Kim M. Shima a, Elizabeth A. Gordon b,aCollege of Business and Economics, Department of Ab Fox School of Business and Management, Departme
motivated by the argument that the lack of portfolio diversity is primarily the result of informationasymmetry issues that can be resolved by providing relevant and reliable information.1 However, itis unclear whether a country using IFRS creates sufcient incentives for increasing foreign investment
Corresponding author. Tel.: +1 215 204 6422; fax: +1 215 204 5587.E-mail addresses: firstname.lastname@example.org (K.M. Shima), email@example.com (E.A. Gordon).
1 As in literature referenced, we consider US GAAP, International Financial Reporting Standards, and its predecessorInternational Accounting Standards, as internationally accepted accounting standards.
J. Account. Public Policy 30 (2011) 481500
Contents lists available at SciVerse ScienceDirect
J. Account. Public Policy
journal homepage: www.elsevier .com/locate/ jaccpubpol0278-4254/$ - see front matter 2011 Elsevier Inc. All rights reserved.document that US investment is associated with IFRS only whenit is combined with a strong regulatory environment, specicallya strong enforcement regime. We also nd that mandating IFRS isattractive to US investors only when combined with a strong regu-latory environment.
2011 Elsevier Inc. All rights reserved.
We analyze whether a countrys use of international nancial reporting standards (IFRS) is associ-ated with increased US investment in foreign equities, related to the home bias effect. Our study isdoi:10.1016/j.jaccpubpol.2011.07.001ccounting and Finance, California State University, East Bay, United Statesnt of Accounting, Temple University, United States
a b s t r a c t
This paper investigates whether a countrys use of internationalnancial reporting standards (IFRS) is associated with increasedUS investment in foreign equities. The recent global trend towardsthe use of IFRS may signal attempts by policy makers to reduceinformation asymmetries for international investors. However, aconcern is that these standards must be accompanied by a strongerregulatory environment in order for them to gain legitimacy. Inves-tor allocation choice is based on US holdings of foreign equities andthe regulatory environment is interpreted using two distinct fac-tors: the legal standards system and the enforcement regime. WeIFRS and the regulatory environment: The case of U.S.investor allocation choice
or if additional conditions are required.2 As more countries use IFRS, the question arises of whether stan-dards alone or in combination with institutions and incentives determine reporting outcomes and ben-ets (Ball et al., 2003; Leuz et al., 2010; Beneish and Yohn, 2008). We provide evidence that a strongenforcement regime is critical for foreign investments, and use of IFRS complements strong enforcement.We nd limited evidence that a countrys legal standards are related to foreign investment.
In most countries, the choice of accounting standards is a country-level public-policy decision.Standard setters and regulators are likely interested in the economic effects of the policy decisionon international investment and capital ows.3 At the national level, greater perceived benets poten-tially entice standard setters to endorse better reporting, such as allowing or mandating IFRS. Improvednancial disclosure regulation could improve market liquidity, which has been associated with higherlevels of economic development (Frost et al., 2006), greater capital accumulation, productivity improve-ments (Levine and Zervos, 1998), and increased capital mobility (Young and Guenther, 2003).4 A cross-country study can offer insights into the aggregate country-level consequences of IFRS.
We examine three hypotheses on the use of IFRS and US investment in foreign equities. We focuson US investment because IFRS is considered an internationally accepted accounting standard compa-rable to US GAAP. Both standards share a common objective of providing information to capital pro-
482 K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500viders and other nancial statement users useful in making economic decisions. While somedifferences exist between the two, US investors would likely nd IFRS statements more familiar thanones prepared under a countrys prior home GAAP.5 So, if we were to nd an investment effect, we aremore likely to detect it for investors familiar with US GAAP.6 Our rst hypothesis asserts that US inves-tors allocate more of their equity investment choices in countries that use IFRS. Our second explores USequity investment and two distinct elements of the regulatory environment: legal standards andenforcement regime. Finally, we examine the joint effect of the use of IFRS and strong regulatory envi-ronment as our third hypothesis.
We test our rst hypothesis estimating a regression model with US investment in foreign equitiesas the dependent variable, IFRS as an independent variable, and include control variables associatedwith foreign investment. We then include the legal standards and the enforcement regime as depen-dent variables to test our next two hypotheses. Our sample period is the years 20032006, a periodwhen numerous countries mandated IFRS.
Overall, our evidence indicates a countrys use of IFRS is not associated with US equity investment,in contrast to our rst hypothesis. This result may not be surprising given the quality of nancialreporting in countries such as Australia and the United Kingdom that have adopted IFRS. Our resultssupport the second hypothesis that the regulatory environment, specically the enforcement regime,is positively associated with greater US investment in foreign equities. The evidence shows the inter-action of IFRS and a strong regulatory environment is signicantly associated with higher US invest-ment consistent with our third hypothesis. Additional analyses nd stronger results when IFRS is
2 Prior research nd that the voluntary use of international accounting standards for nancial reporting reduces informationasymmetries, providing accounting data and disclosures that are more informative, more familiar to foreign investors, and helpfulin cross-country rm analyses than local GAAP (Leuz and Verrecchia, 2000; Ashbaugh and Pincus, 2001; Covrig et al. 2007).
3 Related studies have investigated the investment effect at a rm-level (Bradshaw et al., 2004; Covrig et al., 2007; Yu, 2009)nding rms voluntarily migrate towards improved reporting standards if given sufcient marketplace incentives to do so.
4 Other international research that similarly explores variation across countries using a country-level design includes studiessuch as Young and Guenther (2003), Bushman et al. (2004), Bhattacharya et al. (2003) and Frost et al. (2006). In choosing a countrylevel analysis, we are also aggregating across the rms that make up the population of a country. By aggregating, we avoidproblems with rm specic measures and limit criticisms of not properly controlling or being able to control for all rm-specicfactors associated with foreign investment. Additionally, the measure of investment we use is a comprehensive measure of USinvestment in foreign equities, not limited to only institutional investment in specic rms, based on mandatory reporting of US-resident issuers of securities and US-resident custodians of foreign securities. Data is based on a mandatory survey jointlyconducted by the US Dept. of the Treasury, the Federal Reserve Bank of NY, and the Board of Governors of the Federal ReserveSystem with signicant penalties for non-compliance.
5 Literature nds that overall US GAAP and IFRS are comparable, e.g. a current working paper by Barth et al. (2011) attribute thegreater value relevance of IFRS reporting to the increase in comparability with US GAAP.
6 Yu (2009) examines security-level holdings of international mutual funds in 31 countries from 200 to 2007 to assess the effectof IFRS adoption. Unlike our study, she nds that adoption of IFRS increases holdings, attributing this to a reduction in informationprocessing cost and increasing comparability of nancial information. Similar to us, she nds that results are more pronounced in
countries with good investor protection and legal enforcement.
mandated. These ndings support the argument that IFRS reduces information asymmetries if institu-
tions. Overall, results of these checks support the main ndings except we nd some sensitivity to the
K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500 483measure of US investment.This paper contributes to the literature examining the relation between attributes in the reporting
environment and investments in foreign equities. First, the country-level effects of IFRS and the reg-ulatory environment are examined separately and jointly for their association with a broad measure ofUS ownership of foreign equities. Several studies investigate the relation between different proxies foraccounting quality and foreign investment by examining institutional ownership in non-US compa-nies generally nding that higher quality is related to greater investment (Bradshaw et al., 2004;Aggarwal et al., 2005; Covrig et al., 2007; Yu, 2009). We further explore this issue by investigatingthe investment by US residents in foreign equities from countries where IFRS is used. Thus far, evi-dence is mixed with some studies using rm-level data and nding an increase investment allocation(Yu, 2009), others nding no effect (Beneish et al., 2011) and still others nding only an effect condi-tional on certain factors (Florou and Pope, 2009; DeFond et al., 2011).
The US setting is interesting because of its relative size and inuence on other capital markets, aswell as being one of the few remaining large economies that is still considering mandating IFRS. Likeour study, Amiram (2009) examines country-level investments holdings7 nding an increase in coun-tries that use international accounting standards, where the relationship is stronger if the foreign inves-tor is from countries that also use international accounting standards (IFRS) and in countries with lowercorruption and better investor protection. In contrast, we provide evidence that a countrys use of IFRSalone is not associated with US equity investment. Rather, using IFRS is associated with an increase ininvestment when institutional factors are in place that better align reporting incentives with foreignstock ownership needs. Thus, our focus on country-level aggregate investments by US investors comple-ments this work by providing insights into the economy-wide effects allowing and mandating IFRS.
Second, we complement previous work on the role of legal institutions and security regulationenforcement (e.g. Hope, 2003; Frost et al., 2006; Hail and Leuz, 2006; Daske et al., 2008). We analyzea broad range of factors and isolate legal standards system and the enforcement regimes as two dis-tinct dimensions within a countrys regulatory environment.
Our study should be of interest to companies, standard setters, and regulators. Our nding that for-eign investment will not necessarily increase with the use of IFRS when the regulatory environment isnot strong implies that standards setters and regulators should consider strengthening institutions aswell as the potential benets of IFRS for foreign investment. Similarly, companys benets may belimited, subject to their reporting incentives (Beneish et al., 2011; Yu, 2009).
The remainder of the paper is organized as follows. Section 2 discusses related literature and thehypotheses. Section 3 describes the methodology and variable constructs. Section 4 presents theregression results and robustness checks. Section 5 concludes.
2. Hypotheses development
2.1. Potential benets of IFRS
IFRS has the potential to increase foreign investment by providing high quality nancial reportsand reducing information processing costs. High quality nancial reporting can reduce information
7 Amiram (2009) uses a different investment measure than ours from the IMF CPIS website to measure foreign portfoliotional features are in place that align reporting incentives with foreign investors needs.Our analyses control for country characteristics such as prior accounting and reporting environ-
ment, economic development, and attractiveness of the market to foreign investors. We incorporatetwo distinct aspects of investors information processing costs related to country specic differencesin accounting standards: disclosure and measurement. We nd limited evidence that when disclo-sures, but not measurement choices, are more similar to US GAAP investment is higher. We performseveral robustness checks of model specication, alternative dependent variables, and variable deni-holdings.
asymmetries for outside investors (Saudagaran and Diga, 2000; El-Gazzar et al., 1999; Street andBryant, 2000; Tarca, 2004; Ashbaugh and Pincus, 2001; Gordon et al., 2010). Similar accountingstandards can decrease investors processing costs (Choi and Levich, 1991; Lainez and Callao, 2000;Bradshaw et al., 2004). Investment responses have generally been positive to rms voluntarily usingor conforming to international accounting standards for nancial reporting (Covrig et al., 2007;
about the quality of nancial reporting claiming that in countries with strong investor protection and
484 K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500enforcement, nancial reporting quality under IFRS is not likely to dominate that of domestic stan-dards. For weak investor protection countries, they assert that IFRS likely represents a signicantimprovement in nancial reporting quality, but weak implementation and enforcement mechanismslead investors to discount nancial reports. An alternative view is that benets of increased compara-bility due to a common set of accounting standards accrue to all companies, especially those in stron-ger regulatory environments.
To date, empirical evidence on IFRS adoption and investor allocation is mixed. For instance, Beneishet al. (2011) examine countries that mandatorily adopt IFRS and nd that a countrys adoption of IFRSis associated with increases in investment in companies foreign debt but not equity investments. Yu(2009) examines security-level holdings of international mutual funds in 31 countries from 2000 to2007 to assess the effect of IFRS adoption nding that the adoption of IFRS increases holdings andattributes this to a reduction in information processing cost and increasing comparability of nancialinformation. Similarly, DeFond et al. (2011) examine rm-level mutual fund ownership in 14 EU(mandatory) and 9 non-EU (non-adoption) countries and also nd ownership increases where thereare large improvements in comparability, i.e. in those countries that had standards that were dissim-ilar to IFRS. However, they also do not nd mandating IFRS leads to increased US mutual fund invest-ment for rst time adopters except where rms previously adopted IFRS on a voluntary basis. Florouand Pope (2009) examine rm-level investments in over 24 countries from 2003 to 2006 nding in-creased investment with mandatory IFRS adoption but only in countries with strict legal enforcement,low levels of corruption and earnings management. It is uncertain whether those rm-level resultswill hold for a broad range of countries and for US investors in particular.
As more countries use IFRS,8 an issue is whether standard setters in these countries are herding toIFRS to extract perceived benets of IFRS such as attracting foreign capital.9 So, examining the country-level investment effects provides direct evidence on whether the IFRS adoption is an effective strategy forincreasing overall foreign investment. In general, if US investors nd more value in IFRS reporting, i.e.information processing costs are reduced and/or if IFRS are perceived as being of high quality, thenwe predict that US investment will increase.
Hypothesis 1. US investors allocate more of their equity investment towards countries that use IFRSfor their domestically listed standards.
2.2. Regulatory environment
Adopting international accounting standards may be a relatively easy method of attracting foreigninvestment as opposed to more substantial reform of other institutions in the regulatory environment.Ball et al. (2003) and others have argued that a countrys institutional structures have a strong
8 A recent global survey shows a trend by countries to adopt IFRS. Deloitte website for country use of IFRS (http://www.iasplus.com/country/useias.htm) shows that more than half the 129 countries and territories have adopted IFRS for domestic listingcompanies in some form for their reporting requirements.
9 Herding behavior refers to the phenomenon of decision makers being inuenced by what others are doing, regardless ofAggarwal et al.2005; Bradshaw et al., 2004). Whether requiring IFRS produces similar benets is anopen question.
One view is that mandating IFRS will not reduce home bias. For instance, Beneish and Yohn (2008)discuss uncertainty about nancial reporting quality as one type of information friction relating to for-eign equity investment relevant to IFRS. They assert that IFRS will not signicantly reduce uncertaintyindividual information suggesting something else (Banerjee, 1992, p. 798).
breadth to hold responsible parties accountable. Legal system has been associated with differences
in levels of investor protection, patterns of ownership concentrations, and strength of the capital mar-kets (La Porta et al., 1997, 1998). Rules and regulations are important for capital markets because theyhelp reduce enforcement costs and managements opportunistic behavior by mandating the disclosureof important information.
Furthermore, La Porta et al. (2006) argue that strict lawenforcement and stronger investor protectionlaws increase the likelihood of lawsuits, which carries the threat of monetary losses. Thus, incentives ofreporting parties aremore likely to be alignedwith investors in a country that have strong legal institu-tions. Legal systems and shareholder rights have been linked to more relevant earnings, greater corpo-rate transparency, and lower cost of capital (Ball et al., 2000; DeFond et al., 2007; Bushman et al., 2004;Hail and Leuz, 2006). Similarly, legal system and security regulation proxies have been associated withincreased foreign investment (Leuz et al., 2010; Aggarwal et al., 2005).
We hypothesize that US investors prefer equities from countries with a stronger legal standardssystem, based on the argument that the legal standards system aligns reporting incentives of rmswith outside investors and reduces monitoring costs.
Hypothesis 2a. US investors allocate more of their equity investment towards countries that have astronger legal standards system.
A strong enforcement regime ensures adherence to a countrys laws and other regulations. Impor-tantly, the quality of laws cannot substitute or compensate for the quality of enforcement (LaPorta etal., 1998). In general, effective monitoring and prosecuting rule breakers sends a powerful deterrentsignal to potential offenders because it increases the probability of detection. A strong enforcement re-gime will align the incentives for high quality nancial reporting with those of outside investors thusreduce monitoring costs. Firm level studies have found that auditor quality increases foreign invest-ment (e.g. Bradshaw et al., 2004), and that a strong enforcement environment is associatedwith greateranalyst forecast accuracy and greater capital market development (Hope, 2003; Frost et al., 2006). Wehypothesize that US investment will be greater in countries with a stronger enforcement regime.
Hypothesis 2b. US investors allocate more of their equity investment towards countries that have astronger enforcement regime.
IFRS combined with a strong regulatory environment could increase investor allocation because astrong regulatory environment reduces the risk that those standards will not be complied with. Joshi(1998) ranks enforcement issues as the most important obstacle for the international harmonizationof accounting practice. Therefore, we hypothesize the combination of IFRS and a strong regulatoryenvironment increases US investment because it provides greater incentives for compliance with IFRSand reduces monitoring costs.
Hypothesis 3. US investors allocate more of their equity investments towards countries that use IFRSand have a strong regulatory environment.
3.1. Data and descriptive statistics
Table 1 describes variables and lists data sources. Table 2 summarizes the sample selection. Theinuence on the preparers incentives to produce high- or low-quality nancial reporting. Building onthese themes, we address the importance of the regulatory environment and its interaction withaccounting standards. Our factor analysis, discussed later, identies two distinct factors: the legalstandards system and the enforcement regime.
Legal standards system is interpreted to characterize the way rules are promulgated and their
K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500 485sample consists of observations of US holdings of foreign equities for the years 20032006 in 44
Table 1Variable description and source.
Variable Description Source
DependentEQUITY The total dollar amount of US portfolio
investment in foreign equities, scaled bycountry GDP
US Treasury Department, Federal ReserveBank of New York and the Board ofGovernors of the Federal Reserve Systemsurvey (December 20032006)
IndependentAccounting standardIFRS A countrys relative degree of using IFRS for
domestic reporting requirements (score02)Deloitte Touche and Tohmatsu survey(20022005)
IFRS_ALLOW An indicator variable equal to unity when acountry allows IFRS; zero otherwise
Deloitte Touche and Tohmatsu survey(20022005)
IFRS_MAND An indicator variable equal to unity when acountry mandates IFRS; zero otherwise
Deloitte Touche and Tohmatsu survey(20022005)
Regulatory environmentLegal standards system (LEGAL) A measure of rule making and breadth to
hold responsible parties accountable, basedon a factor analysis of: (1) legal system; (2)liability enforcement; (3) publicenforcement; and (4) anti-director rights
LG_SYS An indicator variable equal to unity forcommon law legal system; zero otherwise
CIA country website
LIAB_ST Measure of procedural difculty inrecovering losses from directors,distributors and accountants (score 010)
La Porta et al. (2006)
PUB_ENF Measures the characteristics, rule-making,investigative and criminally responsiblepower of the security market supervisor(score 010)
La Porta et al. (2006)
ANTI_DIR Measures regulation limiting management/insider actions against minorityshareholders (score 05)
La Porta et al. (1998)
Enforcement regime (ENFORCE) Measures of the ability to implementexisting laws and regulations, based on afactor analysis of: (1) insider trading lawenforcement; (2) judicial efciency; and (3)audit spending
INS_ENF A measure of the existence andenforcement of insider trading laws on orbefore 1995 (score 02)
Bhattacharya and Daouk (2002) as adaptedby Bushman, Piotroski and Smith (2004)
JUD_EFF Assessment of the efciency and integrityof the legal environment as it affectsbusiness (score 010)
La Porta et al. (1998)
AUDIT Dollar sum of total assets or sales auditedby international accounting rms scaled bycountry GNP
Control variables for the reporting environmentADR Measure of visibility based on percentage
of (US Share) investment traded as ADRUS Treasury Department Survey (20032006)
CONF_Meas Measure of familiarity of accountingmeasurement standards to US investorsbased on the country mean of number ofaccounting method choices consistent withUS GAAP divided by total number ofmethod choices disclosed
Bradshaw et al. (2004)
CONF_Disc Measure of familiarity of accountingdisclosures to US investors based on 12country disclosures consistent with USGAAP divided by total number ofdisclosures
Expanded number of countries andadapted based on Young and Guenther(2003) using (2001) A Survey of NationalAccounting Rules Benchmarked againstInternational Account Standards
(continued on next page)
486 K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500
Table 1 (continued)
Variable Description Source
Control variables for economic development and attractiveness to foreign investorsLIST/POP Number of domestic listed companies
scaled by populationWorld Development Indicators for prioryear (The World Bank)
LIQUID Liquidity of capital markets measured bythe total dollar value traded as apercentage of GDP
World Development Indicators for prioryear (The World Bank)
TURNOVER Turnover ratio measured by the dollar World Development Indicators for prior
K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500 487countries obtained from the US Treasury Department survey.10 Sample size is limited due to number ofcountry observations for independent variables LEGAL and ENFORCE, which reduce sample size in mostregressions to 176. Including variables for measurement conformity and disclosure conformity furtherreduces our sample to 152.11 Table 3 presents a summary of dependent and independent variables atthe country level.
The dependent variable, EQUITY, is measured as the total dollar amount of US portfolio investmentin foreign equities as of December 31 for the years 20032006, scaled by country GDP.12 Canada, Fin-land, Hong Kong, Netherlands, Singapore, Switzerland, and the United Kingdom represent the highestequity holdings by US investors. Accounting standards, IFRS, is measured by a countrys ordinal response
value of shares traded as a percentage ofmarket capitalization
year (The World Bank)
INFLAT Average ination rate for consumer pricesfor previous 10 year period
World Development Indicators for prior10-year average (The World Bank)
Table 2Sample selection.
Observations for which US portfolio investment data is available 428Less: missing data for LEGAL and ENFORCE 252Observations for which US investment, IFRS, LEGAL and ENFORCE are available 176Less: missing data for measurement conformity to US GAAP 12Less: missing data for disclosure conformity to US GAAP 12Number of observations in sample with full data available 152on a Deloitte Touche Tohmatsu survey indicating level of IFRS use for domestic reporting for years 20022005. The scale ranges from 0 to 2, with zero being not allowed, one is allowed, and two being requiredfor all entities. Recognizing that countries could allow IFRS or require it, and consistent with prior re-search that nds investment effects for voluntary adoption, we code the choice of a country allowingIFRS separately from requiring it. We also model the choice to allow or mandate IFRS as individual bivar-iate variables, IFRS_ALLOW and IFRS_MAND, respectively.
Because of signicant correlations among regulatory environment variables we employ factor anal-ysis to determine whether these variables represent an underlying construct or constructs. We applyfactor analysis to seven distinct regulatory environment variables and identify two underlying latentconstructs: legal standards system (LEGAL) and enforcement regime (ENFORCE). Identifying twounderlying constructs, rather than seven separate variables, also allows for parsimony in our model.Varimax factor loadings representing the correlation of each variable and the factor are reported alongwith factor scoring in Table 5. Two factors were retained with eigenvalues above one.13
10 Data years 2003 and 2006 are benchmark surveys of all US residents with foreign holdings, while the interim years are basedon larger reporting institutions grossed up to represent full sample. Direct investment where ownership consists of direct orindirect control of 10% or more of voting stock is excluded. Foreign holdings of the largest institutions represent approximately 90%of market value of equity of full benchmark survey.11 We delete regression outliers, so the number of observations reported in Table 6 is lower than 152. Outliers are identied asthose observations with an absolute value of the studentized residual greater than three.12 Ahearne et al. (2004) use an earlier version of this data in their analyses of home bias and US holdings of foreign equities. Wealso test alternatives to the GDP scaler in the robustness section.13 The Cronbachs alpha score, a measure of the coefcient of reliability (Hair et al., 1998), for each factor indicates a reasonablehigh level of internal consistency with both scores exceeding 0.50.
Table 3US investment and nancial reporting environment variables (average) by country.
Country EQUITY (%) IFRS LG_SYS LIAB_ST PUB_ ENF ANTI_DIR INS_ENF JUD_EFF AUDIT ADR CONF_Meas CONF_Disc
Argentina 0.76 0.00 1 0.22 0.58 4 2 6.00 0.214 0.78 0.79 0.73Australia 10.85 0.50 1 0.66 0.90 4 1 10.00 2.152 0.12 0.73 0.82Austria 3.51 2.00 0 0.11 0.17 2 1 9.50 3.101 0.02 0.74 0.64Belgium 5.45 1.25 0 0.44 0.15 0 2 9.50 4.088 0.05 0.78 0.64Brazil 8.44 0.00 0 0.33 0.58 3 2 5.75 0.584 0.61 0.74 0.64Canada 21.22 0.00 1 1.00 0.80 5 2 9.25 2.532 0.01 Chile 3.14 0.00 0 0.33 0.60 5 1 7.25 1.044 0.73 0.78 0.82Colombia 0.41 0.00 0 0.11 0.58 3 1 7.25 0.122 0.70 0.72 Denmark 6.30 1.25 0 0.55 0.37 2 1 10.00 3.252 0.06 0.64 0.64Egypt 2.45 2.00 1 0.22 0.30 2 1 6.50 0.304 0.25 0.52 0.91Finland 22.96 1.25 0 0.66 0.32 3 2 10.00 3.240 0.46 0.68 0.73France 9.89 0.50 0 0.22 0.77 3 2 8.00 4.463 0.14 0.76 0.73Germany 5.59 1.25 0 0.00 0.22 1 2 9.00 4.105 0.11 0.78 0.73Greece 4.09 1.50 0 0.50 0.32 2 1 7.00 1.174 0.04 0.73 0.64Hong Kong 29.54 1.25 1 0.66 0.87 5 2 10.00 2.579 0.07 0.66 1.09India 4.11 0.00 1 0.66 0.67 5 1 8.00 0.212 0.25 0.51 0.82Indonesia 2.74 0.00 0 0.66 0.62 2 1 2.50 0.077 0.10 0.84 1.00Ireland 18.10 0.50 1 0.44 0.37 4 1 8.75 3.213 0.31 0.63 1.00Israel 18.96 0.00 1 0.66 0.63 3 2 10.00 1.883 0.55 0.90 0.73Italy 3.74 0.50 0 0.22 0.48 1 1 6.75 2.354 0.08 0.80 0.64Japan 8.99 0.00 0 0.66 0.00 4 2 10.00 6.473 0.07 0.71 0.82Jordan 0.28 2.00 0 0.22 0.69 1 0 8.66 1.588 0.00 Kenya 0.03 2.00 1 0.44 0.70 3 1 5.75 0.151 0.00 South Korea 11.60 0.00 0 0.66 0.25 2 2 6.00 1.731 0.19 0.64 1.00Malaysia 5.75 0.00 1 0.66 0.77 4 1 9.00 0.767 0.00 0.71 0.91Mexico 7.16 0.00 0 0.11 0.35 1 1 6.00 0.528 0.68 0.71 1.00Netherlands 23.34 1.25 0 0.89 0.47 2 2 10.00 5.290 0.25 0.73 1.00New Zealand 4.53 0.25 1 0.44 0.33 4 1 10.00 1.206 0.17 0.81 1.00Norway 7.75 0.50 0 0.39 0.32 4 2 10.00 2.800 0.09 0.76 0.91Pakistan 0.31 0.00 1 0.44 0.58 5 1 5.00 0.193 0.06 0.62 0.64Peru 0.97 2.00 0 0.66 0.78 3 2 6.75 0.149 0.81 0.71 1.00Philippines 2.17 0.50 0 1.00 0.83 3 1 4.75 0.279 0.22 0.94 0.91
Portugal 3.22 0.50 0 0.66 0.58 3 1 5.50 1.957 0.13 0.71 0.73Singapore 27.25 1.50 1 0.66 0.87 4 2 10.00 3.246 0.04 0.72 1.00South Africa 12.33 1.25 1 0.66 0.25 5 1 6.00 2.192 0.20 0.68 1.00Spain 6.13 0.50 0 0.66 0.33 4 1 6.25 1.570 0.10 0.73 0.64Sri Lanka 0.32 0.50 1 0.39 0.43 3 1 7.00 0.035 0.00 0.65 Sweden 12.03 0.50 0 0.28 0.50 3 2 10.00 3.979 0.12 0.72 0.62Switzerland 49.82 1.00 0 0.44 0.33 2 2 10.00 11.624 0.15 0.68 0.54Thailand 4.89 0.00 0 0.22 0.72 2 2 3.25 0.532 0.04 0.90 0.69Turkey 2.39 0.75 0 0.22 0.63 2 1 4.00 0.301 0.06 0.62 0.62UK 25.00 0.50 1 0.66 0.68 5 2 10.00 4.384 0.26 0.64 0.92Venezuela 0.52 0.50 0 0.22 0.55 1 1 6.50 0.404 0.91 0.86 0.69Zimbabwe 1.33 1.00 1 0.44 0.42 3 0 7.50 0.287 0.00 0.69
See Table 1 for explanations of variables.
Table 4Country level nancial reporting environment summary statistics.
Variable Obs Mean Median Std. dev. Minimum Maximum
Panel A: descriptive statisticsDependentEQUITY 152 0.099 0.059 0.109 0.000 0.716
IndependentIFRS 152 0.664 0.000 0.837 0.000 2.000LEGALLG_SYS 152 0.342 0.000 0.476 0.000 1.000LIAB_ST 152 0.477 0.470 0.231 0.000 1.000PUB_ENF 152 0.501 0.525 0.226 0.000 0.900ANTI_DIR 152 3.000 3.000 1.342 0.000 5.000
ENFORCEINS_ENF 152 1.474 1.000 0.501 1.000 2.000JUD_EFF 152 7.724 8.000 2.208 2.500 10.000AUDIT 152 2.306 1.920 2.251 0.077 11.624
Controls for reporting environmentADR 152 0.243 0.135 0.254 0.000 0.960CONF_Meas 152 0.726 0.725 0.092 0.507 0.941CONF_Disc 152 0.803 0.775 0.150 0.540 1.000
Controls for economic development and attractiveness to foreign investorsLIST 151 26.978 14.221 31.666 1.462 155.143LIQ 152 59.470 45.400 58.533 0.100 269.300TURN 152 74.113 55.500 67.773 0.500 375.700INFLAT 152 7.764 3.100 17.184 2.100 134.100
EQUITY IFRS LG_SYS LIAB_ST PUB_ENF ANTI_DIR INS_ENF JUD_EFF AUDIT ADR CONF_Meas CONF_Disc
Panel B: correlation matrixEQUITY 0.209 0.159 0.270 0.040 0.141 0.461 0.511 0.707 0.113 0.174 0.090
(0.010) (0.050) (0.001) (0.622) (0.084) (0.000) (0.000) (0.000) (0.167) (0.033) (0.272)IFRS 0.154 0.059 0.011 0.142 0.194 0.050 0.220 0.174 0.123 0.189 0.006
(0.059) (0.470) (0.891) (0.081) (0.016) (0.542) (0.006) (0.032) (0.131) (0.020) (0.942)LG_SYS 0.156 0.070 0.202 0.317 0.622 0.129 0.223 0.184 0.025 0.320 0.387
(0.055) (0.389) (0.013) (0.000) (0.000) (0.114) (0.006) (0.023) (0.762) (0.000) (0.000)LIAB_ST 0.324 0.003 0.228 0.266 0.404 0.028 0.133 0.053 0.141 0.036 0.422
(0.000) (0.971) (0.005) (0.001) (0.000) (0.732) (0.101) (0.515) (0.084) (0.665) (0.000)PUB_ENF 0.037 0.205 0.312 0.274 0.344 0.005 0.181 0.348 0.130 0.162 0.178
(0.655) (0.011) (0.000) (0.001) (0.000) (0.952) (0.026) (0.000) (0.112) (0.046) (0.028)ANTI_DIR 0.181 0.214 0.632 0.429 0.355 0.040 0.167 0.102 0.002 0.279 0.329
(0.026) (0.008) (0.000) (0.000) (0.000) (0.623) (0.039) (0.213) (0.985) (0.001) (0.000)INS_ENF 0.497 0.071 0.129 0.057 0.002 0.022 0.365 0.466 0.079 0.118 0.006
(0.000) (0.386) (0.114) (0.483) (0.977) (0.787) (0.000) (0.000) (0.331) (0.148) (0.941)JUD_EFF 0.626 0.232 0.202 0.201 0.101 0.192 0.388 0.621 0.148 0.170 0.041
(0.000) (0.004) (0.029) (0.013) (0.187) (0.018) (0.000) (0.000) (0.069) (0.037) (0.617)AUDIT 0.711 0.250 0.144 0.050 0.338 0.061 0.471 0.726 0.300 0.108 0.205
(0.000) (0.002) (0.076) (0.544) (0.000) (0.458) (0.000) (0.000) (0.000) (0.185) (0.011)ADR 0.005 0.140 0.042 0.020 0.070 0.081 0.096 0.101 0.221 0.108 0.115
(0.956) (0.086) (0.608) (0.810) (0.391) (0.321) (0.238) (0.218) (0.006) (0.185) (0.159)CONF_Meas 0.236 0.190 0.311 0.200 0.090 0.293 0.094 0.111 0.097 0.063 0.095
(0.003) (0.019) (0.000) (0.014) (0.269) (0.000) (0.251) (0.172) (0.234) (0.440) (0.244)CONF_Disc 0.200 0.022 0.394 0.439 0.178 0.329 0.024 0.071 0.103 0.285 0.094
(0.013) (0.788) (0.000) (0.000) (0.028) (0.000) (0.765) (0.384) (0.210) (0.000) (0.247)
In Panel B Pearson (Spearman) correlations are presented above (below) the diagonal of the matrix. Two-tailed p-values are presented in parenthesis.See Table 1 for explanations of variables.
Table 5Regulatory environment measures.
Factor loadings (varimax rotation) Factor scoring coefcients
492 K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500Factor 1 (LEGAL) Factor 2 (ENFORCE) Factor 1 (LEGAL) Factor 2 (ENFORCE)
Panel A: factor analysisLG_SYS 0.6954 0.1070 0.3242 0.0429LIAB_ST 0.5089 0.1745 0.1843 0.0480PUB_ENF 0.4519 0.2641 0.1607 0.0532ANTI_DIR 0.7661 0.0437 0.4072 0.0457INS_ENF 0.0560 0.5244 0.0381 0.1537JUD_EFF 0.1878 0.6896 0.0992 0.2889AUDIT 0.1384 0.8272 0.0827 0.5466Eigenvalue 1.6043 1.5351Cronbachs alpha 0.6667 0.5375
EQUITY IFRS LEGAL ENFORCE ADR CONF_Meas CONF_Disc
Panel B: correlation matrixEQUITY 0.209 0.159 0.201 0.113 0.174 0.090
(0.010) (0.050) (0.013) (0.167) (0.033) (0.271)IFRS 0.154 0.059 0.135 0.123 0.189 0.006
(0.059) (0.470) (0.099) (0.131) (0.020) (0.942)LEGAL 0.209 0.147 0.035 0.001 0.235 0.450
(0.010) (0.070) (0.668) (0.992) (0.004) (0.000)ENFORCE 0.725 0.248 0.019 0.244 0.121 0.107
(0.000) (0.002) (0.821) (0.003) (0.139) (0.188)The rst factor characterizes the way rules are promulgated and their breadth to hold responsibleparties accountable, and is based on four elements: LG_SYS, LIAB_ST, PUB_ENF, and ANTI_DIR.14 Welabel this factor LEGAL. The rst element is legal system (LG_SYS), which has been associated with highlevels of investor protection and greater public disclosures (La Porta et al., 1998; Ball et al., 2000).Liability standard (LIAB_ST) measures the procedural difculty in recovering losses from directors,distributors, and accountants (La Porta et al., 2006). Public enforcement (PUB_ENF), measures thecharacteristics, rule making, investigative and criminally responsible power of the security market super-visor. Stronger regulatory measures against accountants, market supervisors and others increase theirincentives to monitor management actions. The fourth element is a proxy for anti-director rights(ANTI_DIR) and is based on another rules-based index measuring corporate regulations limitingmanagement/insider actions against minority shareholders.
The second factor captures a countrys ability to implement the existing laws and regulations and islabeled the enforcement regime (ENFORCE). Enforcement regime concerns the ability to implementthe existing laws and regulations. It is composed of three variables INS_ENF, JUD_EFF, and AUDIT.The rst variable is insider trading law enforcement (INS_ENF). The second element, judicial efciency(JUD_EFF), assesses the efciency and integrity of the legal environment as it affects business(La Porta et al., 1998, p. 1124). An efcient judiciary system encourages management to followprescribed rules/regulations and should increase investors condence in rm reporting. The last ele-ment is audit spending (AUDIT), and captures the credibility of nancial reporting.
ADR 0.005 0.140 0.129 0.124 0.169 0.115(0.956) (0.086) (0.115) (0.127) (0.038) (0.159)
CONF_Meas 0.236 0.190 0.262 0.099 0.063 0.095(0.003) (0.019) (0.001) (0.225) (0.440) (0.244)
CONF_Disc 0.200 0.022 0.476 0.012 0.285 0.094(0.013) (0.788) (0.000) (0.884) (0.000) (0.247)
In Panel B Pearson (Spearman) correlations are presented above (below) the diagonal of the matrix. Two-tailed p-values arepresented in parenthesis.See Table 1 for explanations of variables.
14 The four measures of the legal standards systemsLG_SYS, LIAB_ST, PUB_ENF, and ANTI_DIRare positively related to eachother and signicant at least at the 1% level (untabulated). Similarly, all three measures of the enforcement environmentINS_ENF,JUD_EFF, and AUDITare positively related to each other and signicant at least at the 0.1% level (untabulated).
Based on prior literature, we control for investors information processing costs related to differ-ences in reporting environments. Specically, we include variables to capture rm visibility, similarityof measurement standards to US GAAP, and similarity of disclosures to US GAAP. Visibility measuresthe awareness that the rm exists and equity shares are available. Our proxy for visibility is the per-centage of the foreign equities in the US portfolio trading as an American Depositary Receipt (ADR).Recognizing investors information processing costs could be related to differences in both accounting
Guenther (2003) that measures whether similar disclosures to US GAAP are required under local
K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500 493GAAP. They argue that countries where nancial accounting environments lead to greater disclosureof value-relevant nancial accounting information (and thus low information costs for foreigninvestors) are more likely to have higher investment capital mobility. Similarly, in countries that havesimilar disclosure requirements to US GAAP, US foreign investment is likely to be greater. Given therecent convergence program between IFRS and US GAAP, to the extent that US GAAP measurementsand disclosures are similar to those under IFRS these variables are also proxies for the differencebetween local GAAP and IFRS. Finally, other variables are included in analysis that control forvariations in a countrys economic development and attractiveness to foreign investment.
Table 4, panel A, presents descriptive statistics for dependent and country nancial reporting envi-ronment variables. Panel B in Table 4 presents Pearson and Spearman correlations and their two-tailedprobability levels. Table 5 present factor analysis and related correlations for the regulatory environ-ment variables.
3.2. Regression model
Based on the hypotheses presented, four models of US portfolio investment allocation are below:
(1) EQUITY = b0 + b1IFRS + b2Controls + e(2) EQUITY = b0 + b1LEGAL + b2Controls + e(3) EQUITY = b0 + b1ENFORCE + b2Controls + e(4) EQUITY = b0 + b1IFRS + b2LEGAL + b3ENFORCE + b4IFRS LEGAL + b5IFRS ENFORCE + b6IFRS
LEGAL ENFORCE + b7Controls + e
Models 1 through 3 test the single variable hypotheses, and Model 4 includes the interactions ofIFRS and regulatory environment constructs.15 We also estimate similar regressions with separate IFRSbivariate variables to capture the relative degree of adoption.
4.1. OLS regression analysis
In relation to H1, results of Model 1 in Table 6, panel A, show that the use of IFRS is not associatedwith higher US foreign investments.16 Further analyses in panel B show a positive relation with
15 In models presented, we cluster standard errors at a country-level as we have multiple observations per country.16 A possible explanation for these differences from previous studies may be differences in proxy choice, as in the Aggarwal et al.study that uses US GAAP or IAS to measure accounting quality. Similarly, the differences in samples may also be driving the results.The sample in Aggarwal et al. is comprised of emerging market equities, and the Covrig et al. study excludes Canadian investments.Although this study tries to control for conformity of a countrys domestic standards with US GAAP, the fact that the US has notadopted IFRS may still have some inuence in investment decisions. Finally, these prior studies used rm-level data. It may bemeasures and disclosures, we include two control variables. The variable capturing the conformity ofmeasurement choices to US GAAP (CONF_Meas) is the similarity of a countrys accounting measure-ment standards that are consistent with US GAAP similar to one used in Bradshaw et al.s (2004).Familiarity with measurement methods reduces information asymmetry by decreasing the informa-tion processing costs associated with foreign nancial reports.
Similarity of disclosures to US GAAP (CONF_Disc) is based on a construct used in Young andpossible that IFRS adoption by rms may proxy for other effects, such as increased voluntary disclosure levels.
Table 6Regression analysis of US portfolio investment and reporting environment.
Variable Predicted sign Model 1 Model 2 Model 3 Model 4 Model 4a Model 4a substitutingIFRS_MAND for IFRS
Panel A: level regressiona
IFRS + 0.013 0.006 0.003 0.000(0.200) (0.256) (0.349) (0.498)
LEGAL + 0.022* 0.011 0.002 0.004(0.093) (0.155) (0.408) (0.264)
ENFORCE + 0.072*** 0.054*** 0.037*** 0.049***
(0.000) (0.001) (0.005) (0.001)IFRS LEGAL + 0.023** 0.006 0.020*
(0.012) (0.205) (0.086)IFRS ENFORCE + 0.034*** 0.030*** 0.035**
(0.007) (0.002) (0.033)IFRS LEGAL ENFORCE + 0.018* 0.013 0.041**
(0.094) (0.121) (0.037)ADR + 0.031 0.034 0.019 0.323 0.035
(0.500) (0.414) (0.601) (0.260) (0.213)CONF_Meas + 0.091 0.072 0.067 0.018 0.016
(0.537) (0.633) (0.420) (0.786) (0.799)CONF_Disc + 0.130 0.078 0.139** 0.078* 0.065
(0.188) (0.455) (0.027) (0.072) (0.260)LIST + 0.000 0.000
(0.319) (0.302)LIQ + 0.001*** 0.001***
(0.002) (0.000)TURN + 0.000* 0.000**
(0.050) (0.014)INFLAT 0.000* 0.000*
(0.057) (0.080)Y04 0.002 0.004* 0.004** 0.002 0.002 0.005
(0.519) (0.085) (0.048) (0.655) (0.762) (0.448)Y05 0.005 0.007 0.012** 0.010* 0.009 0.012
(0.591) (0.472) (0.006) (0.087) (0.136) (0.141)Y06 0.016 0.028** 0.034*** 0.022** 0.021* 0.024**
(0.407) (0.013) (0.000) (0.044) (0.055) (0.047)F-test (Sig) 3.40 3.13 10.98 26.74 18.25 29.26
(0.007) (0.011) (0.000) (0.000) (0.000) (0.000)Adj. R2 0.110 0.135 0.560 0.697 0.775 0.805N 149 149 149 149 148 148
Variable Predicted Sign Model 1* Model 4* Model 4a*
Panel B: IFRS_ALLOW and IFRS_MANDb
IFRS_ALLOW + 0.078** 0.013 0.000(0.035) (0.327) (0.654)
IFRS_MAND + 0.017 0.009 0.003(0.282) (0.300) (0.600)
LEGAL + 0.009 0.004(0.222) (0.445)
ENFORCE + 0.049*** 0.029**
(0.003) (0.006)IFRS_ALLOW LEGAL + 0.044** 0.003*
(0.039) (0.078)IFRS_ALLOW ENFORCE + 0.063** 0.033**
(0.019) (0.038)IFRS_ALLOW LEGAL ENFORCE + 0.003 0.021
(0.457) (0.915)IFRS_MAND LEGAL + 0.043** 0.005**
(0.011) (0.021)IFRS_MAND ENFORCE + 0.063** 0.057**
(0.006) (0.012)IFRS_MAND LEGAL ENFORCE + 0.059*** 0.041**
494 K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500
Table 6 (continued)
Variable Predicted Sign Model 1* Model 4* Model 4a*
K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500 495ADR + 0.025 0.028(0.571) (0.331)
CONF_Meas + 0.080 0.035(0.588) (0.596)
CONF_Disc + 0.202* 0.108**
(0.064) (0.043)LIST + 0.000
(0.111)LIQ + 0.001***
(0.000)TURN + 0.000***
(0.023)Y04 0.001 0.004 0.003
(0.767) (0.543) (0.681)Y05 0.016** 0.018** 0.015*
(0.006) (0.020) (0.085)non-mandatory use of IFRS (IFRS_Allow), but this effect diminishes in subsequent regressions unlikendings of Florou and Pope (2009), Amiram (2009) and Yu (2009) but consistent with DeFond et al.(2011) and Beneish et al. (2011).
In relation to H2a and H2b, results of Models 2 and 3 reveal that a strong enforcement regime(ENFORCE) is positively and signicantly related to an increase in US investments, while there isweaker evidence that a strong legal system (LEGAL) is important.
Models 4 and 4a provide evidence related to H3. Although IFRS and LEGAL variables are no longerindividually signicant, all interaction terms in Model 4 are signicantly related to US investment.Similarly, Amiram (2009) nds an interaction effect with IFRS adoption and two variables capturingdesirable country characteristics: low corruption and strong property rights. Also consistent withour ndings, Florou and Pope (2009) nd increased investment in countries with strict legal enforce-ment, low levels of corruption and earnings management. These ndings are in line with prior studiesthat show that a strong institutional environment will reduce information and monitoring costs asso-ciated with investing and forecasting cash ows (e.g. Hail and Leuz, 2006; Hope, 2003). The LEGALinteraction terms are no longer signicant when other control variables are added in Model 4a.17
Y06 0.039** 0.039*** 0.032***
(0.038) (0.000) (0.005)F-test (Sig) 8.51 23.55 103.20
(0.000) (0.000) (0.000)Adj. R2 0.1739 0.491 0.670N 149 149 148
This table presents results from regression of US portfolio investment in foreign equity (scaled by country GDP) and variouscountry reporting environmental variables. See Table 1 for explanations of variables. One-sided p-values are presented inparentheses for test variables, with signicant ndings (p < 0.10) highlighted in italics. Two-sided p-values are presented inparentheses for control variables, with signicant ndings (p < 0.10) highlighted in italics.In Panel B, regressions include IFRS_ALLOW bivariate (1 = allowed, 0 otherwise) and IFRS_MAND (1 = required, 0 otherwise).
a EQUITY = b0 + b1IFRS + b2LEGAL + b3ENFORCE + b4IFRS * LEGAL + b5IFRS * ENFORCE + b6IFRS * LEGAL * ENFORCE + b7Controls + e.b EQUITY = b0 + b1IFRS_ALLOW + b2IFRS_MAND + b3LEGAL + b4ENFORCE + b5IFRS_ALLOW * LEGAL + b6IFRS_ALLOW * ENFORCE +
b7IFRS_ALLOW * LEGAL * ENFORCE + b8IFRS_MAND * LEGAL + b9IFRS_MAND * ENFORCE + b10IFRS_MAND * LEGAL * ENFORCE + b11-ENFORCE + b8IFRS_MAND * LEGAL + b9IFRS_MAND * ENFORCE + b10IFRS_MAND * LEGAL * ENFORCE + b11Controls + e.
* Signicant at the 10% level.** Signicant at the 5% level.
*** Signicant at the 1% level.
17 Although there is some correlation of the LEGAL interaction term with the LIQ variable, relatively low VIF scores (below threefor all models) indicate that these results do not appear to be driven by multicollinearity.
Table 7Panel analysis of US portfolio investment and reporting environment. EQUITY = b0 + b1IFRS + b2LEGAL + b3ENFORCE + b4IFRS LEGAL + b5IFRS ENFORCE + b6IFRS LEGAL ENFORCE + b7Controls + e.
Variable Predicted sign Panel RE Panel Tobit RE
Model 4 Model 4a Model 4 Model 4a
IFRS + 0.001 0.001 0.001 0.001(0.417) (0.428) (0.403) (0.401)
LEGAL + 0.026*** 0.020** 0.027*** 0.020**
(0.001) (0.020) (0.003) (0.031)ENFORCE + 0.066*** 0.067*** 0.066*** 0.067***
(0.000) (0.000) (0.000) (0.000)IFRS LEGAL + 0.003 0.002 0.023 0.002**
(0.199) (0.259) (0.249) (0.285)IFRS ENFORCE + 0.018*** 0.018*** 0.017*** 0.018***
(0.000) (0.000) (0.000) (0.000)IFRS LEGAL ENFORCE + 0.005 0.005 0.005 0.005
(0.154) (0.150) (0.153) (0.148)ADR + 0.018 0.017
496 K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500The last column in Table 6, panel A, redenes the dependent variable as unity if a country mandates IFRSand zero otherwise. In this regression, the legal system interaction terms become signicant.
To further explore the differences in the signicance of the legal system interaction terms, Table 6,panel B, presents the results using separate IFRS bivariate variables. IFRS_ALLOW is an indicator vari-able that equals unity if IFRS is allowed for nancial reporting, and IFRS_MANDmeasures whether IFRSis required and takes the value of one when IFRS is mandated. Results are generally consistent with theprevious panel except that the interaction term with LEGAL is negative when IFRS is only allowed inModel 4a. This nding indicates that IFRS and the legal system are substitutes when IFRS is allowed;it is not consistent with the view that investors discount nancial reports in weak regulatory environ-ments (Beneish and Yohn, 2008). However, since the sign is opposite to what we predict in H3, atwo-sided test would be more appropriate and the result is less signicant. We also observe a negativeestimated coefcient on the triple interaction term but it is not signicant.
In contrast, for mandatory IFRS both LEGAL and ENFORCE interaction and the triple interactionterms with IFRS remain positive and even increase in signicance. We interpret this as evidence thatIFRS and a strong regulatory environment are complementary, in support of H3.
Results also indicate that greater conformity to US GAAP disclosure (CONF_Disc) is positivelyrelated to investment. However, there is little evidence that ADR listings (ADR) and similarity ofmeasurement to US GAAP (CONF_Meas) signicantly inuence regression results. This nding is
CONF_Meas + 0.050 0.049(0.528) (0.590)
CONF_Disc 0.050 0.072(0.528) (0.233)
Y04 0.004 0.004 0.004 0.004(0.324) (0.298) (0.262) (0.245)
Y05 0.015*** 0.015** 0.015*** 0.018***
(0.000) (0.000) (0.000) (0.008)Y06 0.035*** 0.035*** 0.035*** 0.049***
(0.000) (0.000) (0.000) (0.000)Chi-square (Sig) 234.48 236.33 237.10 243.84
(0.000) (0.000) (0.000) (0.000)R-square/log likelihood 0.6361 0.6575 332.58 333.68N 149 149 149 149
This table presents results from regression of US portfolio investment in foreign equity (scaled by country GDP) and variouscountry reporting environmental variables. See Table 1 for explanations of variables. One-sided p-values are presented inparentheses for test variables, with signicant ndings (p < 0.10) highlighted in italics. Two-sided p-values are presented inparentheses for control variables, with signicant ndings (p < 0.10) highlighted in italics.
Signicant at the 10% level.** Signicant at the 5% level.
*** Signicant at the 1% level.
somewhat consistent with evidence in Amiram (2009) that investors from countries that use IFRS aremore likely to invest in other countries that use IFRS. Specically, we address the conformity of USGAAP to IFRS, nding similarity of disclosures is more relevant for US investment than similar mea-
and MSCI Index Capitalization (MSCI) have been employed as a scale for equity investment in tests of
K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500 497home bias behavior (e.g. Ahearne et al., 2004; Aggarwal et al., 2005). Results (untabulated) usingMCAP, MSCI19 and BIAS (dened as share of US portfolio/Share in world Market Capitalization) as thedependent variable are generally consistent with prior regressions with regard to ENFORCE (p
portfolio allocations in foreign equities. On further examination, we nd that the mandatory use of
the 2011 EAA Annual Congress, 2010 AAA Annual Meeting, and 2009 AAA International Accounting
Section Mid-year Meeting.
Aggarwal, R., Klapper, L., Wysocki, P.D., 2005. Portfolio preferences of foreign institutional investors. Journal of Banking &Finance 29, 29192946.
Ahearne, A.G., Griever, W.L., Warnock, F.E., 2004. Information costs and home bias: an analysis of US holdings of foreign equities.Journal of International Economics 62, 313336.
Amiram, D., 2009. Financial Information Globalization and Foreign Investment Decisions. Working Paper, University of NorthIFRS is attractive to US investment only with a strong regulatory environment (legal standards andenforcement) but may compensate for a weaker legal system when IFRS use is allowed.
We also contribute to the governance literature by examining the wider regulatory environmentand its relation to US investment in foreign equities. Through factor analysis, we identify two dimen-sions of the regulatory environment, legal standards system and enforcement regime. Both factorsindividually and when interacted with IFRS are associated with US foreign investment. The ndingsin this study are robust to alternative variable constructs and the inclusion of additional controls.
Our evidence is consistent with the argument that countries may be herding to require the use ofIFRS as a means of attracting foreign capital, but US investors do not increase their investments basedon IFRS alone. Our results are consistent with the view that IFRS provides value to investors, but itmust be in a setting that ensures the standards will be followed and enforced.
The analyses are subject to three main caveats. First, IFRS development and the regulatory environ-ment is a dynamic process. The US is currently undergoing substantive efforts to converge to IFRS fordomestic nancial reporting. As the differences between US GAAP and IFRS diminish, the reduction ininformation processing costs may inuence US investments and anticipated results. Second, althoughwe perform numerous sensitivity and robustness tests to variable and model specications, someissues like endogeniety and self-selection are difcult to address and could affect results and theirimplications. Third, this study focuses on US equity investing, but does not consider debt nancing.In general, the issues relating to debt nancing may be different than those for equity nancing.
Parts of this paper are based on Dr. Shimas dissertation at the University of Hawaii, Manoa. She isextremely grateful to the members of her dissertation committee, David Yang (Chair), Sharon Cox,Liming Guan, Sang-Hyop Lee and John Wendell.
We gratefully acknowledge comments received from the editor and two anonymous reviewers,from Andrei Filip, Ian Hague, Elaine Henry, Jenice Prather-Kinsey and Jana Raedy and participants atand (5) rm size for 30 largest rms listed in country exchanges.
We investigate the relation between US foreign equity investment, the use of IFRS, and two factorsof the regulatory environment: legal standards system and enforcement regime. We contribution tothe literature by documenting that US investment is positively associated with IFRS only when it iscombined with a strong regulatory environment, specically a strong enforcement regime. That is,accounting standards are not a substitute for a strong enforcement regime when US investors make4.2.3. Other Control VariablesWe include other institutional features as control variables. The following variables were not sig-
nicant in regressions and did not change results (untabulated): (1) development, equal to unity ifcountry is reported as developed in the International Monetary Fund yearbook, (2) the relative burdenof income tax, measured by the highest corporate tax in effect, (3) intensity of a countrys capital con-trols (Edison and Warnock, 2003), (4) variables associated with analyst activity (Chang et al., 2000)
498 K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500Carolina, Chapel Hill.
K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500 499Ashbaugh, H., Pincus, M., 2001. Domestic accounting standards, international accounting standards, and the predictability ofearnings. Journal of Accounting Research 39 (3), 417434.
Ball, R., Kothari, S.P., Robin, A., 2000. The effects of international institutional factors on properties of accounting earnings.Journal of Accounting and Economics 29, 151.
Ball, R., Robin, A., Wu, J.S., 2003. Incentives versus standards: properties of accounting income in four east Asian countries.Journal of Accounting and Economics 36, 235270.
Banerjee, A.V., 1992. A simple model of herd behavior. Quarterly Journal of Economics 107 (3), 797818.Barth, M., Landsman, W., Lang M., Williams, C., 2011. Are International Accounting Standards-based and US GAAP-based
Accounting Amounts Comparable? Working Paper, Stanford Graduate School of Business, University of North Carolina,Chapel Hill, and University of Michigan.
Beneish, M.D., Yohn, T.L., 2008. Information friction and investor home bias: a perspective on the effect of global IFRS adoptionon the extent of equity home bias. Journal of Accounting and Public Policy 27, 433443.
Beneish, M.D., Miller, B.P., Yohn, T.L., 2011. The Effect of IFRS Adoption on Cross-border Investment in Equity and Debt Markets.Working Paper, Indiana University.
Bhattacharya, U., Daouk, H., 2002. The world price of insider trading. The Journal of Finance 107 (1), 75108.Bhattacharya, U., Daouk, H., Welker, M., 2003. The world price of earnings opacity. The Accounting Review 78 (3), 641678.Bradshaw, M.T., Bushee, B.J., Miller, G.S., 2004. Accounting choice, home bias, and US investment in non-US rms. Journal of
Accounting Research 42 (5), 795841.Bushman, R.T., Piotroski, J.D., Smith, A.J., 2004. What determines corporate transparency? Journal of Accounting Research 42 (2),
207252.Chang, J., Khanna, T., Palepu, K., 2000. Analyst Activity Around the World. Working Paper. University of Pennsylvania, and
Harvard Business School.Choi, F.D.S., Levich, R.M., 1991. Behavioral effects of international accounting diversity. Accounting Horizons 5 (2), 113.Central Intelligence Agency (CIA). The Word Factbook 20002007. .Center for International Analysis and Research (CIFAR), 1995. International Accounting and Auditing Trends, fourth ed. CIFAR
Publications Inc., Princeton, NJ.Covrig, V.M., DeFond, M.L., Hung, M., 2007. Home bias, foreign mutual fund holdings, and the voluntary adoption of
international accounting standards. Journal of Accounting Research 45 (1), 4170.Daske, H., Hail, L., Leuz, C., Verdi, R., 2008. Mandatory IFRS reporting around the world: early evidence on the economic
consequences. Journal of Accounting Research 46 (5), 10851142.DeFond, M., Hung, M., Trezevant, R., 2007. Investor protection and the information content of annual earnings announcements:
international evidence. Journal of Accounting and Economics 43 (1), 3767.DeFond, M., Hu, X., Hung, M., Li, S., 2011. The impact of mandatory IFRS adoption on foreign mutual fund ownership: the role of
comparability. Journal of Accounting and Economics 51 (3), 240258.Deloitte Touche Tohmatsu, 20022005. IFRS in Your Pocket. .Edison, H.J., Warnock, F.E., 2003. A simple measure of the intensity of capital controls. Journal of Empirical Finance 10, 81103.El-Gazzar, S.M., Finn, P.M., Jacob, R., 1999. An empirical investigation of multinational rms compliance with international
accounting standards. The International Journal of Accounting 34 (2), 239248.Florou, A., Pope, P., 2009. Mandatory IFRS Adoption and Investor Asset Allocation Decisions. Working Paper. University of
Macedonia and Lancaster University.Frost, C.A., Gordon, E.A., Hayes, A.F., 2006. Stock exchange disclosure and market development: an analysis of 50 international
exchanges. Journal of Accounting Research 44, 437483.Gordon, E., Jorgensen, B., Linthicum, C., 2010. Could IFRS Replace US GAAP? A Comparison of Earnings Attributes and
Informativeness in the US Market. Working Paper. Temple University, University of Colorado, and University of Texas SanAntonio.
Hail, L., Leuz, C., 2006. International differences in the cost of equity capital: do legal institutions and security regulation matter?Journal of Accounting Research 44 (3), 485531.
Hair, J.F., Anderson, R.E., Tatham, R.L., Black, W.C., 1998. Multivariate Data Analysis, fth ed. Prentice Hall, Upper Saddle River,NJ.
Heckman, J., 1979. Sample selection bias as a specication error. Econometrica 47, 153161.Hope, O.K., 2003. Disclosure practices, enforcement of accounting standards and analysts forecast accuracy: an international
study. Journal of Accounting Research 41 (2), 235272.Joshi, P.L., 1998. Empirical research on the international harmonization of accounting standards and practices: survey ndings.
Journal of Financial Management & Analysis 11 (2), 4462.Kaufmann, D., Kraay, A., Mastruzzi, M., 2005. Governance Matters IV: Governance Indicators for 19962004. World Bank Policy
Research Working Paper 3630.La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1997. Legal determinants of external nance. The Journal of Finance
52 (3), 11311150.La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1998. Law and nance. Journal of Political Economy 106 (6), 1113
1155.La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 2006. What works in security laws? The Journal of Finance 61 (1), 132.Lainez, J.A., Callao, S., 2000. The effect of accounting diversity on international nancial analysis: empirical evidence. The
International Journal of Accounting 35 (1), 6583.Leuz, C., Verrecchia, R.E., 2000. The economic consequences of increased disclosure. Journal of Accounting Research 38 (Suppl.),
91124.Leuz, C., Lins, K.V., Warnock, F.E., 2010. Do foreigners invest less in poorly governed rms? Review of Financial Studies 23 (3),
32453285.Levine, R., Zervos, S., 1998. Stock markets, banks, and economic growth. The American Economic Review 88 (3), 537558.Saudagaran, S.M., Diga, J.G., 2000. The institutional environment of nancial reporting in ASEAN. The International Journal ofAccounting 35 (1), 126.
Street, D.L., Bryant, S.M., 2000. Disclosure level and compliance with IASs: a comparison of companies with and without USlistings and lings. The International Journal of Accounting 35 (3), 305329.
Tarca, A., 2004. International convergence of accounting practices: choosing between IAS and US GAAP. Journal of InternationalFinance Management and Accounting 15 (1), 6091.
United States Treasury Department and Federal Reserve Board. December 20042007. United States Portfolio Holdings ofForeign Securities. .
World Bank. World Development Indicators, 20022005. International Bank for Reconstruction and Development/The WorldBank. Washington, DC.
Young, D., Guenther, D.A., 2003. Financial reporting environments and international capital mobility. Journal of AccountingResearch 41 (3), 553579.
Yu, Gwen, 2009. Accounting Standards and International Portfolio Holdings: Analysis of Cross-border Holdings FollowingMandatory Adoption of IFRS. Working Paper. Harvard Business School.
500 K.M. Shima, E.A. Gordon / J. Account. Public Policy 30 (2011) 481500
IFRS and the regulatory environment: The case of U.S. investor allocation choice1 Introduction2 Hypotheses development2.1 Potential benefits of IFRS2.2 Regulatory environment
3 Methodology3.1 Data and descriptive statistics3.2 Regression model
4 Results4.1 OLS regression analysis4.2 Robustness checks and additional analysis4.2.1 Model specification and self selection checks4.2.2 Alternative variable specifications4.2.3 Other Control Variables