Ea

La

b

5c

KFDFTTT

1

titv

1h

Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108

Contents lists available at ScienceDirect

Journal of International Accounting,Auditing and Taxation

arnings, institutional investors, tax avoidance,nd firm value: Evidence from Taiwan

ing-Ling Changa,1, Fujen Daniel Hsiaob,∗, Yann-Ching Tsai c,2

Department of Accounting, Ming-Chuan University, 250 Sec. 5, Zhong Shan N. Road, Taipei 111, TaiwanDepartment of Accounting, Labovitz School of Business and Economics, University of Minnesota Duluth, 1318 Kirby Drive, Duluth, MN5812, United StatesDepartment of Accounting, National Taiwan University, No. 1, Sec. 4, Roosevelt Road, Taipei 106, Taiwan

a r t i c l e i n f o

eywords:oreign institutional investoromestic institutional investorirm valueransfer pricing audit regulationax avoidanceax havens

a b s t r a c t

This study examines the valuation of earnings from China and Taiwan by foreign anddomestic institutional investors across a sample of Taiwanese electronics firms. We furthercompare the valuation of firm earnings reported in tax havens and non-tax havens, andwhether these firms have changed tax avoidance activities since 2004 when the Taiwanesegovernment enacted stricter auditing of transfer pricing regulation.

Our findings show that both operating income from the home country and investmentincome are positively associated with firm value. Operating income from China, however, isnot significantly related to firm value when institutional ownership of the firm exceeds fiftypercent. This result indicates that operating income is valued differently, depending on thelocation from which the income was generated. Non-operating income enhances firm valueregardless of the revenue source. We also report that foreign institutional investors favoroperating income from domestic and investment sources over earnings generated fromnon-domestic sources and other non-operating income. Furthermore, our results suggestthat firms rearrange reported profits from subsidiaries located in tax havens to affiliatesin other countries following the transfer pricing audit guide Taiwan implemented in 2004.Results also indicate firms may have been shifting profits to other low-tax-rate countries,or to countries which do not require firms to pay taxes, even if they are not doing businessin that country.

© 2013 Published by Elsevier Inc.

. Introduction

Prior studies document that foreign institutional investment affects firm value more than domestic investment, par-

icularly in emerging markets. Gillan and Starks (2003, p. 15) argue that “foreign institutional investment has become anmportant influence in emerging equity markets, and firms may be motivated to improve their corporate governance in ordero attract foreign capital.” Internalization theory assumes foreign direct investment (hereafter FDI) increases a firm’s marketalue (Doukas & Travlos, 1988; Fatemi, 1984; Kim, 2000). For example, Bodnar and Weintrop (1997) find that firm earnings

∗ Corresponding author. Tel.: +1 218 726 7454; fax: +1 218 726 8510.E-mail addresses: llchang@mail.mcu.edu.tw (L.-L. Chang), fhsiao@d.umn.edu (F.D. Hsiao), yanntsai@ntu.edu.tw (Y.-C. Tsai).

1 +886 2 28824564.2 +886 2 33661118.

061-9518/$ – see front matter © 2013 Published by Elsevier Inc.ttp://dx.doi.org/10.1016/j.intaccaudtax.2013.07.001

L.-L. Chang et al. / Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108 99

association coefficients are more constant and have a greater impact on incremental foreign earnings than domestic earningsdue to relative growth opportunities overseas. Garrod and Rees (1998) document that earnings and net assets are valuedmore highly for multinational firms than for purely domestic firms because of international expansion opportunities. Usinga sample of U.S. multinationals, Callen, Hope, and Segal (2005), however, assert that domestic earnings more accuratelyexplain the variance of unexpected returns than do foreign earnings. In light of inconclusive results from evaluating U.S.multinational companies (hereafter MNCs), it remains unclear whether firm valuation will vary due to its composition ofdomestic or foreign institutional investment, or domestic earnings and foreign earnings from emerging markets.

MNCs have incentives to use transfer pricing arrangements to shift income to lower tax countries or tax havens, thusmaximizing after-tax profits (Sikka & Willmott, 2010). This type of tax avoidance by MNCs has received considerable atten-tion. Tax authorities have responded with stricter audits on transfer pricing arrangements for MNCs, particularly involvingtransactions with affiliates or related parties in tax havens. This audit response raises the question of whether MNCs are ableto continue tax avoidance activity.

This study uses high-tech electronics firms in Taiwan to examine the variation in firm value associated with institutionalinvestment within emerging markets. A unique feature of the Taiwanese sample is that almost all Taiwanese electronicsfirms have business operations in both China and Taiwan, which attracts both foreign and domestic institutional investment.In particular, many of these firms, with multiple geographic business operations, derive significant foreign earnings fromoperations in tax havens. Thus transfer pricing arrangements for tax avoidance are a common strategy to maximize after-tax profits. Additionally, the setting of these Taiwanese firms, unique from other emerging markets, may provide significantinsights into firm valuation.

The first objective of this study is to examine differences in valuation of earnings by foreign and domestic institutionalinvestors, when comparing operating income from China and operating income from Taiwan. The second objective is toexamine the differences in firm value when operating income is reported from tax havens or non-tax havens. Finally, athird objective is to investigate whether firms have changed reported earnings due to tax avoidance activities subsequentto Taiwanese stricter audit regulation on transfer pricing effective in 2004.

Using a sample of Taiwanese electronics firms that includes 1346 firm-years of observation spanning a six-year periodfrom 2000 to 2005, we find evidence supporting an incremental increase of firm value related to each component of earningstied to significant foreign institutional ownership (more than twenty percent). However, domestic operating income andinvestment income enhance firm values only for firms with significant domestic institutional ownership (more than twentypercent) and with majority institutional ownership (more than fifty percent). The findings suggest that the value of oper-ating income differs depending on where the operating income originated, but non-operating income enhances firm valueregardless of revenue source. Therefore, operating income from Taiwan has a greater impact on firm value than operatingincome from other locations, including from China. The results also indicate that firms which disclose operating incomefrom Chinese operations are valued differently by foreign and domestic institutional investors.

Results indicate that operating earnings from tax havens decreased after the change in Taiwanese transfer pricing auditregulation in 2004. This finding suggests that Taiwanese electronics firms have shifted income from tax havens to othercountries through affiliated transactions in response to new government audit guidelines. Thus, our study of Taiwanesemultinational firms adds to the extant literature on the impact of institutional ownership on firm value of MNCs, andcontributes to accounting practices within the global business environment. Secondly, this paper contributes to a greaterunderstanding of firms’ transfer pricing arrangements to shift income for tax avoidance. Lastly, our empirical findings supporthost countries’ accounting policies that include stricter government audit regulations on firms’ transfer pricing and suggestguidelines for MNCs operating in emerging markets.

The remainder of this paper is organized as follows: Section 2 presents recent literature for review and develops hypothe-ses. Section 3 includes sample selection and research design. Section 4 presents empirical results and analysis. Section 5concludes with a brief summary.

2. Literature review and hypotheses development

In emerging markets, foreign investment has become an important channel to raise capital. In general, foreign institutionalinvestors play an important role in the market. They spend considerable time analyzing the fundamentals of their investmentbefore allocating substantial capital into emerging markets. FDI strategies regarding foreign institutional investors’ thatoverbuy and oversell information have often been followed by domestic institutional investors or individual investors,leading us to explore this valuation behavior as well.

Taiwan is among the most economically-charged emerging markets, and the electronics firms located there have unsur-passed reputations for high quality manufacturing and productivity (Ernst, 2003; Lowe & Kenney, 1999). On the other hand,Taiwanese firms operating in China, one of the biggest markets in the world, provide good business opportunities that ben-

efit from a common language and similar cultures. Thus, China has become the main investment location for Taiwanesefirms. According to the Investment Commission of Ministry of Economic Affairs in Taiwan (hereafter MOEA), the high-techelectronics industry represents the majority of outward investments in Taiwan. These considerations are relevant to ourdiscussion of firm valuation behavior of various institutional investors of Taiwanese firms with operations in China.

1

2

(fitioaH

ivpmdpaiii

Sfiplifi

Ht

2

iAbbia2tpit

tpi3tsbi

ai

00 L.-L. Chang et al. / Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108

.1. Earnings and firm valuation

The relationship between ownership structure and firm value is important to corporate governance. Gillan and Starks2003, p. 15) argue that “foreign institutional investment has become an important influence in emerging equity markets, andrms may be motivated to improve corporate governance in order to attract foreign capital.” Shleifer and Vishny (1986) arguehat large shareholders have greater incentives to monitor the actions of managers and can benefit minority shareholdersn order to limit agency problems. Chung, Firth, and Kim (2002) find that a prevalence of institutional ownership reducespportunistic earnings management. Some empirical evidence on the monitoring role played by institutional investorslso reveals that institutions have greater influence on executive compensation contracts (Bertrand & Mullainathan, 2001;artzell & Starks, 2003).

Internalization theory (Buckley & Casson, 1976) assumes a firm’s proprietary ownership of firm-specific advantages (FSAs)n knowledge and products, extends across international boundaries of the enterprise, thus FDI will enhance a firm’s marketalue. For example, Bodnar and Weintrop (1997) argue that the earnings coefficient is higher for foreign income and moreersistent than the coefficient for domestic income. Garrod and Rees (1998) assert that earnings and net assets are valuedore highly for multinational firms than earnings and net assets of firms that are strictly domestic. Callen et al. (2005) claim

omestic earnings are more important than foreign earnings in explaining the variance of unexpected returns. However,revious research did not address valuation by different kinds of institutional investors. Empirically, institutional investorsre expected to monitor the company and are more concerned about the operating income of each component. While foreignnstitutional investors more carefully weigh fundamental analysis than domestic institutional investors when evaluatingnvestments in developing countries, domestic investors tend to be more familiar and have a greater comprehension of localndustry.

Traditionally the Taiwanese stock market is dominated by many individual investors that own small numbers of shares.tatistics from the Taiwan Stock Exchange (TSE), however, show that foreign equity ownership of Taiwanese electronicsrms has more than doubled in recent years, increasing from 15.2% in 2000 to 32.9% in 2006.Thus, foreign ownershiplays an important role in Taiwan’s stock market. According to the MOEA, China has become the predominant investment

ocation for the Taiwanese electronics industry. Investment of Taiwanese firms into China reached approximately $12 billionn 2006, of which approximately 40.8% was directed toward the electronics-related industry. This discussion motivates ourrst hypothesis:

1. Foreign and domestic institutional investors value earnings components differently, especially when comparing domes-ic and foreign operating income.

.2. Tax havens and income shifting

The Taiwanese electronics industry enjoys an important position as the country’s economic backbone, which justifiesnvestigation of its behavior in response to the Transfer Pricing Audit Regulation issued by the Taiwanese government.ccording to the MOEA, Taiwan’s investment in tax havens (such as Bermuda, Cayman Islands, and British Virgin Islands)ecame the most popular outward investment for the electronics industry, representing around 50% of outward investmentetween 2000 and 2006. During this period, outward investment of the electronics industry is second to only one other

ndustry: finance. Electronics firms, heavily dependent on the international market, are expected to be more nimble whendjusting strategies in response to export demands and rapidly changing economic environments (Guerrieri & Pietrobelli,004). Taiwanese electronics firms locate a significant portion of their foreign operations in tax havens, providing the abilityo shift income from high-tax countries to low-tax countries to maximize after-tax profits. By using tax havens and transferricing, MNCs can avoid or defer corporate taxes. For example, a corporation may use a tax haven subsidiary to accumulate

ncome, or it may use transfer pricing to shift income from high-tax countries to tax havens to postpone or avoid the corporateax of the home country.

The Organization for Economic Co-operation and Development (OECD) applies three key factors (no or only nominalaxes, protection of personal financial information, and lack of transparency) to identify a jurisdiction as a tax haven. In thisaper, a tax haven is defined as a country that imposes no taxes, even though it is transparent, exchanges personal financial

nformation, or requires substantial activity in the country. The OECD issued a black list of tax havens in 2000 which included8 jurisdictions (OECD, 2000). For our research purposes, we define 40 countries and regions as tax havens. In addition tohose listed by the OECD, we include from the United States, the states of Delaware and Nevada in our tax haven list. Thetate of Nevada has no corporate tax and the state of Delaware has no state income tax for corporations incorporated inut not transacting business in Delaware. These two U.S. states are the most popular investment locations in the world for

ncorporating.3 The 40 jurisdictions considered tax havens in this study are listed in Appendix.

Rousslang (1997) reports income tax avoidance is an important objective of income-shifting. Harris, Morck, Slemrod,

nd Yeung (1991) and Scholes, Wilson, and Wolfson (1992) report U.S. firm’s income-shifting due to tax differentials. MNCsn high-tax countries tend to report lower income, and MNCs in low-tax countries therefore tend to report higher income.

3 During the study period of this research, the states of Delaware and Nevada were both considered as tax havens regions by the Taiwanese Tax Authority.

L.-L. Chang et al. / Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108 101

Table 1Summarizes the industry distribution of sample firms by year. The total number of firms is 336 with 1346 firm-year observations across the 6-year period2000–2005.

Industry classification Year 2000 Year 2001 Year 2002 Year 2003 Year 2004 Year 2005 Total

Semi-conductor 24 28 33 37 41 41 204PC& Peripheral 35 39 45 53 55 55 282Photo-electrics LCD/LED 16 20 29 37 39 39 180Communication 10 15 22 23 27 29 126Electronic Components 30 38 48 50 52 53 271Network 6 13 14 17 17 17 84

Computer Service 6 7 10 11 10 11 55Others 16 20 23 28 28 29 144Total 143 180 224 256 269 274 1346

Harris (1993) finds that MNCs shift income among different tax jurisdictions in reaction to the tax regime. Hines and Rice(1994) further reveal that U.S. firms shift reported profits to tax haven affiliates. Rego (2003) finds that MNCs are able toavoid taxes and, in general, have lower worldwide effective tax rates than other firms. In addition, studies also report thattax policies can affect FDI locations (Blonigen, 2005; Hines, 1999).

Prior studies on transfer pricing and tax havens address the economic effect, income-shifting, transferring pricing deter-minants and audits (Li, 2006; Oyelere & Emmanuel, 1998; Reese, Henneberry, & Russell, 1989; Rousslang, 1997). None ofthe previous studies involve the reported earnings location change to avoid audit. In 2004, the tax authorities of the Tai-wanese government regulated the Transfer Pricing Audit Guide (TP regulation); it requests that firms prepare related partytransaction reports for audit for the year 2004, and explicitly asserts that the firm’s transactions with affiliates registeredin tax havens will be extensively audited. The high risk of being audited under this policy may have motivated firms tore-arrange earnings through related party transactions or transfer pricing policy so as to re-allocating earnings from sub-sidiaries located in tax havens to other countries. Anti-avoidance audit policies imply that the MNCs abuse their tax havenaffiliates to reallocate taxable incomes from their home country. Such firms may thus be at a high risk of rigorous audit bytax authorities from the home country or high tax jurisdictions.

The above research supports that MNCs have opportunities to avoid taxes by investing in low-tax countries and by shiftingincome from affiliates in high-tax countries to subsidiaries in low-tax countries or tax-havens. However, previous researchhas not discussed the effect of anti-tax avoidance regulation, which may alter the profit reported location of MNCs that shiftincome from tax haven countries to other countries. Therefore, we include this as a focus of our research, which leads to oursecond hypothesis:

H2. The firm values of Taiwanese firms are negatively associated with their earnings component from tax havens followingtransfer pricing audit regulation in Taiwan.

3. Sample selection and research model

3.1. Sample selection

The study utilizes all publicly listed electronics firms, mainly consisting of high technology and information servicebusinesses, from the Taiwan Stock Exchange (TSE). Electronics firms were selected because high-tech and informationservice related industries comprise approximately 49.5% of all listed firms in the TSE in 2006, and employ the main portionof indices in the TSE in recent years. The sample firms must meet the following selection criteria: (1) listed as electronicfirms on the TSE for the entire period spanning 2000–2005; (2) have multiple earnings sources, with multiple operationallocations in Taiwan (home country), China, and other countries or regions. The period 2000–2005 was chosen because Taiwandrastically increased outward investments to China during this time.4 Furthermore, these six years occur before Taiwanadopted Accounting Standard No. 7 “Consolidated Financial Statements” following International Accounting Standard (IAS27) “Consolidated and Separate Financial Statements,” which re-defined the evidence of an entity with substantial controllingpower and required it to consolidate investee entities’ financial statements. This was a significant change in reportingstructure for many firms’ accounting procedures.

Accounting and financial data (including related party financial status, consolidated statements, and institutional owner-ship structure) are retrieved from the Taiwan Economic Journal (TEJ) database. After eliminating 10 outliers with marketto book values (MB) exceeding 20, and excluding another six firm-years of missing values in institutional ownership, ourfinal sample includes 336 firms representing 1346 firm-years. Table 1 presents the industry distribution of sample firms

and their concentration, which mostly includes semi-conductor manufacturing, computer or PC peripheral, and electroniccomponents related industries.

4 The investment to Mainland China increased at a noticeable speed during this period. The amount of such investment was $1.25 billion in 1999, $2.61billion in 2000, $2.78 billion in 2001, $3.86 billion in 2002, $4.59 billion in 2003, $6.94 billion in 2004, and $6.01 billion in 2005 (MOEA).

1

3

bpo

aped

tCntttv

wttot

m

wyOifiit

ftamtniovo

w5

02 L.-L. Chang et al. / Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108

.2. Valuation models

Dechow, Hutton, and Sloan (1998) suggest that Ohlson’s valuation model provides a simplified framework for earnings-ased valuation empirical research. Since this research examines firm valuation in Taiwanese electronic firms, we followrior studies to adopt the Feltham–Ohlson model (Feltham & Ohlson, 1995; hereafter F–O model) to investigate the relationf market value, different income components, and institutional ownership.

The Ohlson model (Ohlson, 1995) assumes all book value changes satisfy the Clean Surplus Relation (CSR), which indicatesll changes in book value can be reported as income or dividends. In addition, the Present Value Relation (PVR) argues thatrice is equal to the present value of expected future dividends. Therefore, market value of the firm can be specified asxpected dividends discounted at the risk-free interest rate. The F–O model assumes accounting relations, CSR, and PVRenote the value of a firm’s equity and can be represented as the book value plus the present value of abnormal returns.

A firm’s annual reported earnings consist of operating income and non-operating income. To determine the value ofhe firm’s operating income by geographic location, we decompose operating income into domestic income (from Taiwan),hina income (from Chinese subsidiaries), and others (affiliates neither from Taiwan nor China operations). In addition,on-operating income is decomposed into investment income and other non-operating income. We use the share price athe close of the last business day of April in the following year to reflect the firm’s market value, according to requirementshat annual financial reports be declared within four months of the balance sheet date. Using these specifications, we statehe F–O model as model 1 and expect that the firm’s operating income variables will be positively associated with its firmalue.

MVit = ˛0 + ˇoBVit + ˇ1OPCit + ˇ2OPTWit + ˇ3OPOTHERit + ˇ4INVit + ˇ5OTHERit + ˇ

2004∑

t=2000

YDt + ε (1)

here MVit is the firm i’s market value of year t and set as its closing price at the end of April in the following year; BVit ishe firm i’s book value of equity at the end of year t; OPCit is the firm i’s operating income from Chinese subsidiaries for year; OPTWit is the firm i’s domestic operating income from Taiwan for year t; OPOTHERit is the firm i’s operating income fromthers for year t; INVit is the firm i’s investment income for year t; OTHERit is the firm i’s other non-operating income for year; YD2000–YD2004 is the indicator variables to control yearly economic condition for the respective year from 2000 to 2004.

After scaling each variable by the firm’s book value at the end of year t − 1 (BVit − 1), our F–O model 1 can be rewritten asodel 2:

MVit = ˇ0 + ˇ1OPC BVit + ˇ2OPTW BVit + ˇ3OPOTHER BVit + ˇ4INV BVit + ˇ5OTHER BVit + ˇ

2004∑

t=2000

YDt (2)

here MBit is the MVit/BVit−1, firm i’s market to book value at year t; BVit−1 is the firm i’s book value of equity at the end ofear t−1; OPC BVit is the OPCit/BVit−1, firm i’s operating income from Chinese subsidiaries for year t, scaled by book value;PTW BVit is the OPTWit/BVit−1, firm i’s domestic operating income from Taiwan for year t, scaled by book value; OPOTHER BVit

s the OPOTHERit/BVit−1, firm i’s operating income from others for year t, scaled by book value; INV BVit is the INVit/BVit−1,rm i’s investment income for year t, scaled by book value; OTHER BVit is the OTHERit/BVit−1, firm i’s other non-operating

ncome for year t, scaled by book value; YD2000–YD2004 is the indicator variables to control yearly economic condition forhe respective year from 2000 to 2004.

The primary hypothesis investigated in this study is that firm value is a function of earnings and institutional ownership,urther distinguishing between domestic and foreign institutional investors. Therefore, the first empirical test is to examinehe valuation behavior of institutional investors. Since ownership greater than 50% of the investment in a firm is regardeds majority ownership or control, we expect firm value to be significantly enhanced reflecting the benefit of increasedonitoring by majority institutional investors. We therefore include a total institutional investor indicator variable GOV

o identify firms with institutional ownership of more than 50%. The 50% or more of majority institutional ownership doesot need to be a single institutional owner investment in the firm, and it is usually a combination of various institutional

nvestors. (The same concept also applies to the following 20% domestic institutional ownership and 20% foreign institutionalwnership.) We then apply interaction terms to indicate the effect of the institutional investors interacting with each earningsariable on firm valuation. We expect that all earnings variables are positively associated with firm value when a firm’swnership exceeds 50% institutional investors. Thus the valuation model can be rewritten as the following in model 3.

MBit = ˇ0 + ˇ1OPC BVit + ˇ2OPTW BVit + ˇ3OPOTHER BVit + ˇ4INV BVit + ˇ5OTHER BVit + ˇ6GOV

+ ˇ7OPC BVit ∗ GOV + ˇ8OPTW BVit ∗ GOV + ˇ9OPOTHER BVit ∗ GOV + ˇ10INV BVit ∗ GOV

2004∑

+ ˇ11OTHER BVit ∗ GOV + ˇ

t=2000

YDt (3)

here GOV is the indicator variable for total institutional investor, set as one if the firm’s institutional ownership exceeds0%; zero otherwise. All other variables with definitions are the same as in model 2.

L.-L. Chang et al. / Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108 103

We further examine domestic institutional investor’s valuation by using a domestic institutional investor indicator vari-able DD to identify firms with domestic institutional ownership over 20%. This level of ownership constitutes a significantinfluence (i.e., ownership of 20% or more then accounting equity method is used). It is expected that all earnings variables arepositively associated with firm value when a firm has more than 20% domestic institutional ownership. Thus the valuationmodel can be specified as the following model 4.

MBit = ˇ0 + ˇ1OPC BVit + ˇ2OPTW BVit + ˇ3OPOTHER BVit + ˇ4INV BVit + ˇ5OTHER BVit + ˇ6DD + ˇ7OPC BVit ∗ DD

+ ˇ8OPTW BVit ∗ DD + ˇ9OPOTHER BVit ∗ DD + ˇ10INV BVit ∗ DD + ˇ11OTHER BVit ∗ DD + ˇ

2004∑

t=2000

YDt (4)

where DD is the indicator variable for domestic institutional investor, set as one if the firm’s domestic institutional ownershipexceeds 20%; zero otherwise. All other variables with definitions are the same as in model 2.

Similarly, a foreign institutional investor indicator variable FF is used to specify foreign institutional investor’s valuationon firm earnings. It is expected that all earnings variables are positively associated with firm value when a firm has foreigninstitutional ownership exceeding 20%. Thus the valuation model can be specified as the following model 5.

MBit = ˇ0 + ˇ1OPC BVit + ˇ2OPTW BVit + ˇ3OPOTHER BVit + ˇ4INV BVit + ˇ5OTHER BVit + ˇ6FF + ˇ7OPC BVit ∗ FF

+ ˇ8OPTW BVit ∗ FF + ˇ9OPOTHER BVit ∗ FF + ˇ10INV BVit ∗ FF + ˇ11OTHER BVit ∗ FF + ˇ

2004∑

t=2000

YDt (5)

where FF is the indicator variable for foreign institutional investor, set as one if the firm’s foreign institutional ownershipexceeds 20%; zero otherwise. All other variables with definitions are the same as in model 2.

The longitudinal nature of our data may lead to autocorrelation. Wooldridge (2001) asserts that in the case of autocorrela-tion, the error term includes a time-omitted factor. Following Wooldridge (2008, p. 310–312) we employ a lagged-dependentvariable in the cross-sectional model to account for unobserved factors, and our regression models are modified to considerthe inclusion of MBit-1 to control the potential misspecification effect.

Another main hypothesis from the study is to examine the outcome when these electronic firms shift income from taxhavens to other countries in response to audit guidelines from the Taiwan government on newly issued transfer pricingaudit regulation effective in 2004. We decompose consolidated earnings into four components: taxed operating income, un-taxed operating income (operating income from tax havens), investment income, and other income. We define tax havensaccording to the standards from OECD and also include two more from the United States, the states of Nevada and Delaware(see Appendix).

After scaling all variables by the firm’s equity book value at the end of year t−1(BVit-1), we specify the following model6:

MBit = ˇ0 + ˇ1TAX BVit + ˇ2UNTAX BVit + ˇ3INV BVit + ˇ4OTHER BVit + ˇ5TAX BVit ∗ YD + ˇ6UNTAX BVit ∗ YD

+ ˇ7INV BVit ∗ YD + ˇ8OTHER BVit ∗ YD + ˇ

2004∑

t=2000

YDt (6)

where UNTAX NIit + TAX NIit is the firm i’s operating net income for year t; UNTAX BVit is the firm i’s operating net incomefrom tax havens subsidiaries for year t, scaled by book value; TAX BVit is the firm i’s operating net income for year t subtractUNTAX NIit, scaled by book value; INV BVit is the firm i’s investment income for year t, scaled by book value; OTHER BVit is thefirm i’s earnings for year t subtract UNTAX NIit, TAX NIit, and INV NIit, then scaled by book value; YD is an indicator variablefor firm year, set as one if the firm year is 2004 or 2005; zero otherwise.

All other variables with definitions are the same as in model 2.Since the empirical test examines the subsequent effect of Taiwan audit regulation on transfer pricing effective in 2004,

we expect that the untaxed earnings components from 2004 and 2005 should reflect weaker association to firm value thanthe association before the regulation was implemented in 2004. Thus, the coefficient ˇ6 for interaction term UNTAX BVit *YDfrom model 6 is expected to be negative.

4. Empirical results and analysis

Table 2 summarizes the sample firms’ descriptive statistics. Note that the maximum market-to-book-ratio is 14.73 andthe minimum is 0.1; the average is 1.88. It is expressed by market value of the maximum (minimum) NTD $1453 million

(NTD $0.06 million) and by book value of the maximum (minimum) NTD $398 million (NTD $0.11 million).

We also find that correlation coefficients from our correlation matrix (untabulated) are consistently less than 0.7 and thevariance inflation factors are between 1 and 3, lower than the conventional cutoff points, indicating a multicollinearity maynot pose a concern in our regression analysis (Belsley, Kuh, & Welsch, 1980; Tabachnick & Fidell, 1996).

104 L.-L. Chang et al. / Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108

Table 2Summarizes the results of descriptive statistics for testing variables. The variables are defined in more details in Section 3.

Variables N Min Max Mean Std. Dev.

MB 1346 0.13 14.73 1.88 1.55MVt ($NT million) 1346 0.06 1452.64 21.21 76.65BVt−1 ($NT million) 1346 0.11 398.97 11.20 34.73OPC BV 1346 −0.12 3.27 0.11 0.09OPTW BV 1346 −3.04 1.00 0.09 0.18OPOTHER BV 1346 −0.86 1.70 0.06 0.15INV BV 1346 −0.52 0.73 0.01 0.08OTHER BV 1346 −0.81 0.90 −0.05 0.11UNTAX BV 1346 −0.51 0.64 0.01 0.04TAX BV 1346 −0.86 1.70 0.15 0.19UNTAX BV*YD 1346 −0.51 0.55 0.01 0.03TAX BV*YD 1346 −0.86 0.80 0.06 0.14INV BV*YD 1346 −0.52 0.73 0.01 0.06OTHER BV*YD 1346 −0.81 0.90 −0.02 0.08Y00 1346 0.00 1.00 0.11 0.31Y01 1346 0.00 1.00 0.13 0.34Y02 1346 0.00 1.00 0.17 0.37Y03 1346 0.00 1.00 0.19 0.39Y04 1346 0.00 1.00 0.20 0.40

Variable definition: MBit = MVit/BVit , firm i’s market to book value at year t; MVit = firm i’s market value of year t and set as its closing price at the endof April in the following year; BVit−1 = firm i’s book value of equity at the end of year t−1; MBit = MVit/BVit , firm i’s market to book value at year t;OPCit = firm i’s operating income of Chinese subsidiaries for year t; OPC BVit = OPCit/BVit; OPTWit = firm i’s domestic operating income from Taiwan foryear t; OPTW BVit = OPTWit/BVit; OPOTHERit = firm i’s operating income of others for year t; OPOTHER BVit = OPTORit/BVit; INVit = firm i’s investment incomefor year t; INV BVit = INVit/BVit; OTHERit = firm i’s other non-operating incomes for year t; OTHER BVit = NONTORit/BVit; MVit = firm i’s market value at the nextend of April of year t; UNTAX BVit = operating net income from tax havens subsidiaries/BVit; TAX BVit = (operating net income – untaxni)/BVit; operating netincome = UNTAXNIit + TAXNIit; INV BVit = investment income/BVit; OTHER BVit = (earnings – untaxni – taxni – invni)/BVit; MVit = firm i’s market value at thenext end of April of year t; YD = indicator variable for firm year, set as one if the firm year is 2004 or 2005; zero otherwise; UNTAX BVit*YD = operating netiie

iis

o

TS

VfioIYit2

ncome from tax havens subsidiaries/BVit * YD year dummy; TAX BVit*YD = (operating net income – untaxni)/BVit * YD year dummy; INV BVit*YD = investmentncome/BVit * YD year dummy; OTHER BVit*YD = (earnings – untaxni – taxni – invni)/BVit * YD year = dummy; Y00–Y04 = indicator variables to control yearlyconomic condition for the respective year from 2000 to 2004. Descriptive statistics for variables.

Table 3 presents results from F–O model 2. Consistent with expectations, the results indicate that components of operatingncome from China and Taiwan are positively associated with firm value at the 1% significance level. However, non-operatingncome components (including investment income and other non-operating income) are not associated with firm value. This

uggests that non-operating income may be less significant than operating income when establishing firm value.

Table 4 presents empirical results of distinctive institutional investors’ valuation on earnings: institutional investorwnership over 50% (model 3 with GOV); domestic institutional ownership over 20% (model 4 with DD); and foreign insti-

able 3ummarizes the results of F–O regression model 2.

Variable ˇi t-Value Prob.

C 1.73 14.14 0.0000***

OPC BV 4.33 4.44 0.0000***

OPTW BV 3.03 7.49 0.0000***

OPOTHER BV 0.70 0.61 0.5388INV BV −0.28 −0.20 0.8391OTHER BV −0.57 −0.66 0.5117MBit−1 0.06 2.21 0.0272**

Y00 −0.22 −1.28 0.2001Y01 −0.07 −0.44 0.6610Y02 −0.96 −7.57 0.0000***

Y03 −0.46 −3.58 0.0004***

Y04 −0.83 −6.84 0.0000***

R2 0.23Adj. R2 0.22

ariable definition: MVit = firm i’s market value at the next end of April of year t; BVit = firm i’s book value of equity at the beginning of year t; MBit = MVit/BVit ,rm i’s market to book value at year t; OPCit = firm i’s operating income from Chinese subsidiaries for year t; OPC BVit = OPCit/BVit; OPTWit = firm i’s domesticperating income from Taiwan for year t; OPTW BVit = OPTWit/BVit; OPOTHERit = firm i’s operating income from others for year t; OPOTHER BVit = OPTORit/BVit;NVit = firm i’s investment income for year t; INV BVit = INVit/BVit; OTHERit = firm i’s other non-operating incomes for year t; OTHER BVit = OTHERit/BVit;00–Y04 = yearly dummy variables indicate the respective year from 2000 to 2004; GOV = indicator variable for total institutional investor, set as one

f firm’s institutional ownership over 50%; zero otherwise; DD = indicator variable for domestic institutional investor, set as one if firm’s domestic institu-ional ownership over 20%; zero otherwise; FF = indicator variable for foreign institutional investor, set as one if firm’s foreign institutional ownership over0%; zero otherwise.** Significant at 5%.

*** Significant at 1%.

L.-L. Chang et al. / Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108 105

Table 4Summary of results for institutional investors on firm value.

Variable Model 3 Model 4 Model 5

ˇi t-Value Prob. ˇi t-Value Prob. ˇi t-Value Prob.

C 1.48 15.01 0.0000*** 1.57 14.30 0.0000*** 1.49 15.78 0.0000***

OPC BV 4.18 2.91 0.0037*** 5.12 2.24 0.0252** 4.16 7.13 0.0000***

OPTW BV 3.35 9.29 0.0000*** 2.71 4.42 0.0000*** 3.49 10.10 0.0000***

OPOTHER BV 2.12 2.63 0.0087*** 0.69 0.64 0.5251 1.89 2.12 0.0346**

INV BV −2.57 −1.59 0.1118 −3.67 −1.79 0.0744* −1.93 −1.37 0.1708OTHER BV 0.08 0.08 0.9356 −1.81 −1.49 0.1366 0.09 0.09 0.9254MBit−1 0.09 3.07 0.0022*** 0.09 3.24 0.0012*** 0.09 3.06 0.0022***

GOV −0.19 −1.14 0.2543OPC BV *GOV 3.03 1.46 0.1458OPTW BV *GOV 2.73 2.89 0.0039***

OPOTHER BV *GOV 1.39 0.52 0.6050INV BV *GOV 9.15 3.37 0.0008***

OTHER BV *GOV 3.84 2.21 0.0276**

DD −0.17 −1.57 0.1169OPC BV *DD 1.23 0.49 0.6237OPTW BV *DD 2.24 2.98 0.0030***

OPOTHER BV *DD 1.57 0.79 0.4279INV BV *DD 9.37 3.53 0.0004***

OTHER BV *DD 6.11 4.15 0.0000***

FF −0.28 −1.57 0.1178OPC BV *FF 5.20 1.93 0.0541*

OPTW BV *FF 4.22 3.56 0.0004***

OPOTHER BV *FF 3.99 1.80 0.0723*

INV BV *FF 9.57 3.18 0.0015***

OTHER BV *FF 4.56 2.27 0.0234**

Year Fixed Effects Yes Yes YesR2 0.37 0.37 0.40Adj. R2 0.36 0.36 0.39

Variable definitions: MVit = firm i’s market value at the next end of April of year t; BVit = firm i’s book value of equity at the beginning of year t; MBit = MVit/BVit ,firm i’s market to book value at year t; OPCit = firm i’s operating income of Chinese subsidiaries for year t; OPC BVit = OPCit/BVit; OPTWit = firm i’s domesticoperating income from Taiwan for year t; OPTW BVit = OPTWit/BVit; OPOTHERit = firm i’s operating income of others for year t; OPOTHER BVit = OPOTHERit/BVit;INVit = firm i’s investment income for year t; INV BVit = INVit/BVit; OTHERit = firm i’s other non-operating incomes for year t; OTHER BVit = OTHERit/BVit;Y00–Y04 = yearly dummy variables indicate the respective year from 2000 to 2004; GOV = indicator variable for total institutional investor, set as one iffirm’s institutional ownership over 50%; zero otherwise; DD = indicator variable for domestic institutional investor, set as one if firm’s domestic institutionalownership over 20%; zero otherwise; FF = indicator variable for foreign institutional investor, set as one if firm’s foreign institutional ownership over 20%;zero otherwise.

* Significant at 10%.

** Significant at 5%.

*** Significant at 1%.

tutional ownership over 20% (model 5 with FF). Overall, the results from these regression models are consistent with resultsfrom model 2, where components of major operating income OPC and OPTW are positively associated with firm value, butcomponents of non-operating income INV and OTHER are for the most part not associated with firm value, except for invest-ment income in model 4. The association of major operating income with firm value is positive and is minimally supported atthe 5% significance level, indicating most investors continue to value the firm’s core businesses over non-core operations orother revenue sources. This finding meets the expectation of our first hypothesis that operating income is positively relatedto firm value.

Considering the weighting of various institutional investors, the results related to interaction terms of domestic operatingincome OPTW*GOV, OPTW*DD, and OPTW*FF across all models, support strong and positive associations with firm valueat the 1% significance level (model 3 with ˇ9 = 2.73, t-value = 2.89; model 4 with ˇ9 = 2.24, t-value = 2.98; and model 5 withˇ9 = 4.22, t-value = 3.56). Conversely, operating income from China does not demonstrate statistically significant positiveassociation of firm value with the interaction terms, except OPC*FF (ˇ8 = 5.20, t-value = 1.93 in model 5). This finding indicatesthat a firm’s core earnings directly from Taiwan contribute to the essentials of firm value even among various institutionalinvestors. Furthermore, results from model 4 suggest domestic institutional investors favor domestic operating earnings andinvestment income over operating earnings from non-domestic countries (i.e. from China and other affiliates). The positiveassociation of domestic earnings with firm value in model 4 is strongly supported at the 1% significance level, while othercorrelations to firm value are not significant for operating income generated from China and from other income sources.These results indicate that the components of earnings have different effects on firm valuation.

Taiwanese firms may invest in China through third-party countries and may re-allocate earnings through the firms’

investment income or other revenue sources. During the study period, many sensitive and critical electronics industrieswere still prohibited by the Taiwan government from directly investing in China. However, there were firms using loopholesto create shell companies and then indirectly investing in China. Such behavior may explain the domestic institutional

106 L.-L. Chang et al. / Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108

Table 5Summary of results for taxed earnings and untaxed earnings on firm value.

Variables Coefficient t-Statistic Prob.

C 1.64 9.95 0.0000***

TAX BV 2.37 5.41 0.0000***

UNTAX BV 4.95 2.81 0.0050***

INV BV −4.27 −2.42 0.0159**

OTHER BV −0.84 −0.73 0.4645TAXNI*YD 1.25 1.38 0.1688UNTAXNI*YD −4.36 −1.67 0.0957*

INVNI*YD 6.88 2.41 0.0159OTHERNI*YD 0.58 0.34 0.7337MBit−1 0.07 2.38 0.0173**

Year fixed effects YesR2 0.26Adj. R2 0.25

Variable definition: MBit = market value at period t/BVit; operating net income = UNTAXNI + TAXNI; UNTAX BVit = operating net income from tax havenssubsidiaries/BVit; TAX BVit = (operating net income – UNTAXNI)/BVit; INV BVit = investment income/BVit; OTHER BVit = (earnings – UNTAXNI – TAXNI –INVNI)/BVit; YD = indicator variable for firm year, set as one if the firm year is 2004 or 2005; zero otherwise; Y00–Y04 = indicator variables to controlyearly economic condition for the respective year from 2000 to 2004.

iiaTtisiwitiia

oc

ROaHe(tit

ttif−fietmic

* Significant at 10%.** Significant at 5%.

*** Significant at 1%.

nvestor’s higher valuation of investment income (INV*DD) and other non-operating income (OTHER*DD) over operatingncome from China (OPC*DD). Despite Taiwan’s politicized climate toward investments in China, Taiwanese firms remainttracted to China operations mainly due to an abundant lower-wage Chinese labor force that lowers manufacturing costs.hough lower cost operations in China may enhance firm value in general (i.e. OPC BV is positive and significant, ˇ2 = 5.12,-value = 2.24 in model 4), this earnings component, over the long term, may not be highly valued by domestic institutionalnvestors (i.e. OPC*DD is positive but not significant) since the advantage of the low-cost Chinese labor force may not beustainable. The same analysis applies to the firm’s operating income from other affiliates, an income component domesticnstitutional investors may not highly value (i.e. OPOTHER*DD is not significant). However, results of firm value associated

ith the interaction term of investment income (INV*DD, ˇ8 = 9.37, t-value = 3.53 in model 4) and other non-operatingncome (OTHER*DD, ˇ8 = 6.11, t-value = 4.15), show these components are positive and significant at the 1% level. We arguehat some Taiwanese firms benefitted from regulatory loopholes allowing them to create shell investment companies thatndirectly invest in their China operations, and that these shell companies thereby generated a major source of investmentncome or another type of non-operating income. This explains the higher values placed on the firm’s investment incomend other types of revenue sources over legitimate operating income from China.

The results of model 3 demonstrate the valuation behavior of majority institutional investors. In general, the resultsf model 3 are similar to those of model 4 and suggest that Taiwanese firms with majority institutional ownership areomprised mainly of domestic institutional investors.

Results for model 5 support our first hypothesis and show that foreign institutional investors value all types of income.egardless of the earnings component, the interaction terms of operating income, and non-operating income (OPC*FF,PTW*FF, OPOTHER*FF, INV*FF, and OTHER*FF) are significantly positively related to firm value, The coefficients of ˇ8–ˇ12re all positive and significant at least at the 10% level, while ˇ9 (OPTW*FF) and ˇ11 (INV*FF) are significant at the 1% level.owever, the variance of these significance levels also demonstrates that foreign institutional investors prefer operatingarnings of the home country (Taiwan) and investment income (at the 1% level) to operating earnings from non-domestici.e. from China and other affiliates, at the 10% level) and other non-operating income (at the 5% level). This may suggesthat firm value is enhanced due to overall performance regardless of earnings components. It also compares the preferencen valuation by domestic and foreign institutional investors, the former which benefits from local industry knowledge, andhe latter which relies on fundamental analysis.

Table 5 indicates that operating income is positive and significantly related to firm value, regardless of whether or nothe income comes from a tax haven. The coefficients of operating income from tax havens subsidiaries and income not fromax havens, UNTAX BV and TAX BV, are 4.95 and 2.37, respectively, and are significant at the 1% level. However, investmentncome INV BV is negative and significantly associated with firm value (ˇ4 = −4.27, t-value = −2.42). Furthermore, incomerom tax havens subsidiaries is significantly negatively relative to firm value after the year 2004 (ˇ4 of UNTAX BV*YD is4.36, t-value = −1.67). This suggests that the transfer pricing audit regulation does have a significant effect on electronicsrms’ policy while the effects of other components on earnings remain the same. Taiwanese electronics firms prove to befficient at reallocating income sources through transaction models with affiliates. It is worth noting that the interaction

erm of investment income INV BV*YD is significantly positive relative to the firm value (ˇ7 = 6.88 and t-value = 2.41), which

eans the valuation of investment income after 2004 increased. This also provides evidence that firms have again shiftedncome through their affiliates. These firms may rearrange their reported earnings from tax havens subsidiaries to otherountries.

L.-L. Chang et al. / Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108 107

5. Conclusions

This paper uses the F–O model to examine earnings components of Taiwanese electronic firms in relation to firm valueby different kinds of institutional investors. Our sample includes 1346 firm-years from 2000–2005. After distinguishingbetween majority institutional investors, domestic institutional investors, and foreign institutional investors, we find thatoperating income differs in its valuation according to the location of operation, while non-operating income enhances firmvalue regardless of investment income or other income sources. This suggests that if a firm desires to increase institutionalinvestors and foreign capital, it should consider allocating income to domestic income and investment income. Furthermore,domestic institutional investors value operating income from their home country as well as non-operating income. Eachcomponent of earnings, however, regardless of the source, enhances overall firm valuation, and is significantly valued byforeign institutional investors.

Domestic institutional investors value highly on domestic income and investment income, as do majority institutionalinvestors. Allocating income to domestic and investment incomes is a high priority, which suggests that the disclosure ofoperating results from China may have little or no impact on many domestic institutional investors.

In addition, our study also investigates changes of profit-reported locations inspired by transfer pricing audit regulation.Our results support that Taiwanese electronics firms rearrange reported profits from tax havens subsidiaries to affiliatesin other countries in response to the transfer pricing audit guide implemented in 2004. The results do not mean that thetransfer pricing audit guide alters firm transfer of profit to the home country. In contrast, the firms may have been shiftingtheir profits to other low tax rate countries or to regions that do not require payment of taxes. However, the transfer pricingaudit guide regulation did influence firms’ choices in reported earnings locations for the electronics firms.

The above findings provide implications for investors and regulators. Institutional investors benefit from knowing thatdomestic operating income is valued as the foundation of a successful business. Although a Taiwanese firm may transfersome operations to China, lured by inexpensive manufacturing costs, this may not be viewed as positively by domesticinstitutional investors who consider it a distraction to the businesses and detrimental to essential core operations. However,foreign institutional investors appreciate the enhancement to overall firm value. Non-operating income tends to be treatedas a side business; however a MNC is encouraged to increase revenues beyond its core operations by seeking diversifiedinvestments and other types of revenue sources, which are highly valued by institutional investors. Furthermore, our resultsfind that MNCs react quickly when a governmental audit rule is implemented which impacts their reported profits and taxburden. In other words, this policy is very effective for firms’ business activities involving tax shelters.

Lastly, several research limitations may apply to our sample of Taiwanese electronics firms used in this study. Due to thefirms’ specific characteristics of operations in their home country of Taiwan and other locations, as well as investments inChina and other regions, our study results may not be generalized to apply to all other emerging markets. In addition, ourconclusions are limited by the short time period for examining pre- and post-regulatory audit event.

Appendix.

List of tax havens-40 jurisdictionsAndorra Cook Islands Mauritius St. Kitts & NevisAnguilla Cyprus Montserrat St. Vincent and the GrenadinesAntigua and Barbuda Dominica Nauru Turks & CaicosAruba Gibraltar Netherlands Antilles The Principality of LiechtensteinBahamas Grenada Niue The Principality of MonacoBahrain Guernsey Panama (Spanish), (English) The Republic of the Marshall IslandsBelize Isle of Man Samoa US Virgin IslandsBermuda Jersey San Marino VanuatuBritish Virgin Islands Liberia Seychelles State of Delaware in USCayman Islands Malta St. Lucia State of Nevada in US

References

Belsley, D. A., Kuh, E., & Welsch, R. E. (1980). Regression diagnostics: Identifying influential data and sources of collinearity. New York: Wiley.Bertrand, M., & Mullainathan, S. (2001). Are CEOs rewarded for luck? The ones without principals are. The Quarterly Journal of Economics, 116(3), 901–932.Blonigen, B. (2005). A review of the empirical literature on FDI determinants. Atlantic Economic Journal, 33(4), 383–403.Bodnar, G. M., & Weintrop, J. (1997). The valuation of the foreign income of US multinational firms: A growth opportunities perspective. Journal of Accounting

and Economics, 24(1), 69–97.Buckley, P. J., & Casson, M. (1976). The future of the multinational enterprise. London: Macmillan.Callen, J. L., Hope, O., & Segal, D. (2005). Domestic and foreign earnings, stock return variability, and the impact of investor sophistication. Journal of

Accounting Research, 43(3), 377–412.Chung, R. M., Firth, M., & Kim, J. (2002). Institutional monitoring and opportunistic earnings management. Journal of Corporate Finance, 8(1), 29–48.Dechow, P. M., Hutton, A. P., & Sloan, R. G. (1998). An empirical assessment of the residual income valuation model. Journal of Accounting and Economics,

26(1), 1–34.

Doukas, J., & Travlos, N. (1988). The effect of corporate multi-nationalism on shareholders’ wealth: Evidence from international acquisition. Journal of

Finance, 43(5), 1161–1175.Ernst, D. (2003). Pathways to Innovation in the global network economy: Asian upgrading strategies in the electronics industry. East-West Center working paper.

Available on the Internet at http://www.eastwestcenter.org/fileadmin/stored/pdfs/ECONwp058.pdf. Accessed 20.02.12Fatemi, A. M. (1984). Shareholders benefits from corporate international diversification. Journal of Finance, 39(5), 1325–1344.

1

F

GG

GH

H

HHHK

L

L

O

OO

RRRS

S

STWW

08 L.-L. Chang et al. / Journal of International Accounting, Auditing and Taxation 22 (2013) 98– 108

eltham, G. A., & Ohlson, J. A. (1995). Valuation and clean surplus accounting for operating and financial activities. Contemporary Accounting Research, 11(2),689–731.

arrod, N., & Rees, W. (1998). International diversification and firm value. Journal of Business Finance and Accounting, 25(9 & 10), 1255–1281.illan, S. L., & Starks, L. T. (2003). Corporate governance, corporate ownership, and the role of institutional investors: A global perspective. Journal of Applied

Finance, 13(2), 4–22.uerrieri, P., & Pietrobelli, C. (2004). Industrial districts’ evolution and technological regimes: Italy and Taiwan. Technovation, 24(11), 899–914.arris, D. (1993). The impact of U.S. tax law revision on multinational corporations’ capital location and income shifting decisions. Journal of Accounting

Research 31, (Suppl.), 111–140.arris, D., Morck, R., Slemrod, J., & Yeung, B. (1991). Income shifting in U.S. multinational corporations. Working Paper. NBER No.3924. Available on the Internet

at http://www.nber.org/papers/w3924. Accessed 05.01.10artzell, J. C., & Starks, L. T. (2003). Institutional investors and executive compensation. The Journal of Finance, 58(6), 2351–2374.ines, J. R., & Rice, E. M. (1994). Fiscal paradise: Foreign tax havens and American business. The Quarterly Journal of Economics, 109(1), 149–182.ines, J. R. (1999). Lessons from behavioral response to international taxation. National Tax Journal, 52(2), 305–322.im, W. S. (2000). Does FDI increase firm value in emerging markets? Working paper. Weatherland Center for International Affairs, Harvard University.

Available on the Internet at http://www.wcfia.harvard.edu/node/414. Accessed 05.12.10i, J. (2006). Transfer pricing audits in Australia, China and New Zealand: A developed vs. developing countries perspective. The International Tax Journal,

30(4), 21–28.owe, N., & Kenney, M. (1999). Foreign investment and the global geography of production: Why Mexican consumer electronics industry failed. World

Development, 27(8), 1427–1443.ECD. (2000). 33 Jurisdictions committed to improving transparency and establishing effective exchange of information in tax matters: List of unco-operative tax

havens. Available on the Internet at: http://www.oecd.org/document/19/0,3343,en 2649 33753 1903251 1 1 1 1,00.html. Accessed 15.07.09hlson, J. (1995). Earnings, book values and dividends in equity valuation. Contemporary Accounting Research, 11(2), 661–687.yelere, P. B., & Emmanuel, C. R. (1998). International transfer pricing and income shifting: Evidence from the UK. European Accounting Review, 7(4),

623–635.eese, R. A., Henneberry, S. R., & Russell, J. R. (1989). Transfer pricing in multinational firms: A review of the literature. Agribusiness, 5(2), 121–191.ego, S. O. (2003). Tax-avoidance activities of U.S. multinational corporations. Contemporary Accounting Research, 20(4), 805–833.ousslang, D. J. (1997). International income shifting by US multinational corporations. Applied Economics, 29(7), 925–934.choles, M. S., Wilson, G. P., & Wolfson, M. A. (1992). Firm’ responses to anticipated reductions in tax rates: The Tax Reform Act of 1986. Journal of Accounting

Research, 30(Suppl.), 161–185.ikka, P., & Willmott, H. (2010). The dark side of transfer pricing: Its role in tax avoidance and wealth retentiveness. Critical Perspectives on Accounting,

21(4), 342–356.hleifer, A., & Vishny, R. (1986). Large shareholders and corporate control. Journal of Political Economy, 94(3), 461–488.abachnick, B., & Fidell, L. (1996). Using multivariate statistics (third ed.). New York: Harper Collins.ooldridge, J. M. (2001). Econometric analysis of cross section and panel data. Cambridge, MA: MIT Press.ooldridge, J. M. (2008). Introductory econometrics: A modern approach (fourth ed.). Mason, OH: South-Western College Publishing.