of clean surplus accounting, unrecorded goodwill is measured by the present value of expected future abnormal earnings defined as actual earnings minus the cost of capital times the net operating asset at the beginning of the period 7. An empirical version of F&O (1995) Model is presented in Equation (1): (1) Where P it is the market value of firm i at time t, NBV it equals the net book value of shareholders equity per share, and AE it stands for present value of abnormal earnings per share. Error term (µ it ) represents the random shocks that affect share price. If unbiased accounting is used to 6 This model is first introduced in the paper of Feltham and Ohlson (1995) and is then widely known as the F&O (1995) Model. It provides a framework to estimate firm value using clean surplus theory. 7 Here we demonstrate F&O (1995) Model using the following equations: Basic F & O Model: ( ) ( ). Under clear surplus accounting ( clean surplus ensures that all changes in shareholders equity that do not result from transactions with shareholders themselves, such as dividends, are reflected in the incomes statement): ( ) ( ), i.e. Unrecorded goodwill is calculated as actual earnings minus expected earnings (or minimal required earnings by investors). 15

17 measure the book value, and if all available information is reflected in expected earnings, α 0 would be to zero and α 1 and α 2 would equal to one. Having the F&O (1995) Model, Amir et al. (1997) in their paper, argued that deferred taxes is important in the F&O (1995) Model valuation. They thereby developed an equation which is consistent with the F&O (1995) Model to incorporate deferred taxes in the valuation. Their argument runs as follows: The accounting treatment of deferred taxes may affect the calculation of unrecorded goodwill, because even though deferred tax may not explicitly and directly generate abnormal earnings, its presence on the balance sheet affects investors perception of future abnormal earnings. As a result, reported deferred taxes are an important factor to determine the market value. Following this argument, Amir et al. (1997) give a modified model presented by Equation (2): In Equation (2) 8, the market value of a firm i at time t is P it. NOA it is the net operating assets per share, adjusted for net deferred tax assets. NFA it is the net financial assets per share for firm i at time t. AE it is the abnormal operating earnings per share for current period. DT it is the net deferred tax assets per share for firm i at time t. Subject to data availability and other statistical considerations, we modify Amir s equation to obtain our empirical model. First, we combine net operating assets (NOA) and net financial assets (NFA) to get net book value (NBV). We use the aggregation of the two factors because net operating assets (NOA) and net financial assets (NFA) have large correlation which may lead to multi-collinearity problems in regression. The same treatment is adopted in Amir et al. (1997 and 1999) and Chang et al. (2009). Ayers (1998) in his empirical study also dropped the separation of net operating assets and net financial assets, instead used two variables - book value of assets and book value of liabilities. Our principle is to select variables that is most closely consistent with the original F&O (1995) Model, thus, we use net book value (NBV) as in Equation (1). Second, we use current year actual earnings (EARN) as a proxy for abnormal operating earnings 8 See Lynn et al. (2008) Incremental Value Relevance of Unrecognized Deferred Taxes: Evidence from the United Kingdom (The Journal of American Accounting Association, 30(2)), p. 113 Research Design section for the detailed evolvement of the regression equations. (2) 16

18 (AE), since abnormal operating earnings (AE) is not observable. The calculation of abnormal earnings (i.e. actual earnings minus cost of capital times beginning of year net operating assets) depends on the cost of capital of each particular firm. This figure is different among firms and is volatile to measurement error. Amir et al. (1999) calculates the cost of capital using a capital asset pricing model (CAPM) 9. However, we choose a more simple approach to use current earning as a proxy of market expectations, because this approach is well supported by past empirical studies. Barth and Clinch (2009), Chang et al. (2009) and Lynn et al. (2008) used the similar approach in the empirical estimation of the F&O (1995) Model, and obtained satisfactory empirical results. Third, we decompose net deferred taxes to deferred tax liabilities (DTL), deferred tax assets (DTA) and unrecognized deferred tax assets (UDTA), because the original design of our study is to see the effect of each of three factors on market value individually. The first two components - deferred tax liabilities (DTL) and deferred tax assets (DTA) - are those which meet the recognition criteria and are recognized on balance sheets. The third component - unrecognized deferred tax assets (UDTA) - is classified as unrecognized portion and is disclosed in the notes to financial statements. Besides these modifications, we add four control variables to our model. First, we add growth factor (Gr), which is the percentage increase in tangible assets, since growth in assets (Gr) is related to deferred tax liabilities (DTL), especially those arising from the timing difference from depreciation (Chaney and Jeter, 1994). If we omit it, the coefficient on the deferred tax liability (DTL) may be overstated. Also, growth, showing the potential of growing opportunities, is also one of the most important factors that affect investors investment decisions, thus explains stock price fluctuations. Second, we add a loss dummy variable (LOSS) to indicate whether the firm is making positive or negative accounting profit in current year, because this affects investors perceptions of deferred taxes: if the firm reports a profit, investors is likely to be more optimistic about the future tax savings, thus place a higher perceived value for deferred tax assets, and vice versa. In Chang et al. (2009), they even include the loss dummy variable and interact it with other financial variables in their model. They suggest that some coefficients of these financial variables is different in a loss and profit firm context. Third, we add a year-dummy variable 9 In Amir et al. (1999) paper, they first run a market model regression to estimate each firm s systematic risk factor (β i ). Then they apply β i to the CAPM formula (Cost of equity capital = R f + β i Risk premium). 17

19 (Ycrisis) to differentiate years after the global financial crisis from others, because after financial crisis investors may be pessimistic about the overall economy. The pessimism among investors may depress the stock price. Finally, we add a dummy variable (Mat) to control the effect of Materials and Energy industries because our sample includes more companies from these industries compared with the overall industry distribution of Hong Kong Stock Market (see sample descriptions in Sample and Variables section). Moreover, the nature of assets and liabilities of firms in these industries and the industry regulatory requirements may be different compared with other firms. For example, few tax jurisdictions allow companies to claim tax deductions on acquired oil and gas production properties. Therefore, the tax base of such assets would be small. In such case, large deferred tax liabilities would be recognized (PwC, Financial reporting in the oil and gas industry 10 ). The empirical equation is as follows: (3) Where is the share price 11 of the firm i at the trading day when the firm releases its financial reports for the fiscal year end t. We use the price at release time instead of that at balance sheet year-end date, because the financial data disclosed to the market will more likely be reflected in the stock price at the day the financial reports are released to the market than at the year-end date. NBV is the net book value of the firm s equity at the balance sheet date, adjusted for deferred taxes. EARN is the actual earnings for current year, represented by net profit for the year disclosed in the firm s profit or loss statement. DTL represents the deferred tax liabilities, DTA is the deferred tax assets, and UDTA is the amount of unrecognized deferred tax assets. Gr represents the growth of the firm, and is calculated as the growth of tangible assets. Ycrisis takes value 1 for the year 2008, 2009 and 2010, and zero for the previous years. LOSS is 1 if firms that suffer loss in the financial year. Mat takes 1 for firms from Materials or Energy industries, and zero for firms in other industries. 10 This report is available in the official website of PricewaterhouseCoopers We use market value and share/stock price interchangeably in this paper. 18

20 With regression Equation (3), our hypothesis can now be stated in terms of the expected signs of the coefficient estimates. H1: β 3 =0, against the alternative β 3 <0. H2: β 4 =0, against the alternative β 4 >0. H3: β 5 =0, against the alternative β 5 0. In the next section, we describe how we collect and analyze data to test these three hypotheses. 19

21 5. Sample and Variables Data are collected for the years , altogether six years. First, we randomly select 300 firms out of 1,315 listed equity securities on Hong Kong Stock Exchange, Mainboard 12. Then, we obtain financial statement data (i.e. net income, total assets, tangible assets, total liabilities, deferred tax asset and liabilities) as well as price and outstanding shares data from DataStream 13. We delete 69 firms which do not have sufficient data from the database (Table 2). In addition, we exclude 105 firms which offset deferred tax assets and deferred tax liabilities in the statement of financial position 14. We further exclude six banks from the sample because of their highly regulated nature and the different nature of their assets and liabilities. The finalized sample has 120 sample firms. The full list of company code and name is presented in Appendix I. Table 2 Summary of Sample Firms Selection Process Population 1,315 Sample by random selection 300 Less firms with: Missing financial statement or price data (62) Missing deferred taxes data (7) Deferred tax liabilities and deferred tax assets are offset on balance sheet (105) Banks (6) Total sample firms firms are randomly selected from the 1,315 firms listed on Hong Kong Stock Exchange Mainboard. After removing firms with missing data and firms that do not meet our selection criteria, we include 120 sample companies for further analysis. 12 List of securities is obtained from the Security Trading Information section of HKEx website: (The list is dated on 19 March 2012). 13 We get access to DataStream database with the subscription by School of Business, Hong Kong Baptist University, Hong Kong. 14 According to HKAS 12, an entity may offset deferred tax assets and deferred tax liabilities, if they relate to income taxes levied by the same taxation authority on the same taxable entity (para. 74). A number of firms choose to offset deferred tax assets and deferred tax liabilities in the statement of financial position, but separate them in notes to financial statements. These firms are not included in the final sample. 20

22 The industry classification of the firms is presented in Table 3. The classification follows Hang Seng Industry Classification System (HSICS) 15. The sample covers all industries, and the industry distribution of the sample is to a large extent consistent with the population distribution. There are two major deviations. First, the sample firms are heavy in Energy (00) and Materials (05) industry 16.6% of the 120 sample firms are in Energy and Materials industry, while the industry distribution of Hong Kong stock market shows these two industries account for only 9.1% of total listed companies in Second, in our sample, there are only four companies from the financials industry (3.3%, versus 9.3% of the population), and the four companies are all from 503- other financial sector under 50 - Financials industry. The reason is that we deliberately exclude banks (501) and insurance companies (502) from the empirical study, because of the difference in regulations on operations and the structures of financial statements. Table 3 Industry Classification HSICS Code Industry Number of firms Percentage (%) 00 Energy Materials Industrial Goods Consumer Goods Services Telecommunications Utilities Financials Properties & Construction Information Technology Conglomerates Total % A total of 120 samples are included over the periods. Industry classification is based on Hang Seng Industry Classification System (HSICS). 15 HKEx start to use Hang Seng Industry Classification System (HSICS) from 31 December We get individual company classifications from Company/Securities Profile on HKEx website: (Accessed 21 March 2012). 16 Data are obtained from Appendix 2 of HKEx News Release (11 December 2007) HKEx adopts Hang Seng Industry Classification System : Industry Distribution of Hong Kong Stock Market. 21

23 Within the six-year observations of each firm, data are partly missing for some years. After deleting such firm-year observations, we include in our final sample 563 firm-year observations. These samples have complete financial statement data, deferred tax data and price data. Table 4 is the summary of these 563 firm-year observations. The sample contains 495 profit firm-year observations, over 88% of total observations (Panel C). Table 4 Panel A:Pooled sample 120 firms 720 Less firms with: Missing price and outstanding shares data (40) Missing deferred tax data (45) Incomplete deferred tax footnotes (72) Total sample firm-year observations 563 Panel B: Observations by year Total observations 563 Panel C: Observations by Profit/Loss year Profit year 495 Loss year 68 Total observations

24 We define each of the variables closely with the F&O (1995) Model [Equation (1)] and Amir s regression equation [Equation (2)], with some minor adjustments for the specifications of Hong Kong companies. We use unadjusted share price on the trading day when the firm release financial statements, which is usually three or four months after the fiscal year end. Unadjusted means the share price has not been adjusted for right issues and stock splits. It is regarded as the actual or raw price of the stock. Current earnings (EARN) is calculated as the net income available to common shareholders after income from discontinued operations, unusual gains and losses and extraordinary items. Net book value (NBV), deferred tax liabilities (DTL) and deferred tax assets (DTA) are calculated from data obtained through DataStream. We deduct net deferred tax liabilities from the net book value (NBV) to get net book value adjusted for deferred taxes. We manually collect unrecognized deferred tax assets (UDTA) which are disclosed in the Notes to the accounts in the Annual Report 17. The descriptions of raw data are in Appendix II. We deflate all these financial statement variables by the number of shares outstanding at balance sheet date, so that these independent variables are expressed on a per-share basis 18. All numbers are measured in terms of Hong Kong dollars. For firms that use other currencies as their measurement unit, we apply the exchange rate at the balance sheet date to convert them into Hong Kong dollars. 17 The full-text financial statements are obtained from HKEx Advance Search engine: 18 The number of shares outstanding includes all issued shares of the company, minus treasury shares. Some mainland-based companies have both H shares and A shares. They are both included in deflator. 23

25 6. Results and Discussion 6.1 Descriptive Statistics We run a linear multiple-regression using Ordinary Least Squares (OLS) method. After generating the original descriptive statistics (Appendix III), we find out that some variables have extreme values. This may violate the basic assumption of OLS estimation 19. For example, the price (P) has a large range from to 122. Growth (Gr) has extreme value which may be due to sample errors. Therefore, we classify as outliers the values of each variable whose distance from its mean is more than three times its standard deviation. And then we removed them. We removed a total of 30 observations from the 563 sample. We provide descriptive statistics for variables in Table 5. The price distribution shows that although the highest price in the sample is over 50, more than three quarters of the observations have price (P) lower than This also indicates that the price is concentrated at low levels; but a small number of firms have very high prices. Net book value (NBV) has similar distribution pattern with the price (P). A large number of firms in the sample have low net book value (thirdquartile level is slightly higher than the mean); a small number of firms have very high net book value. NVB has an average of 3.20 with a standard deviation of The mean of current earnings (EARN) over one year (0.35) is around 10.9% of the net book value (NBV). This shows that on average listed companies in Hong Kong obtain a rate of return on equity of around 10.9% per year. The lower quartile of current earnings (EARN) is a positive number. It means less than a quarter of the firms are making a loss. This is also shown by the descriptive statistics of the loss dummy variable (LOSS): A majority of Hong Kong listed companies make profit in current years. The average of deferred tax liabilities (DTL), deferred tax assets (DTA), and unrecognized deferred tax assets (UDTA) are 0.17, 0.03 and 0.19, respectively. Deferred tax liabilities (DTL) are much larger than deferred tax assets (DTA). And unrecognized part of deferred tax assets (UDTA) is much bigger than its recognized counterparts (DTA). 19 An important least square assumption in multiple-regression is that large outliers are unlikely. Another way to state the assumption is that the dependent variable and repressors have finite Kurtosis. 24

26 Table 5 Descriptive Statistics (n=533) Mean Minimum Maximum Std. Deviation 25% 50% 75% P NBV EARN DTL DTA UDTA Gr LOSS Ycrisis Mat Sample: Final sample contains 563 firm-year observations. 30 samples are specified as outliers. Variable definitions: P = closing share price of the trading day when the firm released Annual Reports; NBV = the book value of equity before deferred tax assets and liabilities at year end, deflated by the number of year-end shares outstanding; EARN = reported net profit for common shareholders, deflated by the number of year-end shares outstanding; DTL = the book value of deferred tax liabilities at year end, deflated by the number of year-end shares outstanding; DTA = the book value of deferred tax assets at year end, deflated by the number of year-end shares outstanding; UDTA = the balance of the unrecognized deferred tax assets at year end, deflated by the number of year-end shares outstanding; Gr = percentage growth in total tangible assets over the year; LOSS = dummy variable indicating that the firm is making a loss in the year, i.e. LOSS=1 if EARN<0, otherwise LOSS= 0; Ycrisis = dummy variable indicating that the observation is for the corresponding year from 2008 to 2010, i.e. Ycrisis is 1 if the observation refers to 2008, 2009 and 2010, and 0 otherwise. Mat = dummy variable indicating that the corresponding firm is in Materials industry or Energy industry, according to HSICS classifications. 25

27 This result differs from the observations in Chang s study of Australian firms under the income statement method (AASB1020). In Chang et al. (2009), they documented very close average deferred tax assets (DTA) and unrecognized deferred tax assets (UDTA) (0.045 and 0.047) for profit firm years 20. They also documented slightly higher average deferred tax liabilities (DTL) (0.061). This may provide us with some insights about the difference of the two accounting treatment of deferred taxes: There is a stricter requirement for companies to recognize deferred tax assets under balance sheet method (HKAS 12) than under income statement method (AASB 1020 before 2005). The deferred tax liabilities (DTL) account for about 5% of net asset of an average firm (NBV). The deferred tax asset (DTA) is less than 1% of the net book value (NVB). The descriptive statistics of the control variables shows that on average, Hong Kong listed companies have 20% high growth per year. What s more, in our sample firm-year, 57% are from years after the financial crisis, 12% suffer loss, and 11% are from Energy or Materials industries. Table 6 presents the variables correlation matrix. As expected, price is positively correlated with net book value (NBV) and current earnings (EARN) (Pearson = and 0.674). Deferred tax liabilities (DTL), deferred tax assets (DTA) both have positive correlation with the price (P). Unrecognized deferred tax assets (UDTA) have positive correlation with other variables, but insignificant correlation with the price (P). Growth (Gr) does not have significant correlation with either of other variables. The correlation between net book value (NBV) and current earnings (EARN) (0.733), deferred tax liabilities (DTL) and net income (NI) (0.608), and deferred tax liabilities (DTL) and net book value (NBV) (0.765) are high. It may result in the problem of multi-collinearity. Therefore, we will further check Variance Influence Factor (VIF 21 ) for each independent variable in later parts. 20 See Chang et al. (2009) Market s perception of deferred tax accruals [Accounting and Finance 49(1)] p. 661, Table 3: Descriptive Statistics table. 21 VIF stands for variance influence factor, which is an indicator of multi-collinearity. 26

28 Table 6 Correlation Matrix (n=533) P NBV EARN DTL DTA UDTA Gr P ** ** ** ** NBV ** ** ** ** EARN ** ** * DTL ** ** DTA UDTA Gr 1 * Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed). The correlations are pair-wise Pearson correlations. P = closing share price of the trading day when the firm released Annual Reports; NBV = the book value of equity before deferred tax assets and liabilities at year end, deflated by the number of year-end shares outstanding; EARN = reported net profit for common shareholders, deflated by the number of year-end shares outstanding; DTL = the book value of deferred tax liabilities at year end, deflated by the number of year-end shares outstanding; DTA = the book value of deferred tax assets at year end, deflated by the number of year-end shares outstanding; UDTA = the balance of the unrecognized deferred tax assets at year end, deflated by the number of year-end shares outstanding; Gr = percentage growth in total tangible assets over the year; 6.2 Deferred Tax Liabilities and Assets Table 7 presents the results of testing the three hypotheses. The hypotheses are tested by the direction and significance of the coefficient on deferred tax liabilities (DTL), deferred tax assets (DTA) and unrecognized deferred tax assets (UDTA). The coefficient of net book value (NBV) and current earnings (EARN) is 0.4 and 7.3 respectively. The coefficient estimates are 27

29 statistically significant, and is comparable with the estimates in the study by Chang et al. (2009) of Australian firms. The control variables, i.e. growth (Gr), the dummy variable indicating year of financial crisis (Ycrisis), the dummy variable indicating year of loss (LOSS), the dummy variable indicating the firm of Materials industry or Energy industry (Mat), are not statistically significant (p > 0.05). Growth (Gr) is only significant if we extend the confidence level to 10%. The variance influence factor (VIF) for each variable is smaller than 5, so we conclude that the problem of multi-collinearity is not serious. We find strong support for our predictions in Hypothesis 1. The coefficient of deferred tax liabilities (DTL) is negative and statistically significant (-3.45) at the 1% level. The result is consistent with prior studies, e.g. Amir et al. (1997), Ayres (1998), Beaver and Dukes (1972) and Givoly and Hayn (1992). Furthermore, our results extend their findings to the Hong Kong market: for Hong Kong listed companies, investors view deferred tax liabilities as real financial liabilities. Deferred tax liabilities are therefore a liability item that will be recurred and settled in the foreseeable future. They indicate future taxable amounts. The result does not support the opposite point of view, which argues that deferred tax liabilities arising from recurring items are unlikely to reverse and are not treated by investors the same way like other liabilities, e.g. Lasman and Weil (1978). The result also suggests that the accounting treatment of deferred tax liabilities under HKAS 12 provides reliable information for investors in Hong Kong. Consistent with our expectations, there is a positive and statistically significant coefficient (p value is close to zero) for recognized deferred tax assets (DTA). It is consistent with the studies in Canadian and US markets - Zeng (2003) and Amir et al. (1999). This result confirms our prediction that deferred tax assets are viewed by the market to indicate tax savings in the future, and therefore, enhance the firm value. The magnitude of the coefficient on deferred tax assets (DTA) is relatively high at This suggests that investors are quite confident that future deductible amount and loss/credit carry-forward can be utilized by firms in the future. This finding confirms that the recognition of deferred tax assets provide useful information to the market. 22 In an Australia context, the coefficient of DTA documented by Chang et al. is only 5.8 for Australian firms under AASA 1020, less than half of our coefficient estimate in Hong Kong. 28

30 OLS Regression Results Table 7 Collinearity Coefficients Statistics B Std. Error t Sig. VIF (Constant) 1.758** NBV 0.429** EARN 7.299** DTL ** DTA ** UDTA Gr LOSS Ycrisis Mat n 533 R Adjusted R Std. Error of Estimate F-statistic ** * indicates significance at 0.05 level (2-tailed). ** indicated significance at 0.01 level (2-tailed). P is the closing share price at the time when the firm release Annual Reports. NBV is the book value of equity before deferred tax assets and liabilities at year end. EARN is the reported net profit for common shareholders. DTL is the book value of deferred tax liabilities at year end. DTA is the book value of deferred tax assets at year end. UDTA is the balance of the unrecognized deferred tax assets at year end. All of the financial statement variables are deflated by the number of shares outstanding at year end. Gr is the percentage growth in total tangible assets over the year. LOSS is a dummy variable capturing all loss firms, i.e. LOSS=1 if EARN<0, otherwise LOSS= 0. Ycrisis is a dummy variable indicating that the observation is for the corresponding year from 2008 to 2010, i.e. Ycrisis is 1 if the observation refers to 2008, 2009 and 2010, and 0 otherwise. Mat is dummy variable indicating that the corresponding firm is in Materials industry or Energy industry, according to HSICS classifications. 29

31 Contrary to our expectations, the result of unrecognized deferred tax asset (UDTA) does not support our hypothesis. The coefficient of unrecognized deferred tax asset (UDTA) is statistically insignificant (p-value = 0.295). Unrecognized deferred tax assets are not associated with the market value of a firm (P). This means that the inability of the accruals to meet the recognition criteria does not reliably provide a signal to the market about the future profitability of the firm (possible negative association), neither does this disclosure suggest future tax savings for investors (possible positive association). The result is not consistent with prior studies which are in favor of the value-relevance of unrecognized deferred tax assets, e.g. Guenther and Sansing (2004) and Lynn et al. (2008). We try to give some possible interpretations to the above findings about deferred tax assets. There are two basic arguments about deferred tax assets (recognized and unrecognized): If an measurement standpoint is taken, deferred tax assets suggest future tax savings, i.e. investors expect firms to have sufficient future profits to utilize such tax savings; yet, from an informative point of view, deferred tax assets provide signals of future losses of the company. We have several observations from the descriptive statistics and the regression results: Unrecognized deferred tax assets (UDTA) have much larger mean than deferred tax assets (DTA); recognized deferred tax assets (DTA) has high positive coefficient, while the unrecognized part is generally not associated with the market value. One possible explanation to these observations is that the recognition criteria for deferred tax assets are perceived to be very strict, therefore only a small part of deferred tax assets is recognized. The argument runs as follows: The investors of Hong Kong listed companies, knowing that the recognition of deferred tax assets is very strict, will be confident about the future profitability of the firm, even if the firm is not able to prove that it has sufficiently large amount of future profit to utilize the deductible temporary difference, tax losses and tax credits and then recognize deferred tax assets. Therefore, the informative standpoint (signaling-loss effect) fails to apply in the Hong Kong market. Another possible explanation is that investors are optimistic about the current operating status and future prospects of Hong Kong listed companies, given the small number of loss companies (mean of LOSS = 0.1) and the large average growth (mean of Gr = 20%). The investors are confident that companies will generate sufficient profit to utilize the tax losses and credits, so that they value highly of the recognized deferred tax assets. This explains the high coefficient of deferred tax assets (DTA). 30

32 Alternatively, these observations may be driven by a valuation model misspecification. Thus, we run another return model in the next section. 6.3 Sensitivity Tests We perform several sensitivity tests to our regression results. In the previous design, we aggregate net operating assets (NOA) and net financial assets (NFA) to get net book value (NBV) as an independent variable. We run another regression using net operating assets (NOA) and net financial assets (NFA) in order to be more consistent with F&O (1995) Model. The regression result is in Appendix IV. We find no significant difference on the result of other variables. We also test if the result is sensitive to the choice of deflator. Since we use the price at the time when firms release their financial statements as the independent variable, the most proper deflator should be the shares outstanding at the release time. But the number of shares outstanding disclosed in the annual reports may not be equal to the number of shares at the release time of the annual reports. Therefore, as a sensitivity test, we also use the shares outstanding at the most recent financial reports from the release time to deflate the financial variables. The result is not qualitatively different from the result in Table 6. We also conduct a sensitivity test to estimate a regression of returns of stocks on deferred taxes because the return model (returns regressed on changes of financial variables) is less prone to econometric problems than the price model (stock price regressed on financial variables) (Kothari and Zimmerman, 1995). In our model to relate price to deferred taxes, we may be exposed to some statistical problems related to the assumptions of linear regression: possible departure from the normality assumption, i.e. the residuals should be normally distributed about the predicted value (see Appendix V for the normal probability plot of regression residuals) 23. We estimate the empirical return model using the following regression equation: (4) 23 If residuals deviate from the straight line, then the normality assumption may be violated. 31

33 RETURN it is the change in market capitalization in year t, deflated by total shares outstanding. To simplify the calculation, we calculate RETURN it as (P t P t-1 )., are the per share changes in net book value, net profit, deferred tax liabilities, deferred tax assets and unrecognized deferred tax assets, respectively. Control variables are defined in the same manner as in the price model. We use original data to estimate the coefficients using this return model. However, the R 2 is too low (0.04), so that we cannot draw inference from this model (see Model Summary in Appendix VI). Kothari and Zimmerman (1995) suggest while the return model better predicts the coefficients of variables, the low R 2 of the return studies might lead researchers to draw incorrect inferences. However, the return model is worth considering as a complement of the price model when we further our study using another dataset. At last, our study has limitations regarding our sample selection criteria. We only include in our sample the firms who separate deferred tax assets and deferred tax liabilities in the Statement of Financial Position. However, this is not the whole picture, because some firms offset the two items in the Statement of Financial Position, but disclose the separated amount in the Note to accounts. Excluding these firms may lead to bias towards the population. This is a limitation of our current study, and we will research into it in further studies. 32

34 7. Conclusion In this study, we evaluate the informative value of the accounting treatment under HKAS 12 Income Taxes. We examine how Hong Kong market perceive deferred taxes reported by Hong Kong listed companies under balance sheet method. Our model is based on Feltham and Ohlson (1995). Our sample includes 120 companies listed on Hong Kong Stock Exchange Mainboard. We document a negative relationship between the firm market value and the reported deferred tax liabilities. This shows investors view deferred tax liabilities as financial liabilities that indicate future tax payment. The argument that it will not be recurred or settled in the future, thus not a real liability, does not stand in the Hong Kong market. In our sample, a positive relationship exist between the market value and deferred tax assets. This means investors view deferred tax assets as tax savings in the future. However, we do not see a significant relationship between the firm market value and unrecognized deferred tax assets. We can infer from the empirical results that although a significant part of recognized and unrecognized deferred tax assets comes from tax losses, deferred tax assets (both recognized and unrecognized ones) do not signal future loss in the market. In conclusion, our result suggests that, generally, the balance sheet method of HKAS 12 provide useful information to the Hong Kong stock market. Further research need to be done in order to assess whether the recognition criteria of reported deferred tax assets is overly strict that it hampers the reported deferred taxes from providing more relevant information to the market. 33

35 Appendix I Sample Company List 10 HANG LUNG GROUP 419 MEDIACHINA CORP 874 GUANGZHOU PHAR 12 HENDERSON LAND 420 FOUNTAIN SET 883 CNOOC 13 HUTCHISON 432 PCPD 888 ROADSHOW 18 ORIENTAL PRESS 458 TRISTATE HOLD 980 LIANHUA 28 TIAN AN 485 STARLIGHT INT'L 987 CH RENEW EN INV 56 ALLIED PPT (HK) 487 SUCCESSUNIVERSE 992 LENOVO GROUP 63 WINFOONG INT'L 489 DONGFENG GROUP 1001 VAN SHUNG CHONG 66 MTR CORPORATION 494 LI & FUNG 1049 NET2GATHER 86 SUN HUNG KAI CO 496 KASEN 1052 YUEXIUTRANSPORT 97 HENDERSON INV 500 DVN (HOLDINGS) 1088 CHINA SHENHUA 100 CLEAR MEDIA 517 COSCO INTL HOLD 1097 I-CABLE COMM 113 DICKSON CONCEPT 529 SIS INT'L 1100 MAINLAND HOLD 119 POLY HK INV 532 WKK INTL (HOLD) 1138 CHINA SHIP DEV 124 KINGWAY BREW 533 GOLDLION HOLD 1161 WATER OASIS GP 129 ASIA STANDARD 535 FRASERS PPT 1193 CHINA RES GAS 130 MOISELLE INT'L 566 APOLLO SOLAR EN 1199 COSCO PACIFIC 148 KINGBOARD CHEM 592 BOSSINI INT'L 1207 SRE GROUP 179 JOHNSON ELEC H 635 PLAYMATES 1221 SINO HOTELS 210 DAPHNE INT'L 656 FOSUN INTL 1313 CHINARES CEMENT 229 RAYMOND IND 659 NWS HOLDINGS 1812 CHENMING PAPER 258 TOMSON GROUP 665 HAITONG INT'L 1818 ZHAOJIN MINING 289 WING ON CO 678 GENTING HK 1878 SOUTHGOBI-S 308 CHINA TRAVEL HK 679 ASIA TELE-NET 1880 BELLE INT'L 311 LUEN THAI 688 CHINA OVERSEAS 2313 SHENZHOU INTL 315 SMARTONE TELE 697 SHOUGANG INT'L 2331 LI NING 323 MAANSHAN IRON 710 VARITRONIX INTL 2338 WEICHAI POWER 330 ESPRIT HOLDINGS 715 HUTCH HARB RING 2358 MITSUMARU HOLD 336 HUABAO INTL 724 SINO-TECH INT'L 2618 TCL COMM 360 NEW FOCUS AUTO 725 PERENNIAL INT'L 2626 HNC 363 SHANGHAI IND H 728 CHINA TELECOM 2727 SH ELECTRIC 371 BJ ENT WATER 731 SAMSON PAPER 2883 CHINA OILFIELD 374 FOUR SEAS MER 732 TRULY INT'L 2899 ZIJIN MINING 375 YGM TRADING 738 LE SAUNDA HOLD 3318 CHINA FLAVORS 384 CHINA GAS HOLD 751 SKYWORTHDIGITAL 3337 ANTON OILFIELD 386 SINOPEC CORP 754 HOPSON DEV HOLD 3339 LONKING 391 MEI AH ENTER 806 VALUE PARTNERS 3382 TIANJINPORT DEV 393 GLORIOUS SUN 827 KO YO GROUP 3398 CHINA TING 395 SINO DRAGON 857 PETROCHINA 3818 CHINA DONGXIANG 406 YAU LEE HOLD 865 FIRST MOBILE 3833 XINXIN MINING 411 LAM SOON (HK) 869 PLAYMATES TOYS 3938 SAMLING GLOBAL 34

36 Appendix II Raw Data Descriptions Variables Descriptions DataStream Code P Common shares outstanding NBV Closing price which has not been historically adjusted for bonus and rights issues. This figure therefore represents actual or raw prices as recorded on the day. The number of shares outstanding at the company's year end. It is the difference between issued shares and treasury shares. Net book value equals total assets minus total liabilities, excluding deferred taxes. UP WC05301 WC02999 (total assets) WC03351(total liabilities) EARN Net income available to common shareholders WC01751 DTL Deferred tax liabilities WC18183 (deferred taxescredit) DTA Deferred tax assets WC18184 (deferred taxesdebit) UDTA Unrecognized deferred tax assets from Notes N/A Gr Growth equals the percentage change in total tangible assets WC02649 (total intangible other assets) 35

37 Appendix III Original Descriptive Statistics N Minimum Maximum Mean Std. Deviation Mean+3s.d. P u.b.= NBV u.b.= EARN u.b.= DTL u.b.= %DTA u.b.= UDTA u.b.= Gr u.b.= u.b. stands for upper bound. Values larger than this amount are classified as outliers and are removed from the sample. 36

38 Appendix IV Regression Results Sensitivity Analysis Unstandardized Coefficients Standardized Coefficients Collinearity Statistics B Std. Error Beta t Sig. Tolerance VIF (Constant) EARN 7.448** NFA.665** NOA.436** DTL ** DTA ** UDTA Gr R Adjusted R Std. Error of Estimate

39 Appendix V 38