The Determinants Of Euro Portfolio Bias Finance Essay
A voluminous literature in finance and economics documents considerable potential gains from diversification of portfolios internationally. In one of the earliest work, Grubel (1968) establishes that the mean-variance of an internationally diversified portfolio is much larger than the portfolio with purely domestic assets. Subsequent research such as, Levy and Sarnat (1970), Solnik (1974b), van Wincoop (1994), Obstfeld (1994) and Canova and Ravn (1996), provides similar evidence in support of an internationally diversified portfolio for high returns and low risk motives.
It is not surprising that these findings are also backed up by benchmark portfolio theories (e.g. Capital Asset Pricing Model of Sharpe (1964)) which recommend well-diversified portfolios of risky assets. Under standard assumptions  , the international version of Capital Asset Pricing Model (ICAPM) of Solnik (1974a) and Sercu (1980) predict that investors should optimally hold assets of each country in proportion to that country’s share in world market portfolio. However, it has been found that investors usually hold disproportionately large share of assets that are issued from domestic market, with considerably less chunk of foreign assets ― a phenomenon commonly known in academic literature as “home bias”.
The evidence on home bias is long-established in empirical work  . For instance, Cooper and Kaplanis (1994) find that home bias in equity portfolios stands at; 63 percent for USA  , 74 percent for UK, 89 percent for Italy, 94 percent for France and 97 percent for Sweden, for the period 1987-89. More recently, Faruqee et al. (2004) find home bias in all 20 developed countries in their sample for the year 1997. Similarly, Chan et al. (2005) document a strong presence of home bias in mutual funds of 26 developing and developed countries. Surprisingly, home bias is even found within a country, as Coval and Moskowitz (1999) observe that mutual fund managers show preference for holding assets of local firms.
The recent empircal finidngs point towards a declining trend in home bias globally, primarily owing to greater financial integration across countries  . However at the same time when home bias is waning, an interesting trend emerged with the advent of european Economic and Monetary Union (EMU): it is noticed that the euro member countries have largely confined their foreign asset holdings within the euro region rather than diversifying to other non-european countries. For instance, around 54 percent of equity investments of EMU countries are invested within the EMU region in 2006, while it amounts to 60 percent in case of debt securities. More strikingly, about 75 percent (66 percent) of debt (equity) investments of EMU members are allocated within the european region. This phenomenon of excessive holdings of euro countries in assets issued from within the region is referred in literature as “euro portfolio bias” which sets the foundation for the present study.
The transformation from home bias to euro bias simply appears to be the outcome of “re-definition of home” for euro members (Foad 2008). Nonetheless this transformation is of much significance owing to the fact that the patterns of cross-border portfolio holdings determine the degree of international risk sharing (i.e. disconnecting consumption from domestic output shocks) and the transmission of financial shocks (Lane 2005). A relevant case in point is the recent sovereign debt crisis in europe that emanates from heavily debt economy of Greece and have raised fears of spreading all around europe, mainly because of excessively large cross-investments between member countries  . Thus due to its potentially large macroeconomic implications, there is a need to investigate the underlying factors that explain the euro bias phenomenon. Though there is no dearth in literature investigating the explanations for home bias, no study to the best of our knowledge has examined the determinants of euro portfolio bias. Given this background, our study endeavors to make a contribution to the regional bias literature, by exploring the determinants of euro bias in equity and bond markets for the period spanning 1997 to 2008.
2.2 Review of Literature
What are the reasons that compel investors to hold disproportionately high share of domestic assets as contrast to textbook portfolio theories? There is a large volume of literature documenting the possible explanations for the investors’ behavior that influence them to turn down benefits from diversifying their portfolios internationally. The foremost reasons put forward in literature are (1) hedging of currency risk, (2) high transaction costs (fixed and variable) associated with foreign assets, (3) limited information about foreign assets, (4) preference for domestic assets as a better hedge against inflation, and (5) moral hazard and enforcement issues. Lewis (1999), Karolyi and Stulz (2002) and more recently Sercu and Vanpée (2007) present an extensive literature survey on home bias phenomenon.
Numerous studies have found information asymmetry as a leading determinant of home bias. Indicators like geographical proximity, language and culture commonalities are generally employed to capture the impact of informational flows on cross-border portfolio holdings. For example, Faruqee et al. (2004) find that information asymmetry, which is proxied by distance, language and phone lines; significantly explain bilateral equity holdings in 23 developed and emerging economies. Interestingly, Coval and Moskowitz (1999) observe that factors like geography, language and culture explain home bias within a country. Moreover, bilateral trade is considered vital in reducing information asymmetry as high trade flows facilitate more informational exchanges between countries. Both in case of bond and equity holdings, Lane (2005) and Lane and Milesi-Ferretti (2005) find bilateral trade and informational linkages, playing a lead role in determining asset allocations.
While theoretically appealing, some other justifications for home bias are not well-recognized in empirical work. For instance, a high transaction cost of foreign asset holdings is often mentioned as the prime reason for preferring domestic assets. However, Tesar and Werner (1995) negate this argument by concluding that foreign equity has a higher turnover rate than domestic equity, implying less transaction costs associated with holding a foreign asset. Faruqee et al. (2004), on the contrary, disagree with these findings and show that the transaction costs related indicators such as phone costs and number of phones lines are major determinants of foreign portfolio holdings. Likewise, the argument that investors prefer domestic equity as it serves a better hedge against inflation is questioned by Cooper and Kaplanis (1994, p. 45), who argue that this motive can hold only “if investors have very high levels of risk tolerance and equity returns are negatively correlated with domestic inflation”.
With the establishment of european Economic and Monetary Union (EMU) in January 1999, there is an expectation of enhanced financial integration among member countries. Lane (2005) and Lane and Milesi-Ferretti (2005) are the first to explicitly pinpoint the presence of excessively large cross-investments in bonds and equity holdings between EMU member countries when compared with other set of countries ― which they dubbed as “euro-area bias”  . These portfolio investments are much larger amongst euro countries than the predictions made after taking into account factors such as, macroeconomic and trade linkages, and cultural commonalities. De Santis and Gérard (2006) obtain similar findings of higher share of euro holdings by european investors, after controlling for other factors. Specifically, the average increase over the global portfolio rise in euro assets is about 12 percent for equity holdings and 22 percent for bond holdings. Moreover, Lane and Milesi-Ferretti (2007), Giofré (2008) and Balli et al. (2010) present strong evidence of euro portfolio bias.
Whether the leading explanations for home bias are relevant in case of euro portfolio bias? As far as information asymmetries are concerned, greater financial integration among EU countries may translate to more informational flows among them, thereby reducing information associated costs. Furthermore, with the establishment of a common Euronext stock exchange  by merging of Paris, Brussels and Amsterdam stock exchanges, the transaction costs are expected to reduce sharply. Foad (2008) points out that the fixed cost of entering into the foreign market may have reduced, because of a single currency. Moreover an obvious affect to EU investors is the elimination of exchange rate risks which some studies have found to be a strong determinant of portfolio decisions. For instance, Fidora et al. (2006) estimate that elimination in exchange rate volatility (from sample mean to zero) decreases home bias in equity by 20 percent and in bonds by a considerable 60 percent. However, Adjaouté et al. (2000) conclude that the disappearance of exchange rate risk in itself is not a leading factor for portfolio readjustments in euro area.
Another interesting aspect of the formation of EMU is the convergence in asset returns, which is confirmed (among others) in a comprehensive study by Baele et al. (2004)  . They find that euro area government bond yields and equity returns are increasingly converging, as a result of a strong influence of common factors compared to local factors. Thus any convergence in asset returns may likely to divert portfolio allocations to non-euro countries. In addition, a related issue is the convergence in euro area prices, which is expected as an outcome of adopting a common monetary policy, greater coordination in business cycles and increasing trade flows among member countries. For euro investors, this price convergence means that domestic assets will no longer have the superiority in providing hedge against inflation when compared with foreign assets. Nevertheless, price convergence in euro area has not been empirically validated. A Deutsche Bundesbank report (March 2009) finds that at maximum there is only marginal price convergence witnessed in euro region. Similarly, Foad (2010) observes an increase in price convergence only for large euro countries, while no change in case of small euro members.
From the aforementioned discussion, it can be gathered that the advent of EMU has probably lead to, reduction in information and transaction costs, elimination of exchange rate risks and, convergence in asset returns and prices (to some extent). All these factors point towards possible reallocations in portfolio holdings, which has not gone unnoticed in research. For instance, comparing pre and post-euro periods, Foad (2008) attributes information asymmetry as the leading factor behind the drastic decline in home bias witnessed in euro area countries  . Similarly, presenting a strong confirmation of portfolio reallocation in the euro region, De Santis and Gérard (2006) find that the leading determinants for the change in portfolio compositions are the anticipated diversification benefits and the motivation to hold optimal share of foreign assets. However, one of the weaknesses of this study is its short time period of analysis (1997 to 2001) which at best can identify the plausible medium term determinants of portfolio reallocation in euro area.
Considering only informational asymmetries as the possible explanation for euro bias phenomenon, Giofré (2008) distinguishes two channels namely, real and financial  , through which informational spillovers to capital markets can influence equity allocations. The study concludes that financial channel predominantly explains the excessive equity holdings by euro investors. In a more comprehensive attempt employing recent data 1997-2007, Balli et al. (2010), identify market capitalization, transaction costs and credit default risk as main determinants, influencing the transformation from home to euro portfolio bias. The authors argue that a decline in default risk as signalled by lower debt to GDP ratio, along with decreasing transaction costs, induce investors to hold more euro based equity and bond securities.
2.3 Motivation for Study
The fact that euro countries have largely confined their foreign asset holdings in the euro region rather than diversifying to other non-regional countries is important for several reasons. Most prominently, the patterns of cross-border portfolio holdings determine the degree of international risk sharing (i.e. disconnecting consumption from domestic output shocks) and the transmission of financial shocks (Lane 2005). For instance, Demyanyk et al. (2008) estimate that the largest effect on risk sharing comes from those assets and liabilities that have been invested outside the EMU region. Another relevant case in point is the recent sovereign debt crisis in europe that emanates from heavily debt economy of Greece and have raised fears of spreading all around europe, mainly because of excessively large cross-investments between member countries  . Also our preliminary findings from the first essay, in which we observe the ineffectiveness of factor income inflow channel in absorbing output shocks; seems to be the outcome of strong presence of euro portfolio bias  . Moreover, the composition of asset holdings of euro countries with other major developed countries is vital in terms of its potential of causing asymmetric affects in the wake of global financial disturbances. As EMU is the most sophisticated attempt to manage an economic bloc, an enquiry on the determinants of cross-border financial holdings will be valuable for other economic blocs that will follow suit.
Given this background, this study will explore: why euro investors are holding large proportion of assets from euro region compared to assets of non-euro countries? Whether the leading explanations for home bias are relevant in case of euro portfolio bias? Is it information asymmetries (Lane and Milesi-Ferretti 2005; Giofré 2008; Foad 2008); or transaction costs (Faruqee et al. 2004; Balli et al. 2010); or portfolio optimization motives (De Santis and Gérard 2006); or a combination of several factors including those that have been overlooked in research so far?  More specifically, the study endeavors to answer these questions by analyzing the determinants of euro portfolio bias in equity and bond markets for the period spanning 1997 to 2008. Though there is no dearth in literature investigating the determinants of home bias, no study to the best of our knowledge has examined the determinants of euro bias, making it a contribution to the regional bias literature. A distinguishing feature of this study is that we will run bilateral panel regressions using regional (euro) bias indicator as a dependent variable, whereas other studies simply use euro area dummy to capture the impact of portfolio allocations in the euro region  .
2.4 Research Method
2.4.1 Measure of Regional Bias
Several empirical studies such as Bertaut and Griever (2004), Ahearne et al. (2004), IMF (2005), De Santis (2006) and De Santis and Gérard (2006), measure home bias by employing foreign asset acceptance ratio (FAAR)  . FAAR in essence quantifies the amount by which actual share of foreign assets deviates from its optimal share as predicted by international capital asset pricing model (ICAPM)  . This ratio () is simply calculated by dividing the actual share of foreign assets in total home country holdings () by the optimal share of foreign assets in total home country holdings (). A ratio of 1 implies no divergence from ICAPM and thus no home bias. Accordingly home bias is calculated as,
Here is computed as the ratio of total foreign portfolio holdings to total portfolio holdings of home country, where the later is obtained by deducting the domestic portfolio holdings by foreigners from total domestic market capitalization and then adding up the total foreign portfolio holdings by home country. The optimal share of foreign holdings is calculated as one minus the home country’s share in the world market portfolio. From the abovementioned home bias measure, the closer the answer is to one, smaller will be the home bias and vice versa  .
The regional portfolio bias () phenomenon, that is the tendency of excessive investments of a country in assets issued by other member countries within the region; is computed by slightly modifying the home bias measure discussed above. Specifically, regional bias () is measured as,
Where is the ratio of home country’s total foreign holdings of assets that are issued by regional member countries, to total asset holdings of home country  ; and the optimal share of regional holdings is estimated as 1 minus the share of all countries outside the region (including home country) in world market portfolio. By employing the aforementioned regional bias measure in case of european Economic and Monetary Union (EMU) countries which is the region under study; a positive outcome will point towards the excess holdings of euro assets by member countries when weigh against the predictions of ICAPM.
2.4.2 The Model
In order to explore the determinants of euro bias in equity and bond markets, we will run bilateral panel regressions using foreign asset acceptance ratio (FAAR) as a dependent variable. Since FAAR measures the extent to which home country’s actual portfolio holdings of a foreign country () deviates from its optimal share (); the dependent variable of our model is the difference between the two (i.e.). This is one of the main contributions of our study as other studies simply use euro area dummy to capture the impact of excessive portfolio allocations in the euro region. Furthermore, given the considerable heterogeneity found in portfolio holdings among euro member countries  , FAAR will be useful to observe the distinguishing patterns of portfolio reallocations since the establishment of EMU.
The following general form of equation will be estimated to investigate the determinants of euro portfolio bias:
where, andare the country-specific fixed effects for home (h) and foreign (f) country. The real Purchasing Power Parity (PPP) adjusted GDP per capita and log linearized population are the proxies used to gauge the country fixed effects of home and foreign country.
is the share of a foreign country in total world market capitalization. This indicator has been established as a crucial determinant of cross-border portfolio holdings: a country with a higher share in global market capitalization generally attracts more foreign investments as it is considered a fairly liquid market with better earning opportunities  .
is debt-to-GDP differential between home and foreign country which is an indicator to gauge the relative repaying capability of a country as the degree of financial indebtedness considerably influences the cross-border investment decisions.
Portfolio Diversification Motives
is a vector to determine the diversification motives of portfolio allocations
From theoretical perspective, a significantly negative sign on these variables suggests that portfolio holdings decisions are influenced by diversification motives. These indicators include:
(1) is the correlation between GDP growth rates between home and foreign country.
(2) is the equity return correlations between home and foreign country.
is a vector that comprise several variables which will be employed to explore the impact of informational flows on euro portfolio bias. The variables include:
(1) is the physical distance between the capital cities of home and foreign country. Distance is commonly used as a proxy for information cost and is expected to be negatively associated with cross-border portfolio holdings.
(2) is a dummy variable, with value of 1 if official languages are same in home and foreign country, otherwise 0. This is a proxy for measuring communication cost and point towards the accumulation of mutual knowledge between countries.
(3)is the per-minute cost of a call from home country to foreign country. Low phone costs between countries imply low information and transaction costs in the financial market.
(4) is a dummy variable, with value of 1 if home and foreign country share the same border, otherwise 0.
(5) is the sum of export share of home country in the foreign country and the share of foreign country’s export in home country. High trade flows facilitate more informational exchanges between countries, thereby reducing information asymmetries and increasing cross-border portfolio transactions.
is a vector to measure institutional quality of a destination country. Academic literature emphasizes the significant role that institutions play in cross-border portfolio allocations as investment decisions are strongly linked to prevalent government stability, rule of law, property rights and degree of perceived corruption. Countries with better institutional quality receives relatively more foreign investments that those with less impressive institutional framework. The institutional measures comprise:
(1)is the corruption perception index of a foreign country.
(2) is the composite country risk index which captures factors like government stability, socioeconomic conditions, investment profile, internal and external conflicts, law and order and bureaucracy quality.
is the well-known Sharpe ratio which is used to evaluate the performance of a portfolio. This ratio indicates whether a portfolio’s return is due to better investment decision or an outcome of an unwarranted risk taken. A higher ratio points to a better risk-adjusted performance and may likely to increase portfolio inflows into a country. It is calculated as the average excess return of a country stock market relative to global returns, divided by the standard deviation of the excess returns variability (Lane and Milesi-Ferretti 2004).
is the measure of bilateral banking integration between home and foreign country.
This indicator is based on Bank for International Settlements (BIS)’s consolidated international banking claims data, which is broken down in terms of instrument, sector, maturity and vis-à-vis country. This data is reported by central banks of respective countries, which in turn compile it from internationally active banks in their countries. Lee (2008) utilizes the same data to analyze the patterns of cross-border asset holdings in case of east asian countries.
2.5 Data Issues
The key data relating to bilateral cross-border equity and debt holdings are obtained from the International Monetary Fund’s Coordinated Portfolio Investment Surveys (CPIS). This database is first available in 1997 covering 29 source countries, and since 2001 it is available on annual basis with the current participation of 75 countries. Specifically, CPIS dataset provides a geographical breakdown of international portfolio holdings based on the residence of the issuer of the securities (for preliminary calculations using CPIS, see Appendix B.2 & B.3). This data is widely used in research because of its reliability and consistency as the surveys are conducted using guidelines that are similar across reporting countries  . Based on the availability of CPIS database, this study will use annual data for the period 1997 and 2001-2008  . The sample countries include twelve EMU countries (i.e. Austria, Belgium, Finland, France, Germany, Greece, Luxembourg, Ireland, Italy, Netherlands, Portugal, and Spain) and three non-EMU countries (namely, Denmark, Sweden and United Kingdom) that are members of European Union (EU). The data sources for other variables are mentioned in Table 2.1.
Need help with your literature review?
Our qualified researchers are here to help. Click on the button below to find out more:
In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: