Amiram (2012).pdf

8/18/2019 Amiram (2012).pdf

1/26

JOURNAL OF INTERNATIONAL ACCOUNTING RESEARCH American Accounting AssociationVol. 11, No. 2 DOI: 10.2308/jiar-502822012pp. 57–81

Financial Information Globalization andForeign Investment Decisions

Dan Amiram

ABSTRACT: This paper investigates the association between the adoption of

international accounting standards and foreign investment decisions. Prior research

suggests that information asymmetries between local and foreign investors and

behavioral biases caused by unfamiliarity of the foreign markets contribute to investors

preferring to invest in their home markets. Because one of the goals of the adoption of

international accounting standards is to establish a high-quality, internationally familiar

set of accounting standards, I predict that foreign investments will increase in countries

that adopted International Financial Reporting Standards (IFRS) after the adoption and

that this increase is driven by the familiarity of IFRS. I find that foreign equity portfolio

investments (FPI) increase in countries that adopt IFRS. More importantly, I find that this

relation is driven by foreign investors from countries that also use IFRS. Moreover, the

effect of accounting familiarity is more pronounced when investor and investee countries

share language, legal origin, culture, and region. I also find that countries with lower

corruption and better investor protection experience larger increases in FPI after they

adopt IFRS relative to other IFRS users. These findings are consistent with thehypothesis that familiar accounting information drives foreign investment decisions.

Keywords: foreign portfolio investments; international accounting; IFRS; familiarity;

information asymmetry; home bias; foreign bias; cross-border investments.

I. INTRODUCTION

I

nformation asymmetries between local and foreign investors and differences in investment

environments are the common explanations for the fact that investors tend to invest in their

domestic market rather than diversify their portfolios with foreign investments. This paper empirically investigates what role, if any, the globalization of one of the most important sources of

financial information, accounting information, plays in cross-border investment decisions.

Dan Amiram is an Assistant Professor at Columbia University.

I thank Jeff Abarbanell, Robert Bushman, Mary Margaret Frank, Wayne Landsman, Mark Lang, Chris Lundblad, KatieMcDermott, Mark Maffett, Ed Owens, Jana Raedy, Mike Welker (editor), Frank Warnock, Chris Williams, Emanuel Zur,and two anonymous reviewers for helpful discussions and comments. This study was conducted as part of the second-year requirements of the doctoral program at The University of North Carolina at Chapel Hill and also benefited fromcomments made by workshop participants there.

Editor’s note: Accepted by Michael Welker.

8/18/2019 Amiram (2012).pdf

2/26

Information asymmetries between local and foreign investors can negatively affect foreign

investments (Brennan and Cao 1997; Portes and Rey 2005; Van Nieuwerburgh and Veldkamp

2009). These asymmetries can help local investors to react with more timeliness and accuracy than

foreign investors and thus to exploit asymmetries at the expense of the foreign investors. This in

turn can cause foreign investors to invest less in the foreign market. Prior research also claims that

familiarity—also known as the competence effect—with the investment environment can cause

investors to feel more competent about the foreign market and thus to increase their investments

abroad (Heath and Tversky 1991; Huberman 2001; Graham et al. 2009). ‘‘Familiarity’’ is defined as

an increase in investor understanding of the business environment, leaving the quality of

information constant. Graham et al. (2009) note that when people feel skillful or knowledgeable in

an area, they would rather bet on their own judgment or invest. But when people do not feel

competent, they prefer not to bet.

In the last decade more than 100 countries around the world have permitted or required the use

of International Financial Reporting Standards (IFRS) as their official accounting system.1 IFRS are

issued by the International Accounting Standards Board (IASB) with the goal of being

internationally accepted, high-quality accounting standards.2 The growing use of IFRS can affect foreign investors’ decisions in at least two ways. First, IFRS are in general of higher quality than

local financial reporting standards (Barth et al. 2008) and thus may reduce information asymmetry

between foreign and local investors and thus increase investment. Second, the use of familiar

accounting standards can increase foreign investors’ confidence in their ability to assess the foreign

market and thus can lead them to invest more in the market. 3

Foreign investments in equity can be divided into two large subgroups—foreign portfolio

investments (FPI) and foreign direct investments (FDI). Foreign portfolio equity investments typically

are defined as a purchase of ownership, but not control, in a domestic corporation by a foreign entity or

an individual. This contrasts with the definition of FDI, which requires control over the acquired

corporation. FPI represents a growing portion of global investment; it is even larger than FDI for U.S.investors in most recent years. I focus my investigation on equity FPI investments because, unlike FDI

investors, FPI investors depend solely on standardized information such as accounting data.

Using a large sample of FPI across countries and years, I find a positive relation between the

adoption of IFRS and foreign portfolio investment. More importantly, I predict and find that

investors from countries that use IFRS increase their FPI in countries that adopt IFRS more than

investors from countries that do not use IFRS. I also find that, after controlling for this accounting

familiarity effect, IFRS has only a small effect on foreign investments. This finding supports the

hypothesis that investors seek a familiar accounting language when making their investment

decisions. In addition, I find that the familiarity of accounting standards acts in a similar fashion to

other familiarity factors such as language, geographical region, legal origin, and colonial past. Thefamiliarity of accounting standards has an incremental effect on FPI over and above these other

1 Leuz and Wysocki (2008) suggest using the worldwide adoption of IFRS as a significant event to test the cross-country and macroeconomics effects of accounting.

2 IAS were issued by the International Accounting Standards Committee (IASC) until 2001. The IASB, thesuccessor to the IASC, issues IFRS, which include standards issued not only by the IASB, but also by the IASC,some of which have been amended by the IASB. My sample includes data from both periods. Throughout thispaper, I use the acronym IFRS to describe both IAS and IFRS.

3 McKinsey Global Investor Opinion Survey (McKinsey & Co. 2002) revealed some of the preferences of international investors for a global set of accounting standards. According to the survey, 90 percent of theparticipants thought a single global set of accounting standards is desirable, and 42 percent of them identifiedinternational accounting standards as a ‘‘very important factor ’’ for investment decisions. This percentageexceeded the percentage of investors who identified market liquidity and property rights laws as important factorsfor investment decisions One purpose of this paper is to evaluate empirically whether investors act in a way that

58 Amiram

8/18/2019 Amiram (2012).pdf

3/26

familiarity factors, and accounting familiarity increases the effects of these other factors. Finally, I

predict and find that countries with lower corruption and better property rights laws experience

higher benefits from the use of IFRS than other adopters. These results are consistent with claims

that accounting standards alone are not enough to achieve positive economic outcomes ( Ball 2006).

The paper makes several contributions to the accounting, finance, and economics literature.

First, it highlights the effect of familiarity of accounting standards on investment decisions. The

findings thus complement findings by Huberman (2001) that familiarity with a corporation’s

headquarters location spurs investments, and Graham et al. (2009) that increased competence

influences trading. It also adds to the evidence on competence effects documented in the behavioral

literature (Heath and Tversky 1991). Second, this study provides evidence that a significant portion

of the increase in foreign investment after the adoption of IFRS is associated with familiarity. This

evidence adds to our knowledge on the ways through which IFRS affects foreign investments.

Contemporaneous studies by Florou and Pope (2012), Yu (2010), DeFond et al. (2011), and

Khurana and Michas (2011) also find an increase in foreign investments following the adoption of

IFRS, but they focus on decreased information asymmetry and increased comparability of

accounting information as drivers of the association between investments and IFRS. Third, contraryto these contemporaneous studies, I use a dataset that contains information on essentially all foreign

portfolio investments in a country and thus provide more general evidence. These contemporaneous

studies focus on subsamples of foreign portfolio investments and thus subsamples of the sample I

use. Some of these studies focus on mutual funds. Some focus on available data on institutional

investors. Others focus on U.S. investors. Because investors from these subsamples may be affected

by different economic forces, these foreign investors may increase investments while other foreign

investors decrease investments. Beneish et al. (2009) contemporaneously examine the effect of

mandatory IFRS adoption on cross-border investment in equity and debt markets using an

aggregated version of the dataset I use. They fail to find evidence that IFRS affects FPI in the equity

market, which may be due to the aggregate level of their tests.The paper continues as follows: Section II provides more information about FPI, IFRS, and the

relations between FPI holdings, home bias, information, and accounting. Section III develops the

hypotheses and empirical tests. Section IV provides detail on the data and sample selection. Section

V describes the empirical results, and Section VI contains sensitivity tests. Section VII concludes.

II. INSTITUTIONAL BACKGROUND AND MOTIVATION

Institutional Background

Foreign equity investments can be divided into two large subgroups—foreign portfolio

investments and foreign direct investment. The International Monetary Fund (IMF) considersinvestments of more than 10 percent of the controlling rights to be FDI and those of less than 10

percent to be FPI. At any point in time, the balance that a domestic resident holds in foreign equity

is considered to be a foreign holding. I examine the consequences of the globalization of accounting

information on FPI rather than FDI for several reasons. First, FPI is not as ‘‘sticky’’ as FDI and

tends to change in reaction to significant market events. Second, foreign direct investors have more

ability to change the activities of the investee firms in ways that will fit their needs. Third and most

importantly, foreign direct investors usually have large holdings in the acquired company and are

better positioned to demand and get the information they need. Foreign portfolio investors, in

contrast, may rely more on accounting data and their own interpretation of this information.

Moreover, as will be discussed below, the bilateral nature of the FPI dataset and the dataset

structure is better suited for examining the familiarity effect between local and foreign investors.

The cost of focusing on FPI is that extra caution is needed when one generalizes the results to

Financial Information Globalization and Foreign Investment Decisions 59

8/18/2019 Amiram (2012).pdf

4/26

Although foreign portfolio investments make important contributions to the recipient

economies and represent a growing part of the total foreign investments (FDI and FPI), there is

little evidence on how accounting information affects foreign portfolio investors’ decisions in a

multinational setting and how the globalization of accounting information affects FPI.4

Motivation

Prior literature refers to the tendency of domestic investors to invest in their home market as home

bias. Several explanations for home bias have been offered, such as barriers to capital flows,

withholding taxes, political risk, regulation, information asymmetries, and familiarity, but there is little

evidence regarding how global integration of financial information can mitigate this phenomenon. The

foreign investment literature relates directly to the home bias literature and uses similar arguments to

explain the barriers to and determinants of foreign investments (Amiram and Frank 2012).

Brennan and Cao (1997) suggest that foreign investors are less informed than locals and thus

react more slowly to market developments. This information asymmetry between foreign and

domestic investors can lead to lower foreign investments because of the foreign investor’sdisadvantage. Portes and Rey (2005) discuss the information relevant to evaluating investments in

financial assets such as corporate equity. They find a strong association between information

proxies such as telephone call traffic, geographical distance, and investment bank presence and

equity portfolio transactions. They also claim that accounting practices and corporate culture,

among other factors, help to mitigate the information asymmetries that cause the bias.

French and Poterba (1991) suggest that ‘‘familiarity effect ’’ shapes foreign investment decisions.

They speculate that investors may invest less in foreign markets because they know less about these

markets and their institutions and firms. Heath and Tversky (1991), Huberman (2001), and Graham et

al. (2009) are good examples of the behavioral explanation for the familiarity effect, which suggests

that people invest in familiar stocks, while ignoring the principles of portfolio theory. Heath and

Tversky (1991) argue that people feel more competent to bet in environments they feel they know more

about. Huberman (2001) extends this idea to financial markets and finds that investors invest in more

familiar stocks. Graham et al. (2009) find that investors who feel more competent have more

internationally diversified portfolios. Taken together, these findings suggest that differences in

information between local and foreign investors and lack of familiarity obstruct foreign investment.

Accounting helps firms to communicate information to investors. More precise and familiar

accounting information may mitigate the information asymmetry and familiarity barriers for

investments. The objective of financial reporting is to provide useful information for investment

decisions (IASC 2001). Accounting standards do not discriminate between information provided to

domestic and foreign investors, which means that domestic and foreign investors should receive the

same information from the firm. However, it is the investor’s responsibility to learn and understandthe accounting standards that are used in a foreign country. Countries can mandate which accounting

standards their domestic firms use, i.e., domestic standards, U.S. generally accepted accounting

standards (GAAP), IFRS, or let the individual firms decide which ones they want to use. Although

each set of standards has advantages and disadvantages, which can result in differences in accounting

quality, an important question is whether such differences can mitigate the asymmetric information

and familiarity barriers and thus increase FPI.5 Ex ante, it is plausible that, because IFRS are the most

4 See Errunza (2001) and Stulz (1999) for reviews on the contributions and costs of foreign investments. Graetz andGrinberg (2003) indicate that FPI has become a larger part of international portfolios. In the U.S., for example, the

market value of FPI was higher than that of FDI in most years since 1990.5 See Barth et al. (2008) for the accounting quality differences of accounting amounts from applying local and

International Accounting Standards (IAS) and Barth et al (2012) for accounting quality differences of accounting

60 Amiram

8/18/2019 Amiram (2012).pdf

5/26

commonly used accounting standards worldwide, their use can increase FPI by reducing information

asymmetry and making the investment environment more familiar to foreign investors.

III. HYPOTHESIS AND EMPIRICAL SPECIFICATION

Adoption Hypothesis

As noted above, information asymmetry and familiarity of the investment environment are

important determinants of cross-border investments. Because the adoption of IFRS can reduce

information asymmetry and increase familiarity, I hypothesize the following:

H1: All else equal, investors from foreign countries hold more equity portfolio investments in

countries that adopt IFRS after the adoption, relative to countries that do not use IFRS.

To examine my hypothesis, I exploit data on the foreign portfolio holdings of 73 countries

from the Coordinated Portfolio Investment Survey (CPIS) of the IMF. As described below, these

data provide the allocation of equity portfolio investment holdings for each of the 73 reporting

(investor) countries in 240 target (investee) countries and territories. This dataset is available for the

year 1997 and for the years 2001 through 2006.6 Using different sources of information, I

determine, as described later, the year that countries in my sample adopted IFRS.

Combining these two datasets and different control variables allows me to estimate Equation

(1) below, which follows the natural log, double fixed-effects specification of Lane and Milesi-

Ferretti (2008). Lane and Milesi-Ferretti (2008) develop an analytical model, based on a

generalization of Obstfeld and Rogoff’s (1996) gravity of trade model, which explains bilateral FPI.

They test their model using the CPIS bilateral data.7 The model suggests that the incentive to hold

foreign portfolio investments comes from the need to hedge against changes in costs of trade

imports. If prices and quantity of imports from a foreign country are increasing, foreign portfolio

investments will ensure that the domestic investor, who is also the importer, can offset some of theincrease in import costs by gaining return on his/her investments. Thus the model establishes a

relation between FPI and imports, and uses log transformation to express this relation in a linear

equation that can be estimated using the standard OLS assumptions. This basic model allows me to

establish the association between foreign portfolio investments and the adoption of IFRS. The

estimating equation is thus:

Log of Equity FPI ijt ¼ b0 þ b1 IFRS INVESTEE jt þ b2k CONTROLSijt

þ FixedEffect INVESTORi þ FixedEffect INVESTEE j þ eit : ð1Þ

The dependent variable in Equation (1), Log of Equity FPI ijt , is the natural log of the level of

equity holdings of country i (investor country) in country j (investee country) at time t .8

IFRSINVESTEE jt , the variable of interest, is an indicator variable that takes the value of 1 if country j

uses IFRS at time t , and 0 otherwise, i.e., it takes the value of 1 in the year of and the years after

6 The data are available at: http://www.imf.org/external/np/sta/pi/cpis.htm. Further description of the dataset can befound in Bertaut and Kole (2004) and Lane and Milesi-Ferretti (2008). Seventy-three (240) residence (source)countries is the maximum number of countries available in the data.

7 The log specification is a direct result of the Lane and Milesi-Ferretti (2008) analytical model. My empiricalspecification follows the study’s empirical specification. Desai and Dharmapala (2011) use a similar empiricalmodel with U.S. Treasury data for U.S. investments. Portes and Rey (2005) also use a similar specification.Amiram and Frank (2012) use the same methodology and data.

8 To keep observations with FPI in country j equal to 0 in the log form, I add 1 to all the FPI data, which isconsistent with Lane and Milesi-Ferretti (2008). Desai and Dharmapala (2011) apply a similar procedure.Inferences are not sensitive to the elimination of the zero observations (more than 4 000 country-year


http://www.imf.org/external/np/sta/pi/cpis.htmhttp://www.imf.org/external/np/sta/pi/cpis.htm

8/18/2019 Amiram (2012).pdf

6/26

8/18/2019 Amiram (2012).pdf

7/26

Based on the Lane and Milesi-Ferretti (2008) model and results, I expect the coefficient on

LIMPORT to be positive. I also expect the degree of openness in the investee country,

OPENESS INVESTEE , to positively explain FPI. However, during my sample period, most economies

are relatively open for capital movement, which can cause the coefficient on this index to be

insignificant. All the risk measures should correlate negatively with FPI. However, because the riskmeasures are highly correlated, it would not be surprising that individual risk coefficients are

indistinguishable from zero. Similarly, collinearity can affect the market cap, LMCAP INVESTEE , and

the GDP, LGDP INVESTEE , variables’ coefficients.

Familiarity Hypothesis

The next step in my analysis is to determine whether familiarity of information is associated with

higher levels of foreign portfolio investments. If the familiarity of IFRS significantly determines

foreign investment decisions, then investors from countries that are familiar with IFRS (i.e., those

countries that use IFRS) will increase their equity portfolio investments in countries that adopt IFRS

more than investors from countries that do not use IFRS. This leads me to the second hypothesis:

H2: All else equal, investors from countries that adopt IFRS have higher holdings of equity

portfolio investments in countries that also adopt IFRS than investors from countries that

do not adopt IFRS.

To test this hypothesis I use a variation of Equation (1) and estimate the following equation:

Log of Equity FPI ijt ¼ c0 þ c1 IFRS INVESTEE jt þ c2 IFRS

INVESTORit þ c3 IFRS

BOTH t

þ c4k CONTROLSijt þ FixedEffect INVESTORi þ FixedEffect

INVESTEE j

þ lit ; ð2Þ

where IFRS INVESTEE , as in Equation (1), is an indicator variable that takes the value of 1 if theinvestee country is an IFRS user, and 0 otherwise, i.e., in the years after the adoption. IFRS INVESTOR

is an indicator variable that equals 1 if the investor country is an IFRS user, and 0 otherwise, i.e., in

the years after the adoption. The variable of interest, IFRS BOTH , is the interaction between

IFRS INVESTEE and IFRS INVESTOR, and it equals 1 when both countries are IFRS users. As in

Equation (1), CONTROLS is a vector of control variables identical to those used in Equation (1).

FixedEffect INVESTOR and FixedEffect INVESTEE are investor and investee country fixed effects.

Standard errors are clustered by investor and year.

Equation (2) tests whether foreign equity portfolio holdings from countries that use IFRS

increase their investments in investee countries that adopt IFRS more than other investor countries.

Finding that c3 . 0 suggests that familiarity of accounting standards plays a role in internationalportfolio investment decisions. Finding c1 ¼ 0 can be interpreted as lack of ability of the use of IFRS to explain FPI after controlling for the familiarity effect. In addition, this test further mitigates

the concern of correlated omitted variables since it is difficult to attribute acceptance of this

hypothesis to institutional factors.

Synergy Hypothesis

The literature suggests that accounting standards alone cannot produce economic outcomes

(e.g., Bushman and Smith 2001; Ball 2006; Barth et al. 2008). This argument can be applied

directly to this study. If a country lacks the institutional environment that allows accounting

standards to achieve their goals, it is less likely that IFRS will increase foreign investments. For

example, if a country is corrupt or has weak property rights laws, it is less likely that foreign


8/18/2019 Amiram (2012).pdf

8/26

hand, in countries with strong institutional environments, the use of IFRS should be more

beneficial. This line of reasoning leads to predicted cross-sectional differences in the effects of the

use of IFRS on foreign investments, which are formally stated in the third hypothesis:

H3: All else equal, countries that have low corruption and good property rights laws will

experience a larger increase in FPI when they adopt IFRS than countries that do not havelow corruption and good property rights law.

To test this hypothesis I estimate the following Equation (3):

Log of Equity FPI ijt ¼ c0 þ c1 IFRS INVESTEE jt þ c2STRONGINST jt

þ c3 IFRS INVESTEE jt

STRONGINST jt þ c4k CONTROLSijt

þ FixedEffect INVESTORi þ FixedEffect INVESTEE j þ lit : ð3Þ

IFRS INVESTEE , as in Equation (1), is an indicator variable that takes the value of 1 if the investee

country is an IFRS user, and 0 otherwise. STRONGINST is a variable that changes with the different

specifications to take the values of the underlying strong institution variable (corruption index or property rights index). The corruption index data are obtained from the Transparency International

website, and the property rights index was obtained from the Heritage Foundation Index of

Economic Freedom website.10 One reason to use the corruption and property rights indices is that

they are time varying and thus allow for time-varying controls for institutional features in the

research design. In addition, they are likely to capture the institutional environment in a country. In

all specifications, STRONGINST will take higher values when there is a stronger institution (high

corruption index means less corruption in the country). The variable of interest is the interaction

between IFRS INVESTEE and STRONGINST . As in Equation (1), CONTROLS is a vector of control

variables identical to those used in Equation (1). FixedEffect INVESTOR and FixedEffect INVESTEE are

investor and investee country fixed effects. Standard errors are clustered by investor and year.Equation (3) tests whether indeed IFRS acts in synergistic fashion with other institutional

features in the economy such as lower corruption and stronger property rights laws. This test

identifies predictable cross-sectional differences in the benefits of the use of IFRS. Finding that c3. 0 provides evidence that supports H3.

IV. DATA AND SAMPLE

Dependent Variable: Foreign Portfolio Investments Data

The International Monetary Fund provides data on worldwide holdings of foreign portfolio

investment at the Coordinated Portfolio Investment Survey (CPIS) website (http://www.imf.org/ external/np/sta/pi/cpis.htm).11 The CPIS reports bilateral data on foreign equity portfolio asset

holdings by the residence of the issuer. The first CPIS was conducted at the end of 1997, when 29

economies participated. Since 2001, the survey has been conducted annually and contains holdings as

of the end of 2006 for 73 source countries. For each source, the survey reports holdings in

approximately 240 destination countries or territories. Participants in the CPIS follow definitions and

classifications that are mutually consistent by following the methodology set out in the IMF Balance

of Payment Manual (IMF 1993).12 Prior research suggests most FPI flows to public corporations

10 The corruption index is available at: http://www.transparency.org/research/cpi/ . The property rights index is

available at: http://www.heritage.org/index/download11 The data were last downloaded on August 3, 2008.12 See further discussion about the benefits and limitations of the data in Lane and Milesi-Ferretti (2008) and Bertaut

64 Amiram

http://www.imf.org/external/np/sta/pi/cpis.htmhttp://www.imf.org/external/np/sta/pi/cpis.htmhttp://www.transparency.org/research/cpihttp://www.heritage.org/index/downloadhttp://www.heritage.org/index/downloadhttp://www.transparency.org/research/cpihttp://www.imf.org/external/np/sta/pi/cpis.htmhttp://www.imf.org/external/np/sta/pi/cpis.htm

8/18/2019 Amiram (2012).pdf

9/26

because it is more difficult for foreign investors to purchase nontraded foreign equities. The CPIS

dataset is recognized as the most valuable and accurate source of foreign portfolio investments data,

and it allows me to examine my research question in a worldwide setting with bilateral relations.

Experimental Variable: IFRS Adoption Data

Determining the year in which a country adopted IFRS is not a trivial task and requires several

research design choices and a few sources of information. My aim is to capture the date in which

foreign investors face a critical mass of IFRS adopters in a foreign country. Thus, in most countries,

it will be the date on which a country mandated the use of IFRS for domestically traded

corporations. In a very few cases, I consider the adoption date to be the date on which a country

permits the use of IFRS and on which there is evidence that supports a significant use of IFRS in

that country. While misclassification of the adoption date is likely to make it harder to find an

association between IFRS adoption and FPI, I conduct several sensitivity tests to alleviate the

concern that the results are driven by misclassification and my research design choices. First, I

randomly choose an adoption date for countries that adopt IFRS in my sample and re-estimate myregression. I find no significant association between IFRS and FPI using this method. This suggests

that the results are not driven by random country characteristics. Second, I change the classification

to only countries that mandatorily adopted IFRS, which are almost all of the countries in my

sample, and find that my inferences below are unchanged.

My main source for the year that a country started using IFRS (adoption date) is Deloitte’s

IAS Plus website.13 The website provides information on 162 countries from which 85 jurisdictions

require the use of IFRS for all listed companies. The vast majority of the IFRS adoption dates were

obtained from the country-specific web pages in the IAS Plus website.14 Other adoption dates were

obtained from the International Federation of Accountants (IFAC) website. This website provides

surveys conducted by IFAC to its 157 members for which they provide information about their

accounting system, and some of them provide information of IFRS year of adoption. The last

source for adoption dates data is the World Bank and the IMF ‘‘Reports of the Observance of

Standards and Codes’’ (ROSC), which provides reviews of the accounting systems of several

countries.15 According to the IAS Plus website, many jurisdictions that maintain their own local

GAAP claim that their local GAAP is ‘‘based on,’’ ‘‘similar to,’’ or ‘‘converged with’’ IFRS. Often,

not all IAS/IFRS have been adopted locally. Since the website does not compare national or

regional accounting standards to IFRS in detail, it reports only on the direct use of IFRS in

individual countries or regions. Direct use means that the basis of preparation note and the auditor’s

report refer to conformity with IFRS. I classify a country as an IFRS adopter if one of the sources of

adoption information that I use indicates that the auditor report in that country refers to compliance

with IFRS. For example, when the countries in the EU adopted IFRS in 2005, the auditor report stated that the financial report was prepared in accordance to IFRS as adopted by the EU. In this

case, the EU countries will be classified as adopters because the audit report refers to IFRS.16

Countries around the world vary in the way they implement IFRS (IAS Plus). Some countries

adopt IFRS for all corporations domiciled in that country. Some adopt IFRS only for public firms,

and some allow certain sectors not to adopt. Because prior research finds that foreign portfolio

investments are concentrated in the largest public corporations in a country, I define the country as

13 See global use of IFRS table at: http://www.iasplus.com/country/useias.htm14 IAS Plus is considered to be the most reliable source of IFRS adoption dates, and the IASB itself relies on it as a

data source (Ramanna and Sletten 2010). The data were collected on June 17, 2008.15 IFAC surveys are available at: http://www.ifac.org/ComplianceAssessment/published_surveys.php. ROSC

reports are available at: http://www worldbank org/ifa/rosc more html


http://www.iasplus.com/country/useias.htmhttp://www.ifac.org/ComplianceAssessment/published_surveys.phphttp://www.worldbank.org/ifa/rosc_more.htmlhttp://www.worldbank.org/ifa/rosc_more.htmlhttp://www.ifac.org/ComplianceAssessment/published_surveys.phphttp://www.iasplus.com/country/useias.htm

8/18/2019 Amiram (2012).pdf

10/26

an adopter if public companies are the ones that adopt IFRS. The quality of implementation also

may vary across countries. If the quality of IFRS varies across countries, this may cause the IFRS

adoption effect to be less pronounced in certain countries. I use this notion to test H2.

In almost all countries, I classify the date of adoption as the date on which the country

mandated the use of IFRS for traded corporations. But a few countries, including Switzerland(2005) and Austria (2002), required mandatory adoption of either IFRS or U.S. GAAP for their

firms. In these cases, when there is evidence in either Worldscope or prior research that a significant

amount of traded corporations adopted IFRS, I used this year as the adoption year. 17 Another

example is Germany, which permitted IFRS before the mandatory adoption in 2005, and prior

research suggests that a significant number of firms voluntarily adopted IFRS (Leuz and Verrecchia

2000). Inferences are unaltered if I use only the dates of mandatory adoption in a country. This is

expected, because very few countries are classified based on this exception.

My final list contains only countries for which I have a date of IFRS adoption and countries that

never adopted IFRS. Countries that use IFRS but the date of adoption is not available were excluded

from the sample. Thus the final list contains 105 countries, 60 that adopted IFRS by the end of my

sample period (2006) and 45 that did not. Table 1 provides the full list of sample countries. 18

Sample Selection

I collect and merge all data on foreign portfolio equity investments available in the CPIS. This

step results in a potential of 122,640 country-year observations ([73 reporting countries in 2006]3

[240 possible destinations] 3 [7 years that the survey was conducted]). Merging the year of

adoption data and control variables, and deleting observations with missing holdings data, results in

a dataset of 19,608 country-year observations. Most of the observations are eliminated because

information about the use of IFRS is not available (68,985 observations), but many of these

observations had missing values anyway. By eliminating observations with missing controlvariables, I obtain the final sample that contains 13,992 country-year observations. Table 2 presents

the final 53 investor and 81 investee countries that are included in my main analysis. The countries

in the sample are heterogeneous: they are located on different continents and have different sizes

and different economic characteristics. Table 3 presents descriptive statistics for the data. In the

sample, foreign holdings comprise, on average, 7 percent of the market capitalization of the

investee country ($545 81 out of $646,185). As noted by Lane and Milesi-Ferretti (2008), thispercentage of holdings suggests that the data capture a significant portion of the total foreign

portfolio investments, unlike other datasets. Generally, for most variables, there is evidence of

skewness, as indicated by differences between sample means and medians. Estimation using natural

log helps to mitigate this problem.

V. RESULTS

Testing the Adoption Hypothesis

Table 4 presents the results of the estimation of Equation (1) using OLS. Model 1 includes

investor and investee fixed effects, while Model 2 does not. As predicted, in both models, there is a

17 If the Worldscope dataset shows that at least 20 percent of the firm-level observations in a country have adoptedIFRS in a certain year, or any other source suggests that there is a ‘‘significant adoption’’ in a country in a certain

year, then I classify that country as an IFRS adopter before the official mandatory adoption date.18 In a few countries, the domestic exchange allows foreign issuers to file financial reports based on IFRS. If this fact

makes domestic investors more familiar with IFRS then it should be harder to detect any foreign investments

66 Amiram

8/18/2019 Amiram (2012).pdf

11/26

positive significant relation between countries that adopt IFRS and holdings of foreign portfolio

investments in that country after the adoption. The coefficient on IFRS INVESTEE is 0.25 with t-

statistic of 4.09 in Model 1, and 0.22 with t-statistic of 2.19 in Model 2.

The risk factor coefficients are not significantly different from zero. This can be the result of the

high correlation between the risk factors. When I aggregate all risk factors into one aggregated

TABLE 1

Adoption of IFRS

Adopters Non-Adopters

Armenia Norway Albania Taiwan

Australia Paraguay Argentina Thailand

Austria Poland Azerbaijan Togo

Bahrain Portugal Bangladesh Tunisia

Belgium Romania Belarus United States

Belize Serbia Benin Uruguay

Bosnia Singapore Bhutan Uzbekistan

Botswana Slovak Republic Brazil Vietnam

Bulgaria Slovenia Brunei Darussalam

Croatia South Africa Burkina Faso

Cyprus Spain Burundi

Czech Republic Sri Lanka Cambodia Denmark Suriname Canada

Estonia Sweden Chile

Finland Switzerland China, P.R.

France Tajikistan Colombia

Germany Tanzania Côte d’Ivoire

Greece Trinidad and Tobago Cuba

Greenland Turkey Ecuador

Guatemala United Arab Emirates Ghana

Hungary United Kingdom Guam

Iceland Venezuela Hong Kong

Ireland India Italy Indonesia

Jamaica Iran

Jordan Iraq

Kazakhstan Israel

Latvia Japan

Liechtenstein Korea

Lithuania Mali

Luxembourg Moldova

Malta Niger

Morocco Pakistan

Namibia Philippines

The Netherlands Russia

New Zealand Saudi Arabia

Nicaragua Syria

The table presents the countries in the sample separated into IFRS adopters and non-adopters. Adopters are countries that effectively adopted IFRS by the end of 2006.


8/18/2019 Amiram (2012).pdf

12/26

TABLE 2

Participant Countries

Country A–K INVESTOR INVESTEE Country L–Z INVESTOR INVESTEE

Argentina X X Latvia X X

Armenia X Lithuania X

Australia X X Luxembourg X X

Austria X X Malta X X

Bahrain X X Moldova X

Bangladesh X Morocco X

Belgium X X Namibia X

Bhutan X The Netherlands X X

Botswana X New Zealand X X

Brazil X X Norway X X

Bulgaria X X Pakistan X X

Canada X X Paraguay XChile X X Philippines X X

China X Poland X X

Colombia X X Portugal X X

Croatia X Romania X X

Cyprus X X Russia X X

Czech Republic X X Saudi Arabia X

Côte d’Ivoire X Serbia X

Denmark X X Singapore X X

Ecuador X Slovak Republic X X

Estonia X X Slovenia X

Finland X X South Africa X XFrance X X Spain X X

Germany X X Sri Lanka X

Ghana X Sweden X X

Greece X X Switzerland X X

Guatemala X Tanzania X

Hong Kong X X Thailand X X

Hungary X X Trinidad and Tobago X

Iceland X X Tunisia X

India X X Turkey X X

Indonesia X X United Arab Emirates X

Iran X United Kingdom X X

Ireland X X United States X X

Israel X X Uruguay X X

Italy X X Uzbekistan X

Jamaica X Venezuela X X

Japan X X Vietnam X

Jordan X

Kazakhstan X X

Korea X X

Total 53 81

The table presents the countries that are used in the main analysis. Investor countries are countries that report their FPI

holdings in the CPIS dataset in at least one year of my sample. Investee countries are countries that at least one investor country reported that it holds FPI in it in at least one year of my sample.

68 Amiram

8/18/2019 Amiram (2012).pdf

13/26

between the aggregated measure of risk and FPI.19 In addition, untabulated results from specifications

in which individual risk factors are included generally generate significant negative coefficients for

each. Some of the coefficients on control variables in the regressions are not significant, probably

because of the high correlation between the variables in the regression. An untabulated joint F-test

shows that control variables as a whole are statistically different from zero. The high R2 (70 percent)

in the full model mitigates some of the concerns about uncontrolled factors.

The results of Equation (1) estimation provide evidence that foreign investors increase their

investments in IFRS reporting countries. These findings may have several explanations, among themthe belief that IFRS standards lead to better-quality accounting information and that investors are now

more familiar with the information that firms that use IFRS provide. The results of the estimation of

Equation (2), which relates to my second hypothesis, may shed additional light on this issue.

Testing the Familiarity Hypothesis

Main Test of the Familiarity Hypothesis

Equation (2) is designed to test whether investors from countries that use IFRS are more

sensitive to the adoption of IFRS in the investee countries than investors from non-IFRS adopting

countries. The results of the tests of this hypothesis are presented in Table 5. In this estimation, the

TABLE 3

Descriptive Statistics

Variable n Mean Median Std. Dev.

Log of equity FPI 13,992 2.68 1.40 2.97

HOLDING 13,992 545 3.04 1,439

IFRS INVESTEE 13,992 0.25 0.00 0.43

IFRS INVESTOR 13,992 0.24 0.00 0.43

IFRS BOTH 13,992 0.13 0.00 0.33

LGDP INVESTEE 13,992 11.67 11.80 1.85

MCAP INVESTEE 13,992 646,185 73,250 2,215,628

LMCAP INVESTEE 13,992 24.64 25.02 2.78

LPOP INVESTEE 13,992 2.86 2.76 1.65

EU INVESTEE 13,992 0.39 0.00 0.49

IMPORT 13,992 2,458 239 10,518

LIMPORT 13,992 5.07 5.48 3.02OPENESS INVESTEE 13,992 1.24 2.00 1.51

CORPSSCORE INVESTEE 13,097 5.72 5.50 2.47

POLRISK INVESTEE 13,992 2.17 2.00 0.78

ECORISK INVESTEE 13,992 2.26 2.00 0.79

LEGRISK INVESTEE 13,992 1.99 2.00 0.89

TAXRISK INVESTEE 13,992 2.04 2.00 0.71

OPERISK INVESTEE 13,992 2.26 2.00 0.79

SECRISK INVESTEE 13,992 2.21 2.00 0.91

The descriptive statistics presented above are taken from a pooled sample of countries from the year 1997 and the years

2001–2006.See Appendix A for detailed variable definitions.


8/18/2019 Amiram (2012).pdf

14/26

variable of interest is IFRS BOTH . As in the estimation of Equation (1), Models 1 and 2 use OLS

estimation. Model 1 includes investor and investee fixed effects, and Model 2 does not. As

expected, in both models there is a positive significant relation between foreign portfolio

investments and the fact that both the investor and investee use IFRS. The coefficients on IFRS BOTH

in Model 1 and 2 are 0.24 and 0.39, respectively, with t-statistics of 2.38 and 2.97, respectively.After controlling for IFRS BOTH , the main effect, i.e., the coefficient on IFRS INVESTEE , is not

statistically different from zero. This is consistent with the notion that a significant portion of the

increase in FPI comes from the familiarity to other IFRS users and not necessarily from IFRS

quality or a decrease in information asymmetry affecting all foreign investors. Equation (2)

estimation results provide evidence that foreign investors seek to invest in countries that employ

accounting standards they are familiar with.

A Closer Look at Familiarity

Findings reported above indicate that investors from countries that use IFRS tend to invest

more in countries that also use IFRS (i.e., the positive coefficient on IFRS BOTH in Equation (2)).

This section compares the behavior of IFRS BOTH to other familiarity factors, such as common

TABLE 4

The Adoption Hypothesis

Variable

Model 1 Model 2

Estimate t Estimate t

IFRS INVESTEE 0.25 4.09*** 0.22 2.19**

LIMPORT 0.22 6.63*** 0.43 9.24***

LGDP INVESTEE 1.26 2.11** 0.12 1.00 LMCAP INVESTEE 0.17 2.00** 0.16 2.34**

LPOP INVESTEE 0.36 0.21 0.22 2.15**OPENESS INVESTEE 0.03 0.77 0.10 2.04**

POLRISK INVESTEE 0.14 1.22 0.20 1.13 ECORISK INVESTEE 0.09 0.76 0.03 0.22 LEGRISK INVESTEE 0.15 1.25 0.72 3.12***

TAXRISK

INVESTEE

0.12 0.56 0.04 0.25OPERISK INVESTEE 0.11 0.92 0.29 1.28SECRISK INVESTEE 0.12 1.57 0.04 0.36

Investor and investee fixed effects Yes No

R2 0.70 0.43

n 13,992 13,992

*, **, *** Indicate significance at the 10 percent, 5 percent, and 1 percent levels, respectively.This table presents the results of the estimation of Equation (1):

Log of Equity FPI ijt ¼ b0 þ b1 IFRS INVESTEE jt þ b2k CONTROLSijt þ FixedEffect

INVESTORi þ FixedEffect

INVESTEE j

þ eit :

The dependent variable is the log of foreign portfolio holdings of country i in country j. IFRS INVESTEE is an indicator variable that takes the value 1 if country j uses IFRS.All other variables are described in Appendix A. Intercept is included but not presented. Model 1 does not include fixedeffect, and Model 2 does.Tested hypotheses are indicated in bold.

70 Amiram

8/18/2019 Amiram (2012).pdf

15/26

(COMCOLONY ), shared language (COMLANG), and shared legal origin (COMLEGORIG). Each

of these four familiarity variables is an indicator variable that takes the value of 1 if the investor and

investee share the familiarity factor. Whether the countries share a common language and whether

they share a common colonizer is obtained from the CEPII website. The legal origin of a country

was obtained from La Porta et al. (2008). To examine the interaction between these other familiarity

factors and the IFRS familiarity factor, I construct a familiarity index ( FAMILIARITYINDEX ). The

familiarity index is the sum of the four familiarity indicator variables as in La Porta et al. (1998).

Results from estimating variations of Equation (2) tabulated in Model 1 of Table 6 show that

IFRS familiarity ( IFRS BOTH ) behaves in a similar way to the other four factors described above

(coefficient of 0.33 with a t-statistic of 4.96), i.e., it has strong positive effect on FPI, over and

above the other familiarity factors and with similar magnitude to the other familiarity factors. In

Model 2, I combine the familiarity effects as described above and show that the FAMILIAR-

ITYINDEX explains the FPI holdings between countries (coefficient of 0.56 with t-statistic of

12.47). The interesting result of this subsection comes from estimating the interaction between the

familiarity index and IFRS BOTH . The coefficient on the interaction between FAMILIARITYINDEX

TABLE 5

The Familiarity Hypothesis

Variable

Model 1 Model 2


IFRS INVESTEE 0.07 0.91 0.03 0.21

IFRS INVESTOR 0.15 2.50** 0.05 0.60 IFRS BOTH 0.24 2.38** 0.39 2.97***

LIMPORT 0.22 6.57*** 0.43 9.22***

LGDP INVESTEE 0.98 1.59 0.13 1.04 LMCAP INVESTEE 0.16 2.08** 0.16 2.29**

LPOP INVESTEE 0.26 0.16 0.23 2.23**OPENESS INVESTEE 0.04 0.94 0.11 2.25**

POLRISK INVESTEE 0.16 1.43 0.20 1.12

ECORISK

INVESTEE

0.08 0.67 0.02 0.16 LEGRISK INVESTEE 0.16 1.43 0.71 3.08***TAXRISK INVESTEE 0.11 0.55 0.06 0.33OPERISK INVESTEE 0.10 0.90 0.28 1.24

SECRISK INVESTEE 0.12 1.58 0.04 0.40

Investor and investee in fixed effects Yes No

R2 0.70 0.43

n 13,992 13,992


Log of Equity FPI ijt ¼ c

0 þ c

1 IFRS INVESTEE

jt þ c

2 IFRS INVESTOR

it þ c

3 IFRS BOTH

it þ c

4k CONTROLS

ijt þ FixedEffect INVESTORi þ FixedEffect

INVESTEE j þ lit :

The dependent variable is the log of foreign portfolio holdings of country i in country j. IFRS INVESTEE is an indicator variable that takes the value 1 if country j uses IFRS. IFRS INVESTOR is an indicator variable that takes the value 1 if country i uses IFRS. IFRS BOTH is an indicator variable that takes the value 1 if both country i and country j use IFRS.All other variables are described in Appendix A. Intercept is included but not presented.Tested hypotheses are indicated in bold.


8/18/2019 Amiram (2012).pdf

16/26

which is consistent with the hypothesis that IFRS familiarity adds to the benefits of other familiarity

effects in increasing FPI. The coefficient on IFRS BOTH remains positive and significant even in the

presence of the familiarity index (coefficient of 0.21 with t-statistic of 2.93).

Testing the Synergy Hypothesis

Model 1 in Table 7 presents the result of the estimation of Equation (3) where STRONGINST is

the corruption index. As predicted, the coefficient on the interaction term STRONGINST

TABLE 6

A Closer Look at Familiarity

Variable

Model

All Factors

1

Familiarity Index

2

Interaction

3

Estimate t Estimate t Estimate t

IFRS BOTH 0.33 4.96*** 0.21 2.93***

FAMILIARITYINDEX 0.56 12.47*** 0.53 11.62***

FAMILIARITYINDEX IFRS BOTH 0.19 2.78***COMREGION 1.26 8.87***

COMCOLONY 0.45 2.92***

COMLANG 0.66 3.89***

COMLEGORIG 0.13 1.47

LIMPORT 0.05 1.74* 0.09 3.03*** 0.09 2.91*** LGDP INVESTEE 1.66 2.67** 1.82 2.92*** 1.51 2.49**

LMCAP INVESTEE 0.18 1.96** 0.22 2.17** 0.18 2.06**

LPOP INVESTEE 0.33 0.20 0.22 0.13 0.08 0.05OPENESS INVESTEE 0.03 0.76 0.03 0.55 0.03 0.69

POLRISK INVESTEE 0.13 1.15 0.10 0.87 0.15 1.37 ECORISK INVESTEE 0.01 0.11 0.05 0.40 0.01 0.13 LEGRISK INVESTEE 0.14 1.22 0.09 0.70 0.14 1.22TAXRISK INVESTEE 0.20 0.95 0.36 1.54 0.20 0.94OPERISK INVESTEE 0.13 1.05 0.11 0.91 0.13 1.06

SECRISK INVESTEE 0.12 1.54 0.12 1.43 0.09 1.22

Investor and investee fixed effects Yes Yes YesR2 0.73 0.72 0.72

n 13,512 13,512 13,512

*, **, *** Indicate significance at the 10 percent, 5 percent, and 1 percent levels, respectively.This table presents the results from estimation of variation of Equation (2).The dependent variable is the log of foreign portfolio holdings of country i in country j. IFRS BOTH is an indicator variablethat takes the value 1 if both country i and country j use IFRS. The familiarity factors are common geographical region of the investor and investee (COMREGION ), shared colonial past (COMCOLONY ), shared spoken language (COMLANG),and same legal origin (COMLEGORIG). Each of these four familiarity variables is an indicator variable that takes thevalue of 1 if the investor and investee share the familiarity factor (i.e., are in the same geographic area, shared colonialpast, share common spoken language, and have the same legal origin). The familiarity index ( FAMILIARITYINDEX ) isthe sum of the four familiarity indicator variables.All other variables are described in Appendix A. Intercept is included but not presented. Model 1 contains all familiarityfactors, Model 2 contains only the familiarity index, and Model 3 includes the familiarity index and IFRS BOTH .Tested hypotheses are indicated in bold.

72 Amiram

8/18/2019 Amiram (2012).pdf

17/26

indeed works in synergy with strong institutions such as low corruption. The second institutional

characteristic is strong property rights. The index takes higher values when a country has better

property rights. Model 2 in Table 7 presents the result of this estimation. The coefficient on the

interaction term STRONGINST IFRS INVESTEE is positive (0.01) and statistically significant (t-statistic of 3.11). This result is evidence of an incremental IFRS adoption effect on FPI in countries

with other strong institutions.

VI. SENSITIVITY TESTS

Self-Selection

My results might suffer from self-selection bias. In particular, countries may choose to use

IFRS for reasons that are correlated with foreign investments and their institutional environment.

TABLE 7

The Synergy Hypothesis

Variable

Model

STRONGINST ¼CORPSSCORE INVESTEE

H1

1

STRONGINST ¼ PRIGHTS INVESTEE

H1

2


IFRS INVESTEE 0.08 0.53 0.22 1.31STRONGINST 0.01 0.11 0.00 0.40

IFRS INVESTEE STRONGINST 0.04 2.03** 0.01 3.11*** LIMPORT 0.26 7.44*** 0.22 6.64***

LGDP INVESTEE 1.24 2.72** 1.25 2.13**

LMCAP INVESTEE 0.32 3.81*** 0.18 2.24** LPOP INVESTEE 0.92 0.69 0.29 0.17OPENESS INVESTEE 0.04 0.96 0.03 0.86

POLRISK INVESTEE 0.17 1.51 0.13 1.16 ECORISK INVESTEE 0.01 0.1 0.10 0.89 LEGRISK INVESTEE 0.01 0.05 0.13 1.17TAXRISK INVESTEE 0.01 0.08 0.17 0.77OPERISK INVESTEE 0.07 0.73 0.09 0.84

SECRISK INVESTEE 0.10 1.51 0.08 1.15

Investor and investee fixed effects Yes Yes

R2 0.71 0.70

n 13,097 13,979


Log of Equity FPI ijt ¼ c0 þ c1 IFRS INVESTEE jt þ c2STRONGINST jt þ c3 IFRS

INVESTEE jt

STRONGINST jt þ c4k X ijt þ wi þ n j þ lit :

STRONGINST in Model 1 takes the values of the corruption index (CORPSSCORE INVESTEE ) in country j. STRONGINST in Model 2 takes the values of the property rights index ( PRIGHTS INVESTEE ) in country k. STRONGINST IFRS INVESTEE

is the interaction between STRONGINST and IFRS INVESTEE .All other variables are defined in Table 5. Intercept is included but not presented.Tested hypotheses are indicated in bold.


8/18/2019 Amiram (2012).pdf

18/26

because of the EU directive. In addition, investment happens at the firm level, and the mandatory

adoption of IFRS is exogenous to the firm. Another mitigating factor is the ability of the research

design to use country fixed effects to significantly reduce the likelihood of endogeneity.

To assess the sensitivity of my results to this potential deficiency, I follow a similar procedure

to Leuz and Verrecchia (2000).

20

I estimate, as a first stage, a selection equation using a Probit model. This stage estimates the probability of a country adopting IFRS based on a selection model

equation that resembles the one used by Hope et al. (2006). In the first stage estimation, I regress

IFRS INVESTEE on LMCAP INVESTEE , LGDP INVESTEE , OPENESS INVESTEE , and country fixed effects

(untabulated). This first stage regression yields a high McFadden R2 (0.72), suggesting that the

model captures well the probability of IFRS use. As in Leuz and Verrecchia (2000), the inverse

Mills ratio is calculated and added as an additional explanatory variable to the estimation of

Equations (1) and (2). The results (untabulated) suggest that the inclusion of the inverse Mills ratio

does not alter my results and the inferences in Tables 4 and 5 hold. Note that I do not have an

exogenous variable that determines IFRS adoption but does not determine FPI.

Changes Specification

An alternative test to the fixed effects approach described in H1 is a changes test. The

predictions of the theory of portfolio holdings are stated in levels of holdings and not in changes,

but nevertheless a test that directly implements a changes specification may shed some additional

light on the effects of IFRS adoption on FPI. To do this, I obtain a subsample of countries that

mandatorily adopted IFRS during my sample period. To isolate other elements that affect FPI

holdings, I first obtain the residuals from estimation of a variation of Equation (1) on the entire

sample that includes only the controls and fixed effects from Equation (1). The residuals from this

estimation contain the log of FPI that is not explained by the independent variables. Because

IFRS INVESTEE is excluded from this regression, the residuals also contain the part of Log FPI that isexplained by IFRS INVESTEE .

I then use a simple univariate analysis of the residuals to examine whether countries have

higher FPI holding in year t þ1 compared to year t 1, where t is adoption year. I keep in the sampleonly observations from investor countries that exist in years t 1 and t þ1. This procedure gives me a matched sample for the same investor in the same adopting country for years t 1 and t þ1. Findings(untabulated) show a significant increase in FPI for countries that adopt IFRS from one year before

the adoption to one year after the adoption.

Alternative Specifications

Tobit Specification

Because the sample contains many bilateral zero observations, i.e., some countries do not

invest at all in some other countries, my sample is truncated and clustered at zero. This fact may

imply a problem with OLS assumptions and suggest that I should use a different estimation

technique such as Tobit.21 Tobit estimation yields similar inferences to OLS estimation, i.e., a

strong significant relation between increase in the holding of foreign portfolio equity investments

and the adoption of IFRS, and a strong effect on FPI when both investor and investee countries

adopt IFRS.

20 Larcker and Rusticus (2010) note that no single method can completely solve this problem.21 See Lane and Milesi-Ferretti (2008) for discussion on the advantages and disadvantages of the use of Tobit

74 Amiram

8/18/2019 Amiram (2012).pdf

19/26

Time Trend

Prior research suggests that FPI are increasing over time. The purpose of this section is to

examine how sensitive the estimation of Equations (1) and (2) is to a possible time trend in the data.

To some extent, it is difficult to explain theoretically the inclusion of a linear time trend in the

specification while economic trends are already captured in other control variables (such as GDP,market capitalization, and imports). To examine the time trend in the data, I obtain the residuals

from the regression used in the Changes Specification section. By construction, these residuals are

orthogonal to the control variables I use. Thus, if the control variables capture the possible trend in

the data, I should not see a trend in the residuals. I plot the mean of the residuals for every year in

my sample. The plot (not presented) shows that, until 2006, there is no trend in the data. There is,

however, an increase in the residual FPI in 2006, which could create biases in my results. This

increase can be explained by the wholesale adoption of IFRS around the world in 2005, as

previously suggested, but also by uncontrolled factors. I deal with this issue in two ways. First, I

add a linear increasing time variable to my estimation to check whether the results of the estimation

of Equations (1) and (2) survive after such inclusion. Second, I delete all observations from 2006and re-estimate Equations (1) and (2). The results (untabulated) of these two tests do not change the

inferences obtained in the main analysis.

Alternative Dependent Variables

Foreign investors may choose their portfolio weight in a foreign country rather than the level of

their investments. This portfolio weight should be scaled by the weight of the foreign country

market capitalization relative to the world market capitalization, which is the theoretical portfolio

weight based on the international capital asset pricing model. Models that do not use this scaling

approach may suffer from scale bias.22 This approach however may introduce significant noise to

the estimation (Barth and Kallapur 1996). An alternative procedure, which addresses this issue, isto include the theoretical portfolio weight as one of the regressors in the estimated equations.

Untabulated results indicate that the inferences reported in prior sections are unchanged using both

methods, albeit statistically weaker.

European Union Effect

Most members of the European Union (EU) adopted IFRS in 2005. Including investor and

investee fixed effects captures, to some extent, the effects of bilateral investment among the EU

countries. To examine the sensitivity of my results to this potential problem, I add two additional

indicator variables to the estimation. The first takes the value of 1 if the investor country is from the

EU, and the second takes the value of 1 if the investee country is from the EU. The results(untabulated) indicate that the inferences from my main analysis, which is the positive and

significant coefficient on IFRS INVESTEE in Equation (1) and on IFRS BOTH in Equation (2), remain

unchanged.

Subsample Analysis

Removing U.S. and U.K. from the Sample

In light of the limitations of the FPI data that were presented above, there is a possibility that

investments in financial centers and from financial centers distort the results. The problem is that

investment that appears to be investments in the U.S. or the U.K. are actually investments in a third


8/18/2019 Amiram (2012).pdf

20/26

country that just flowed through these financial centers. To mitigate the possibility that this data

shortcoming influences my results, I delete all observations that involve the U.S. and U.K., both as

investors and investees. There are significant costs for such elimination, as the U.S. and U.K.

investors account for a large relative proportion of the world portfolio investments and that is in

addition to the loss of approximately 1,500 country-year observations. Untabulated results of theestimations of Equations (1) and (2) after the exclusion of the U.S. and U.K. observations reveal

similar inferences to those presented in Tables 4 and 5.

Removing the Year 1997 from the Sample

As discussed in the CPIS data description, 1997 was the first year the survey was conducted

and contained only 27 participants. To be consistent with prior literature that includes 1997 in the

sample, my main results contain the data from 1997. However, I mitigate the possibility that the

inclusion of the first year of the data, which has a low number of participants and is four years apart

from the main time series of the data, may affect the results by deleting all of the 1997 observations.

Untabulated findings indicate that the exclusion of 1997 has no significant implications for my mainresults, as the coefficients on IFRS INVESTEE in Equation (1) and IFRS BOTH in Equation (2) remain

positive and significantly different from zero.

VII. CONCLUSIONS

The results in this paper provide evidence that accounting information plays a role in foreign

portfolio investment decisions. The results also provide evidence that familiarity of accounting

standards is the source of association between IFRS adoption and foreign investment decisions.

This result can be explained by the ability of the common use of IFRS to increase the competence

of foreign investors by making the investment environment more familiar to them. Moreover, after

controlling for accounting familiarity, I find no increase in FPI following the adoption of IFRS. I

also find that IFRS familiarity is associated with foreign investments in a similar fashion to other

familiarity factors, such as shared geographical region, shared colonizer, shared spoken language,

and common legal origin. The IFRS familiarity effect interacts with these other factors to enhance

investments, and has incremental effects on investments over and above these other factors.

Moreover, the results show IFRS have a greater effect on FPI in investee countries with lower

corruption and stronger property rights.

Although the results survive a wide variety of sensitivity tests, there are several other

explanations that are consistent with them. For example, IFRS users may adopt the standards as part

of a variety of reforms that aim to increase levels of foreign portfolio investments. The

heterogeneous nature of my sample, the inclusion of openness index, and the use of the Heckman

procedure help mitigate this concern. An additional concern is that it is more likely for countries to

adopt IFRS if other countries also adopt the standards (Ramanna and Sletten 2010). It is unlikely

that this issue affects my results because the tests provide evidence that there is a significant

increase in FPI after the adoption of IFRS and that other familiarity factors are likely to capture

these effects. An additional alternative explanation is that IFRS is seen by investors as a set of

accounting standards that provides financial reports with higher quality. However, for the results of

the tests of my second hypothesis to be consistent with this explanation, investors from countries

that adopt IFRS would have to be more affected by better quality information than investors from

non-IFRS countries. Although I do not test for this possibility, there is no reason to believe this is

the case.

If one takes the popular position that FPI are beneficial to the recipient countries, then my

76 Amiram

8/18/2019 Amiram (2012).pdf

21/26

global set of accounting standards. The results also give additional insight about which factors

affect global investors and shed additional light on explanations of the home bias phenomenon.

REFERENCESAmiram, D., and M. M. Frank. 2012. The Effects of the Taxation of Dividends on the Allocation of Foreign

Portfolio Investment Around the World . Working paper, Columbia University.

Ball, R. 2006. International Financial Reporting Standards (IFRS): Pros and cons for investors. Accounting

and Business Research 36 (1).

Barth, M. E., W. R. Landsman, and M. H. Lang. 2008. International Accounting Standards and accounting

quality. Journal of Accounting Research 46 (3): 467–498.

Barth, M. E., W. R. Landsman, M. H. Lang, and C. D. Williams. 2012. Are IFRS-based and U.S. GAAP-

based accounting amounts comparable? Journal of Accounting and Economics 54 (1): 68–93.

Beneish, M., B. Miller, and T. Yohn. 2009. The Effect of IFRS Adoption on Cross-Border Investment in

Equity and Debt Markets. Working paper, Indiana University.

Bertaut, C. C., and L. S. Kole. 2004. What makes investors over or underweight? Explaining international

appetites for foreign equities. Available at: http://www.federalreserve.gov/pubs/ifdp/2004/819/

ifdp819.pdf

Brennan, M. J., and H. H. Cao. 1997. International portfolio investment flows. The Journal of Finance LII

(5): 1851–1880.

Bushman, R. M., and A. J. Smith. 2001. Financial accounting information and corporate governance.

Journal of Accounting and Economics 32 (1): 237–333.

Chinn, M. D., and H. Ito. 2008. A new measure of financial openness. Journal of Comparative Policy

Analysis 10 (3): 309–322.

DeFond, M., X. Hu, M. Hung, and S. Li. 2011. The impact of mandatory IFRS adoption on foreign mutual

fund ownership: The role of comparability. Journal of Accounting and Economics 51 (3): 240–258.Desai, M. A., and D. Dharmapala. 2011. Taxes, dividends and international portfolio choice. Review of

Economics and Statistics 93 (1): 266–284.

Errunza, V. 2001. Foreign portfolio equity investments, financial liberalization, and economic development.

Review of International Economics 9 (4): 703–726.

Florou, A., and P. Pope. 2012. Mandatory IFRS adoption and investor asset allocation decisions. The

Accounting Review 87 (6).

French, K. R., and J. M. Poterba. 1991. Investor diversification and international equity markets. The

American Economic Review 81 (2): 222–226.

Graetz, M. J., and I. Grinberg. 2003. Taxing international portfolio income. Tax Law Review 56: 537–586.

Graham, J. R., C. R. Harvey, and H. Huang. 2009. Investor competence, trading frequency, and home bias.

Management Science 55 (7): 1094–1106.Heath, C., and A. Tversky. 1991. Preferences and beliefs: Ambiguity and competence in choice under

uncertainty. Journal of Risk and Uncertainty 4 (1): 5–28.

Hope, O.-K., J. Jin, and T. Kang. 2006. Empirical evidence on jurisdictions that adopt IFRS. Journal of

International Accounting Research 5: 1–20.

Huberman, G. 2001. Familiarity breeds investment. Review of Financial Studies 14 (3): 659–680.

International Monetary Fund (IMF). 1993. International Monetary Fund, Balance of Payments Manual . 5th

edition. Washington, DC: IMF.

Khurana, I. K., and P. N. Michas. 2011. Mandatory IFRS adoption and the U.S. home bias. Accounting

Horizons 25 (4): 729–753.

La Porta, R., F. Lopez-de-Silanes, and A. Shleifer. 1998. Law and finance. The Journal of Political Economy 106 (6): 1113–1155.

La Porta, R., F. Lopez-de-Silanes, and A. Shleifer. 2008. The economic consequences of legal origins.


http://dx.doi.org/10.1080/00014788.2006.9730040http://dx.doi.org/10.1080/00014788.2006.9730040http://dx.doi.org/10.1111/j.1475-679X.2008.00287.xhttp://dx.doi.org/10.1016/j.jacceco.2012.03.001http://www.federalreserve.gov/pubs/ifdp/2004/819/ifdp819.pdfhttp://www.federalreserve.gov/pubs/ifdp/2004/819/ifdp819.pdfhttp://dx.doi.org/10.1111/j.1540-6261.1997.tb02744.xhttp://dx.doi.org/10.1016/S0165-4101(01)00027-1http://dx.doi.org/10.1080/13876980802231123http://dx.doi.org/10.1080/13876980802231123http://dx.doi.org/10.1016/j.jacceco.2011.02.001http://dx.doi.org/10.1162/REST_a_00073http://dx.doi.org/10.1162/REST_a_00073http://dx.doi.org/10.1111/1467-9396.00308http://dx.doi.org/10.1287/mnsc.1090.1009http://dx.doi.org/10.1007/BF00057884http://dx.doi.org/10.2308/jiar.2006.5.2.1http://dx.doi.org/10.2308/jiar.2006.5.2.1http://dx.doi.org/10.1093/rfs/14.3.659http://dx.doi.org/10.2308/acch-50075http://dx.doi.org/10.2308/acch-50075http://dx.doi.org/10.1086/250042http://dx.doi.org/10.1086/250042http://www.federalreserve.gov/pubs/ifdp/2004/819/ifdp819.pdfhttp://www.federalreserve.gov/pubs/ifdp/2004/819/ifdp819.pdfhttp://dx.doi.org/10.1086/250042http://dx.doi.org/10.1086/250042http://dx.doi.org/10.2308/acch-50075http://dx.doi.org/10.2308/acch-50075http://dx.doi.org/10.1093/rfs/14.3.659http://dx.doi.org/10.2308/jiar.2006.5.2.1http://dx.doi.org/10.2308/jiar.2006.5.2.1http://dx.doi.org/10.1007/BF00057884http://dx.doi.org/10.1287/mnsc.1090.1009http://dx.doi.org/10.1111/1467-9396.00308http://dx.doi.org/10.1162/REST_a_00073http://dx.doi.org/10.1162/REST_a_00073http://dx.doi.org/10.1016/j.jacceco.2011.02.001http://dx.doi.org/10.1080/13876980802231123http://dx.doi.org/10.1080/13876980802231123http://dx.doi.org/10.1016/S0165-4101(01)00027-1http://dx.doi.org/10.1111/j.1540-6261.1997.tb02744.xhttp://dx.doi.org/10.1016/j.jacceco.2012.03.001http://dx.doi.org/10.1111/j.1475-679X.2008.00287.xhttp://dx.doi.org/10.1080/00014788.2006.9730040http://dx.doi.org/10.1080/00014788.2006.9730040

8/18/2019 Amiram (2012).pdf

22/26

Lane, P. R., and G. M. Milesi-Ferretti. 2008. International investment patterns. The Review of Economics

and Statistics 90 (3): 538–549.

Larcker, D. F., and T. O. Rusticus. 2010. On the use of instrumental variables in accounting research.

Journal of Accounting and Economics 49 (3): 186 –205.

Leuz, C., and R. E. Verrecchia. 2000. The economic consequences of increased disclosure. Journal of

Accounting Research 38 (Supplement): 91–124.Leuz, C., and P. D. Wysocki. 2008. Economic Consequences of Financial Reporting and Disclosure

Regulation: A Review and Suggestions for Future Research. Working paper, University of Chicago

and Massachusetts Institute of Technology.

McKinsey and Co. 2002. Global Investor Opinion Survey: Key Findings. New York, NY: McKinsey and

Co.

Obstfeld, M., and K. S. Rogoff. 1996. Foundations of International Macroeconomics. Cambridge, MA:

MIT Press Books.

Petersen, M. A. 2009. Estimating standard errors in finance panel data sets: Comparing approaches. Review

of Financial Studies 22 (1): 435–480.

Portes, R., and H. Rey. 2005. The determinants of cross-border equity flows. Journal of International

Economics 65 (2): 269–296.Ramanna, K., and E. Sletten. 2010. Why Do Countries Adopt International Financial Reporting Standards?

Working paper, Harvard Business School.

Stulz, R. M. 1999. International portfolio flows and security markets. In International Capital Flows, edited

by M. Feldstein. Chicago, IL: The University of Chicago Press.

Van Nieuwerburgh, S., and L. Veldkamp. 2009. Information immobility and the home bias puzzle. The

Journal of Finance 64 (3): 1187–1215.

Yu, G. 2010. Accounting Standards and International Portfolio Holdings: Analysis of Cross-Border

Holdings Following Mandatory Adoption of IFRS. Working paper, University of Michigan.

78 Amiram

http://dx.doi.org/10.1016/j.jacceco.2009.11.004http://dx.doi.org/10.2307/2672910http://dx.doi.org/10.2307/2672910http://dx.doi.org/10.1093/rfs/hhn053http://dx.doi.org/10.1093/rfs/hhn053http://dx.doi.org/10.1016/j.jinteco.2004.05.002http://dx.doi.org/10.1016/j.jinteco.2004.05.002http://dx.doi.org/10.1111/j.1540-6261.2009.01462.xhttp://dx.doi.org/10.1111/j.1540-6261.2009.01462.xhttp://dx.doi.org/10.1111/j.1540-6261.2009.01462.xhttp://dx.doi.org/10.1111/j.1540-6261.2009.01462.xhttp://dx.doi.org/10.1016/j.jinteco.2004.05.002http://dx.doi.org/10.1016/j.jinteco.2004.05.002http://dx.doi.org/10.1093/rfs/hhn053http://dx.doi.org/10.1093/rfs/hhn053http://dx.doi.org/10.2307/2672910http://dx.doi.org/10.2307/2672910http://dx.doi.org/10.1016/j.jacceco.2009.11.004

8/18/2019 Amiram (2012).pdf

23/26

A P P E N D I X

A

V a r i a b l e s D e fi n i t i o n s

r i a b l e

N a m e

S h o r t D e s c r i p t i o n

D e s c r i p t i o n

r i a b l e s

R e l a t e d t o t h e D e p e n d e n t V a r i a b l e

H O L D I N G

H o l d i n g s o f

e q u i t y f o r e i g n p o r t f o l i o

i n v e s t m e n

t s b y i n v e s t o r i n t h e i n v e s t e e

T h e l e v e l i n m i l l i o n s o f U S D o f h o l d i n g s o f i n v e s t o r c o u n t r y r e s i d e n t s o f e q u i t y

f o r e i g n p o r t f o l i o i n v e s t m e n t s i n t h e i n v e s t e e c o u n t r i e s . S o u r c e : I M F C P I S w e b s i t e .

H O L D I N G P 1

H O L D I N G p l u s 1

E q u a l s H O L D I N G p l u s 1 . T h i s a d d i t i o n o f 1 h a s b e e n m a d e i n o r d e r

n o t t o l o s e

o b s e r v a t i o n s w i t h t h e v a l u e 0 w h

i l e c a l c u l a t i n g t h e n a t u r a l l o g o f h

o l d i n g . S e e

L a n e a n d M i l e s i - F e r r e t t i ( 2 0 0 8 ) . S o u r c e : I M F C P I S w e b s i t e .

L o g o f E q u i t y F P I

N a t u r a l l o g o f H O L D I N G P 1

E q u a l s t h e n a t u r a l l o g o f H O L D I N G

p l u s 1 . T h i s a d d i t i o n o f 1 h a s b e e n m a d e i n

o r d e r n o t t o l o s e o b s e r v a t i o n w i t h t h e v a l u e 0 w h i l e c a l c u l a t i n g t h e

n a t u r a l l o g o f

h o l d i n g . S e e L a n e a n d M i l e s i - F e r r e t t i ( 2 0 0 8 ) . S o u r c e : I M F C P I S w

e b s i t e .

I N V E S T

O R

I n v e s t o r c o u

n t r y

N

a m e o f i n v e s t o r c o u n t r y i n t h e I M

F C P I S d a t a s e t — T h e s e a r e t h e c o

u n t r i e s w h o s e

r e s i d e n t s h o l d e q u i t y f o r e i g n p o r t f o l i o i n v e s t m e n t s i n t h e i n v e s t e e c o u n t r y

r e s i d e n t s .

I N V E S T

E E

I n v e s t e e c o u

n t r y

N

a m e o f i n v e s t e e c o u n t r y i n t h e I M

F C P I S d a t a s e t — T h e s e a r e t h e c o

u n t r i e s w h o s e

r e s i d e n t s ’ e q u i t i e s a r e b e i n g h e l d

b y t h e i n v e s t o r c o u n t r y r e s i d e n t s .

p e r i

Amiram (2012).pdf

Documents