Benefits of international equity portfolio diversification in a post-crisis scenario

(1)

University of Amsterdam - Amsterdam Business School

Master in International Finance

Master Thesis

Benefits of international equity portfolio diversification in a post-crisis scenario

Rafael Lopes de Albuquerque

Student Number: 11150874

Thesis Supervisor: Dr. Rafael Matta

(2)

ABSTRACT

This study examines and compares regional and global diversification benefits from the perspective of 69 local equity investors – 23 from developed, 23 from emerging-market and 23 from frontier-market countries. We analyze the statistical significance of the diversification possibilities using the “spanning” technique developed by Huberman and Kandel (1987), as well as the economic size of these possibilities, comparing increase in Sharpe ratios and decrease in volatilities after expanding investor’s portfolios regionally and globally. We also assess the effects of differences in country risk and the other country-level characteristics on the gains obtained with global diversification. Empirically, we find that on average, local investors can still slightly improve their mean-variance efficiency and considerably reduce portfolio volatility through international diversification strategies. In general, the benefits are more noticeable in the reduction of volatility. From a geographic perspective, regional and global diversification benefits appear largest on average for countries in Eastern Europe and for the Western Europe countries which suffered the most during the global financial crisis (Portugal, Ireland, Italy, Greece and Spain). The evidence for diversification benefits is strong when market frictions are excluded, but substantially reduced when investors face more realistic restrictions such as home bias and short sales constraints. Our empirical results also show that, from the variables analysed, country risk appears to be the only good determinant of diversification benefits. This finding suggests that the international diversification benefits that could be attained are larger for riskier countries, particularly from emerging and frontier markets. We conclude that the gradual integration of world equity markets we have been experiencing in past decades is reducing the diversification benefits of investing abroad, but these benefits are still far from disappearing. This is consistent with the finding of previous studies, which conclude that investors can still achieve significant international diversification gains and show the attractiveness of emerging and/or frontier markets for international portfolio diversification.

Keywords: international portfolio management; diversification benefits; emerging and

(3)

One of the major trends in the financial capital markets over the past 30 years has been the rise of international investment opportunities available to investors. This trend has occurred because of the increasing integration of markets and the consequent reduction or elimination of investment restrictions faced by investors. International investments are attractive to investors because of the greater portfolio risk reduction opportunities that could be obtained in comparison to what could be obtained through a domestic diversification.

The concepts of portfolio diversification are well-known to the investment community. The idea of “don’t put all of your eggs in one basket” came before any financial theory could formalize it, but still can summarize the idea behind diversification. The Modern Portfolio Theory – MPT – (Markowitz, 1952) and the Capital Asset Pricing Model – CAPM – (Sharpe-Lintner, 1964/1965), two of the most important theories in the finance literature, suggest that investors should hold a well-diversified portfolio. They show how investors should allocate their own overall wealth across a range of assets to minimize volatility for a given level of expected return or how to maximize the expected return for a given level of volatility. The foundation is that diversification reduces risk, and the diversification effect is greater the less correlated are the assets in the investor’s portfolio.

The CAPM theory is built over MPT, and suggests that the covariance of the asset with the (equity) market portfolio is the only risk that matters to capital market, and therefore is the only risk that is priced in equilibrium. The market portfolio is commonly considered to be a portfolio of all public equities available. Theoretically, diversification do not distinguish between equities from different countries. The proposition that the world market portfolio should be the investor’s best choice is derived by assuming that individuals differ very little from each other (they all have mean-variance objectives, differing only in risk aversion and wealth) and, therefore, all face a very similar problem. If the fact that investors live in different countries does not affect the homogeneity of their expectations and opportunities, everyone would still hold the same portfolio of risky assets, which would then have to be the world market portfolio (Cooper et. al (2013)).

Nevertheless, the reality is that investors still hold disproportionate amounts of domestic equities relative to the world market portfolio. This under-diversification phenomenon is

(5)

bias effect, which are also related to the concepts of market segmentation and market integration. In a broad perspective, barriers such as legal, informational, behavioural and economic impediments serve to segment domestic capital markets. Particularly, lack of liquidity, currency controls, special tax regulations, underdeveloped capital markets, exchange risk, expropriation risks, corporate governance, and information asymmetries preclude or reduce the attractiveness of investing abroad.

Whether international capital markets are segmented or integrated is closely related to the issue of international diversification. The understanding of market integration suggests that markets must be at least partially integrated for an investor to be able to benefit from international diversification. But despite the proven gains brought about by diversification, these gains can be very limited because of the mentioned barriers to foreign investing. Fortunately, segmentation have been reduced over the past 30 years and many of these restrictions are gradually being eroded.

In the late 1980s and early 1990s, developing countries gained access to foreign capital after the debt crisis of the mid-1980’s. Capital flows to emerging markets increased sharply, with commercial bank debt being replaced by portfolio flows (fixed income and equity) and foreign direct investment as the dominant sources of foreign capital. This could not have happened without these countries embarking on a financial liberalization process, relaxing restrictions on foreign ownership of assets, and taking other measures to develop their capital markets, together with macroeconomic and trade reforms (Bekaert and Harvey (2003)).

As a result, new capital markets started emerging, impacting the world financial environment. After the start of market integration processes, investors from developed countries could diversify their portfolio to newly available markets, and investors from these newly available markets could diversify over developed markets. Developing countries could also expect more economic growth as a result of inflows of capital.

The other side of market integration concerning diversification benefits is that, when markets become fully integrated, diversification benefits tend to be dramatically reduced, or even disappear. In this sense, there is an extensive literature that focus on analysing the levels of financial markets integration, presenting mixed results. General empirical findings are that the market integration leads to higher correlations with the world, and therefore correlation

(6)

market is still mildly segmented and that the degree of global market integration varies throughout time. Besides, emerging and frontier markets are considered to be relatively inefficient. Also, despite increasing financial integration, the home bias has not decreased sizably over the last years.

So, although over the last two decades investors have had the opportunity to increase the share of foreign assets in their equity portfolios (and they indeed partially took this opportunity), frictions in international markets still exist and the decline of the degree of home bias in international financial portfolios has been very slow, resulting in portfolios which still seem to be under-diversified.

Notwithstanding that the benefits of international equity diversification have been extensively researched, there is not a definitive conclusion about the relevance of well diversified equity portfolio in the current international context. A key issue in portfolio management is whether international diversification benefits is still substantial in the context of increasing equity market correlations and crises occurrence, due to economic and financial integration.

We are then interested in assessing if there are (still to be found) benefits to investing internationally for developed, emerging-market and frontier-market equity investors, regionally and/or globally. Moreover, if we find that there is statistical significance on these diversification benefits, we are also interested in evaluating what are their economic size, and how do they differ by country.

In summary, we aim is to analyze whether adding international investment opportunities provides diversification benefits to a domestic investor in three categories of markets, namely developed, emerging and frontier.

Most of the research on international portfolio diversification assumes a US investor viewpoint (e.g. Grubel (1968); Levy and Sarnat (1970); Huberman and Kandel (1987); De Santis and Gerard (1997); Errunza et al. (1999); De Roon et al. (2001); Li et. al (2003); Chiou (2009)). Differently, we contribute to the literature by taking the perspective of a local investor in different countries. It’s an important distinction, as the gains on foreign investing

(7)

American investors (since the US economy is, itself, the most diversified economy in the world) (Driessen and Laeven (2007).

Some authors such as Eun and Resnik (1994), Liljeblom et. al (1997), Fletcher and Marshall (2005), Driessen and Laeven (2007), and Chiou (2008) also take a local perspective, but most often they study a single country or a relatively small sample of countries. We take a sample of 69 countries, being 23 countries from developed markets, 23 countries from emerging markets and 23 countries from frontier markets. Frontier markets are referred to equity markets in smaller, less accessible, but still investable countries in the developing world. These countries are in their early stages of economic development, typically have long-run growth potential, and are often compared to emerging markets in the late 1990s. When their capital, liquidity and financial integration increase, frontier markets may be reclassified as emerging markets. Considering their geographic diversity and low return correlation with developed markets, these frontier markets are an interesting set of assets to be considered in portfolio diversification. To the best of our knowledge, this is the first study to consider such a broad sample from such different markets, and take a local perspective on the benefits of international diversification.

Another common feature of most researches is that they don’t cover the 2007 to 2009 global financial crisis period. Some studies don’t consider it for obvious reasons, as they were conducted before the crisis, and some others because there was not enough data available, as the crisis is a very recent event. As stated by Bekaert et. al (2014), this financial crisis is “arguably the first truly major global crisis since the Great Depression of 1929 to 1932. While the crisis initially had its origin in the United States in a relatively small segment of the lending market, namely the subprime mortgage market, it rapidly spread across virtually all economies, both advanced and emerging, as well as across economic sectors. It also affected equity markets worldwide, with many countries experiencing even sharper equity market crashes than the United States, making it an ideal laboratory to revisit the debate about the presence and sources of contagion in equity markets”. We therefore contribute to the literature by using more recent data covering this important event in our analysis.

We divide our research in two parts. In the first part we are interested in analysing if there are still benefits to investing abroad (regionally and/or globally) for equity investors from developed, emerging and frontier markets. Our main goal is to assess the utility gain that

(8)

would be obtained if all the restrictions of investing abroad were lifted. We analyze the statistical significance of the diversification possibilities using the “spanning” technique (originally proposed by Huberman and Kandel (1987), and expanded by other authors, such as De Roon et. al (2001) and Driessen and Laeven (2007), as well as the economic size of these possibilities, comparing increase in the Sharpe ratios and the decrease in volatilities (following Chiou (2008)).

Two reasons are presented to separate between a regional diversification and a global diversification: (i) there is evidence that people simply prefer to invest in the familiar investment opportunities (familiarity is associated with a sense of comfort with the known and discomfort with the distant), while often ignoring the principles of Modern Portfolio Theory (Huberman (2001)), and (ii) “measuring the potential benefits of regional diversification is also of interest to policy makers, because many countries are currently considering the setup of regional stock exchanges” Driessen and Laeven (2007). This gradual increase in international diversification is worth analysing. Very few other authors take into account this geographic differentiation, therefore we think our study also provides an important contribution in this sense.

In order to measure the economic size of the diversification benefits, we take an approach suggested by Chiou (2008). This approach allows us to assess the effects on frictionless markets as well as consider the effects of short-sales constraints and home bias investing behaviour. This analysis turns to be important because in some countries there are short-selling restrictions, especially for foreign investors. The already mentioned home bias effect is also an important limitation on international diversification. We measure the benefits for each country in two different ways, by maximizing the increase in Sharpe Ratio and by minimizing the global variance of the portfolio. In both, we also evaluate scenarios in which the investor diversifies his portfolio regionally and globally.

In the second part our goal is to investigate what causes the variation in benefits across the countries. We are interested in the effect of differences in country risk and the other country-level characteristics on the gains from global diversification. Because emerging and frontier countries are less integrated with the world capital markets, our hypothesis is that these countries can benefit more from diversifying internationally, and some country-level factors

(9)

Laeven (2007), we estimate for each country a regression of the differences in Sharpe ratio variable on a constant and a few country-specific factors: a proxy for the country risk, a proxy for the size of the country equity market, a proxy for trade openness of the country, and a proxy for the level of financial sector development of the country.

The data used in the study consists of ten years of historical data from April 2006 to March 2016. As we mentioned, it is an interesting time frame to be studied, because it encompasses a sub-period before the global financial crisis (bull market from April 2006 to July 2007), the period of the crisis (bear market from August 2007 to February 2009) and the recovery sub-period after the crisis (bull market from March 2009 to March 2016). We use U.S. dollar-denominated Morgan Stanley Capital International (MSCI) indexes as collected from MSCI website. The price data from MSCI are value-weighted indexes computed from end-of-month price observations on a very large sample of firms in each market. According to Kasa (1992), these indexes have the additional advantages of being built on a uniform basis across countries and of netting out cross-listed securities. This standardization is important because some factors such as index construction, diversification, industrial structure and exchange rate may cause these indexes to have a disparate behaviour (Roll (1992)). Thus, the MSCI indexes are particularly suitable for this study because they are constructed on a consistent common basis, which simplifies our cross-country comparisons.

The remaining part of this thesis is structured as follows. Chapter 2 provides a literature review on portfolio theory, market integration and the benefits of international diversification, Chapter 3 explains the methodology used in the thesis, Chapter 4 describes the data used, Chapter 5 presents the empirical results, and, finally, Chapter 6 gives a summary of the results and conclusion, limitations of the study and recommendations for future related studies.

(10)

2. Literature review

2.1. From the Modern Portfolio Theory (MPT) to the CAPM

Diversification of risk is a fundamental concept in finance theory. It is a common practice for investors to highly diversify their portfolios, unless there are particular reasons (such legal barriers, informational asymmetries, behavioural biases and economic impediments) forcing them to deviate from this norm. Many empirical studies take the world equity-market portfolio as the standard (Cooper et. al (2013)).

The proposition that the world equity-market portfolio should be the investor’s best choice is based on some basic (but strong) assumptions: markets are efficient, all investors have mean-variance utility functions and differ only in risk aversion and wealth. Given these assumptions, investors with access to the same information and opportunity sets but with different risk preferences should hold the same portfolio of risky assets (the portfolio of all public equities, or the “market portfolio”), and adjust for risk aversion by optimally mixing this “fund” with risk-free borrowing or lending (the two-fund theorem, which will be detailed in this section).Cooper et. al (2013) states that the “standard arguments about diversification do not distinguish between stocks of different countries. If the fact that investors live in different countries does not affect the homogeneity of their expectations and opportunities, everyone would still hold the same portfolio of risky assets, which would then have to be the world market portfolio”. As we will show later, for several reasons this is not what happens in practice.

A good explanation about the MPT and the CAPM is given by Geltner et. al (2006), and will be summarized in the following paragraphs.

The original Modern Portfolio Theory – MPT – (also called Markowitz portfolio theory or mean-variance portfolio theory) was proposed by Harry Markowitz (1952). It deals with the strategic decision of how best investors should allocate their own overall wealth across a range of assets (stocks, bonds, real estate, etc.) to minimize portfolio volatility for a given level of expected return or how to maximize the portfolio’s expected return for a given level of volatility. MPT quantifies the diversification benefits in terms of portfolio risk and return,

(11)

It results intuitive that some combinations of assets are more valuable than others as far as the diversification effect is concerned. Pairs of assets that are not perfectly positively correlated tend to provide some additional expected return for a given amount of risk (or, equivalently, less risk for a given expected return) when combined in a portfolio. Theoretically, the more asset classes we can choose to include in a portfolio, the more diversification potential there is.

According to MPT, all investors should hold portfolios on the efficient frontier, which consists of all asset combinations that maximize return and minimize risk. The risk/return preference of a specific investor determines which portfolio on the efficient frontier he should hold. In practice these preferences are often expressed in terms of a target long-run rate of return for the portfolio (or, equivalently, a target volatility). The (maximum risk-return) indifference curve reflects the highest level of satisfaction (or ‘‘utility’’) the investor can achieve. The indifference curve that is ‘‘tangent’’ to the efficient frontier is the best alternative for the investor. So, the investor’s selected target return (or target volatility) reflects the investor’s risk/return preferences. Figure 1 shows this explanation graphically.

Figure 1: Optimal portfolio without a risk-free asset.

Source: Ang (2014).

In addition to the risky assets mentioned so far, it is often useful to consider also a risk-free asset as a possible component of the portfolio. It can approximate borrowing or short-term

(12)

simplification is known as the two-fund theorem. The risk/return possibilities from a combination of a risk-free asset and any portfolio of risky assets will lie along a straight line connecting the risk-free asset’s risk/return with the risky portfolio’s risk/return, the Capital Market Line (CML), also known as Capital Allocation Line. All investors (no matter what their risk preferences are) will prefer combinations of the risk-free asset and a single particular risky asset portfolio (the “market portfolio”). In this case, the optimal allocation among risky assets is determined not by the investor’s risk preferences, but only by his expectations about asset future returns. The investor’s risk preferences can be met by adjusting the position in the riskless asset only, by either borrowing more (for more aggressive preferences) or lending more (for more conservative tastes), so as to meet the appropriate expected return target. Figure 2 shows the efficient frontier and the CML.

Figure 2: Efficient frontier and the Capital Market Line

Source: Benninga (2010)

The combination of risky assets equivalent to the unique optimal allocation in the two-fund theorem is the portfolio that maximizes the slope of a straight line connecting that portfolio’s risk and return with the risk and return of the risk-free asset. This slope is the risk premium of the risky asset portfolio divided by its volatility ((RPORTFOLIO - Rf)/SigmaPORTFOLIO), and is

known as the Sharpe ratio. When a risk-free asset is present, the optimal combination of risky assets is the one with the highest Sharpe ratio.

The Sharpe Ratio can be seen as a measure of risk-adjusted return. Its numerator measures the investor’s compensation for risk, the excess return over an investment in a “risk-free”

(13)

the investment (the standard deviation of its periodic returns). In the Sharpe ratio framework, volatility is the relevant measure of risk.

As we have seen, MPT is a fundamental tool of strategic investment decision making at the macro-level. We keep following Geltner et. al (2006) reasoning presenting another fundamental tool for macro-level investment analysis, which is the equilibrium asset pricing models. They are important because they can help investors understand what are reasonable expected returns, going forward into the future, on investments in different asset classes. As a consequence, investors are able to identify mispriced assets relative to long-run equilibrium. Besides, “by quantifying how the capital market prices risk, asset pricing models can be used to adjust portfolio returns to reflect the amount of risk in the portfolio, thereby helping to control for risk when evaluating portfolio returns or investment performance, which in turn can provide a more rigorous way to evaluate the performance of investment managers”.

The Sharpe-Lintner CAPM (proposed independently by William Sharpe (1964) and John Lintner (1965), who later shared a Nobel Prize) is one of the most famous theories in all of financial economics. It is built on the MPT, but go beyond it to provide a simplified representation of how the capital market perceives and prices risk in the assets that are traded in the market.

The basic idea underlying the CAPM is that the priced risk is not the isolated volatility of each individual asset, but rather it is related to the co-movements between asset returns. Recalling from MPT that all investors hold the same portfolio (the “market portfolio”), the only risk that matters to investors in any given asset is how that asset affects the risk in the market portfolio. Therefore, the only risk that matters to an investor in a stock can be quantified by that asset’s covariance with the market portfolio. As all investors hold their entire wealth in the market portfolio, the variance in this portfolio quantifies the risk that all investors are exposed to. In Geltner et. al (2006) words, “the main investment insight provided by the CAPM is the irrelevance of, and therefore lack of compensation for, diversifiable risk. The CAPM suggests that, as covariance with the market portfolio is the only risk that matters to the capital market, it is therefore the only risk that will be priced in equilibrium”.

(14)

Notwithstanding its fundamental importance, it is also well known that CAPM is not perfect, in the sense that it is not a complete description of reality of equilibrium asset pricing (as it is based on “false” assumptions, such as the existence of a truly risk-free asset, the perfect efficiency of the market, the common investors’ expectations etc.). Also, although beta (with respect to the stock market) is not the only factor affecting risk premiums within the stock market (other factors such as firm size, book/market value ratio and momentum also contributes to explain variation in average ex post returns to portfolios of stocks), it is probably the most important. “The CAPM is useful in practice presumably because the model’s assumptions, while simplifications of reality, are not terribly far from the truth” Geltner et. al (2006).

From finance theory and extensive empirical research, we know that the benefits of diversification depend largely on relatively low correlations between domestic and foreign assets returns. It is reasonable to assume that if the underlying economies and capital markets become more integrated, the equity markets will become more correlated, negatively impacting the benefits of international diversification (Magas (2007)).

An example of the benefits of international diversification is presented on Figure 3. If an investor starts with a diversified portfolio in efficient frontier A (composed by domestic US stocks only), he can improve his risk-return profiles by investing abroad in portfolios B, C, or D, all of which represent more more efficient portfolios. Portfolio B offers the same amount of return with a lower risk. Portfolio C offers a higher-return and lower-risk combination of assets. Portfolio D offers higher return with the same amount of risk. All three portfolios are on the shifted out efficient frontier, which comprises domestic US stocks and foreign stocks.

(15)

Next, we present the relation between MPT/CAPM and the concepts of market segmentation and integration, focusing in emerging and frontier markets, and also highlighting the important concept of home bias. After that, we review the empirical evidence on the benefits of international portfolio diversification.

2.2. Market integration, emerging and frontier equity markets, and home bias

One of the most debated issues in portfolio theory is the interrelationships and long-run co-movements between international equity markets (either developed or developing) and the real effects of these to international portfolio diversification. As markets become more integrated the co-movements between markets tend to rise, undermining the benefits of international portfolio diversification (Maria and Eva (2012)).

According to Maria and Eva (2012), there are three main approaches to describe the relationships between markets. First, in the “integration through prices” approach, stock market integration can be measured by using a single criterion, namely the law of one price. At the base of this approach lies the CAPM, which states that a common asset pricing model should be able to explain the different price levels of assets on different markets. Second, in the “capital-flow integration” approach, the impact of barriers to capital mobility is measured in order to determine stock market integration. It is therefore linked to the mobility of capital between stock markets. The third approach simply compares the evolution (co-movements) of equity markets in order to observe their degree of integration.

In the late 1980s and early 1990s, developing countries gained access to foreign capital after the debt crisis of the mid-1980’s. Capital flows to emerging markets increased sharply, with commercial bank debt being replaced by portfolio flows (fixed income and equity) and foreign direct investment as the dominant sources of foreign capital. This could not have happened without these countries embarking on a financial liberalization process, relaxing restrictions on foreign ownership of assets, and taking other measures to develop their capital markets, often in tandem with macroeconomic and trade reforms (Bekaert and Harvey (2003)).

(16)

As a result, new capital markets started emerging, impacting the world financial environment. After market integration, investors from developed countries could diversify their portfolio to newly available markets, and investors from these newly available markets could diversify over developed markets. Developing countries could also expect more economic growth as a result of inflows of capital.

Nevertheless, market integration is not an easy and rapid process. Markets are considered integrated, from a financial perspective, if assets of identical risk have the same expected return, irrespective of the country in which they are traded. The underlying risk factor is assumed to be the same for all assets. Theoretically, liberalization1_{should bring about}

emerging market integration with the global capital market. Foreign investors will bid up the prices of local stocks with diversification potential, (overall) the cost of equity capital will go down, investment will be increased, and the economy will grow. But it is unlikely that liberalization will lead to the full integration of any emerging market into the global capital market at once (Bekaert and Harvey (2003)). Also, the phenomenon of home asset preference (“home bias”, which we will discuss briefly later in this section) leads many economists to believe that even developed markets are not well integrated.

Considerable research has focused on the evolution of a country from segmented to integrated with world markets. Bekaert and Harvey (2002) identifies two levels of evolution: economic integration, referring to decreased barriers to trading in goods and services, and financial integration, referring to free access of foreigners to local capital markets (and local investors to foreign capital markets). In our study we are particularly interested in the latter.

Whether international capital markets are segmented or integrated is closely related to the issue of international diversification. The understanding of market integration suggests that markets must be at least partially integrated for an investor to be able to benefit from international diversification. In fully segmented markets, foreign investment is not possible by definition. In completely integrated markets, investors can take advantage from international diversification only if the assets in the domestic and foreign markets have different risk-return profiles. If they do not have, the potential gains from international

(17)

diversification are the same as those that can be achieved through domestic diversification (Shawky et. al (1997)).

Putting it differently: from a financial local perspective, in segmented markets investors can only invest in domestic assets, while in integrated markets investors can invest both in domestic and foreign assets. We can also get an intuition about the impact of market integration on asset prices by considering the Sharpe-Lintner CAPM, as stated by Bekaert and Harvey (2002): “in a completely segmented market, assets will be priced off the local market return. The local expected return is a product of the local beta times the local market risk premium. Given the high volatility of local returns, it is likely that the local expected return is high. In the integrated capital market, the expected return is determined by the beta with respect to the world market portfolio multiplied by the world risk premium. It is likely that this expected return is much lower. Hence, in the transition from a segmented to an integrated market, prices should rise and expected returns should decrease”. Although in practice the CAPM is often implemented by assuming markets fully segmented or fully integrated (Stulz, 1999), neither of these is the reality, being the partially segmented world of Errunza and Losq (1985) a more reasonable assumption (we refer to Bekaert and Harvey (1995) for the inconclusive evidence about whether equity markets are segmented or integrated). In a partially segmented market the risk premium is not exactly proportional to either the local or the global beta indexes, and from a practical valuation point of view the implication is that it is important to examine the information in both types of beta in order to make a judgment about the cost of equity (Bekaert and Harvey (2002)).

There is an extensive literature on market integration/segmentation. Authors assume that markets are completely segmented, entirely integrated, or partially integrated, and test whether the empirical data fits their models.

Authors who assume that perfectly integrated countries exist, and suggest that a single asset pricing model applies in that case, include: Solnik (1974), Sercu (1980), Stulz (1981), Adler and Dumas (1983). In the International CAPM’s (Solnik (1974) and Adler and Dumas (1983)) models, the expected return on every asset depends on its beta relative to the world market and on currency risk factors.

(18)

Errunza, Losq, and Padmanabhan (1992), Bekaert and Harvey (1995), Bekaert, Harvey, and Ng (2005), De Jong and De Roon (2005), and Pukthuanthong and Roll (2009) suggest that the world market is mildly segmented and that the degree of global market integration varies throughout time.

Shawky et. al (1997) find that the increased worldwide integration of financial markets might have caused stronger co-movements among international equity markets, thus reducing the potential gains from international diversification. But empirical studies do not provide unanimous evidence. It seems that the conclusions depend on the sample period, the intervals between measurements and the statistical methods applied. The authors also suggest that “there is compelling evidence in favour of international portfolio diversification as a reasonable method to reduce the risk of an investment portfolio without negatively affecting its expected return. The case for international diversification may be even stronger when emerging equity markets are included as part of the investor’s investment opportunity set. However, we remain convinced that due to the unstable structural relation between markets, investors are not likely to be able to take full advantage of the entire scope of possible international diversification gains”.

Bekaert and Harvey (2002) make a broad review of research on advances and integration process of emerging markets over the past years. Their main findings are that:

(i) “Dating market integration is complicated”: regulatory liberalizations are not necessarily defining events for market integration. For instance, the approval of an act that eliminates barriers on foreign capital flows may have little or no effect because of other risk factors. On the other hand, such a law may come after the market is already reasonably integrated (foreign investors might be already able to access the local market through other means such as ADRs or country funds when the barriers are lifted). (ii) “Market integration is often a gradual process”: there is a lot of sources of barriers that

limit and discourage foreign investments in a segmented country, and it is very unlikely that all of them will be lifted at a single point in time. The authors suggest three different types of barriers to emerging market investment: (a) direct (legal) barriers; (b) indirect barriers originated on information asymmetry, accounting standards and investor protection; and (c) general risks associated to developing countries, such as liquidity risk, political risk, economic policy risk and currency risk.

(19)

(iii) “Market integration impacts expected returns”: the authors find that with the process of integration, the expected returns of the assets should decrease.

(iv) “There is no obvious association between market integration and volatility”: they show that empirical evidence shows no significant changes in volatility when a capital market goes from segmented to integrated.

(v) “Capital flows increase after liberalization”: the authors find evidence that as after liberalization the barriers to entry decrease in emerging equity markets, the result is an expressive inflow of foreign capital. They also suggest that the initial foreign capital flows bid up prices and help create a ‘return to integration’.

(vi) “Emerging markets are relatively inefficient”: as expected, the authors find that many emerging equity markets do not behave like developed markets. Some informational efficiency exists in new and smaller equity markets, but it cannot be fully compared to efficiency of developed markets.

(vii) “Market integration leads to higher correlations with the world”: although theoretically the increase in market integration does not necessarily implies increase in the correlation with the world (they provide an example of a country with an industrial structure much different than the world’s average structure, which might keep little or no correlation with world equity returns even after liberalization), they find empirical evidence that, on average, correlations increase (17 of 20 markets experienced increased correlation with the world).

When correlations increase, the benefits of diversification tend to decrease. However, Rajan and Friedman (1997) conclude that international diversification opportunities exist. Market segmentation plays a role in creating those benefits, but as world markets become less segmented over time, international markets will become more highly correlated and the benefits of international diversification may be reduced. Bekaert and Harvey (2002) also find that the correlations of emerging market returns are still sufficiently low to provide important portfolio diversification. More recently, Chiou (2009) also stated that the degree of international market integration is critical to the effectiveness of international diversification, and that overall the integration of international financial markets has gradually increased. Notwithstanding, he argues that the emerging markets are still more segmented than developed countries.

(20)

Besides emerging markets, it is worth noting the importance of frontier markets. We briefly introduce the concept here, and in the following section we present researches on the subject and empirical evidence on diversification effects arising from investing in frontier markets.

Frontier markets are referred to equity markets in smaller, less accessible, but still investable countries in the developing world. About 60 frontier markets exist, located in Africa, Asia, Europe and South America (Speidell, 2011). The average market capitalization of frontier markets within our sample (23 countries) over the past 10 years (2005-2015) was about US$ 25.9 billion. These countries are in their early stages of economic development, typically have long-run growth potential, and are often compared to emerging markets in the late 1990s. When their capital, liquidity and financial integration increase, frontier markets may be reclassified as emerging markets. Considering their geographic diversity and low return correlation with developed markets, these frontier markets are an interesting set of assets to be considered in portfolio diversification. Indeed, the concept received significant emphasis with the launch of S&P Extended Frontier Index in 2007 and the Select Frontier Index in 2010. More recently, other frontier market indexes and ETFs have been launched. Despite this increasing attention to frontier markets among the investment community, very little research takes them into consideration, partially because of the lack of data available, and partially because only recently these markets became available for investment. Frontier markets may offer promising diversification benefits (Berger et al. (2011)). Therefore, the topic researching on frontier equity markets has become very important.

Berger et al. (2011) study the frontier markets and find little evidence of frontier market integration within the world market and a lack of consistent integration dynamics. Their results show that, in contrast with the developing and emerging counties, frontier markets are not integrated and offer no signal of increased integration over time. The authors suggest that frontier markets provide international diversification benefits when they appear in a diversified portfolio that includes developed and emerging equities.

Chen et. al (2014) study the co-movements between frontier markets and leading equity markets. They find that frontier markets in different regions have distinct relationships with leading markets. Population growth, industry value, interest rate, tax rate, and tariff of the frontier markets significantly influence the integration between both markets. Energy, gross

(21)

national income, stock traded value, and high-technology exports of leading markets saliently influence the integration.

To conclude our review on market integration, we make reference again to Bekaert and Harvey (2002), who present interesting links between the real economy and finance: as we showed earlier, market integration is associated with lower expected returns, and therefore, lower cost of capital. Lower cost of capital enables more investment projects to have positive net present value (NPV), and as a result, more investments opportunities can be undertaken. Besides the increase in investment, the authors provide evidence that the trade balance worsens after equity market liberalizations, suggesting that the additional investment is indeed financed by foreign capital, whereas personal consumption does not increase. Evidence suggests that real economic growth, measured by real GDP, increases, on average, by 1% per year over the five years following the liberalization of equity markets (Bekaert, Harvey and Lundblad, (2002)).

There is a significant literature on the linkage between finance and economic growth. Some researchers have focused on financial liberalization of the banking sector, stock market indicators and capital account reforms, such as Demirgüç-Kunt and Levine (1996), Levine et al. (2000) and Laeven (2001). Other researchers such as Rajan and Zingales (1998), Wurgler (2000), Lins et al. (2005) and Galindo et al. (2007) have focused on how more-developed financial markets help relax financing constraints and improve the allocation of capital by increasing investment in growing industries and decreasing investment in declining industries. The impact on the real economy is an average effect: some countries grow faster than others, and depend on different factors (Bekaert and Harvey (2002)). For instance, GDP growth gains are possibly negatively influenced by the degree of development of the country’s financial markets (Bekaert, Harvey and Lundblad (2002) suggest that if the bank loan market in a country is active and robust, this will mute the impact of opening an equity market on GDP prospects). “While it is difficult to attribute causality from the financial sector to the real economy, the evidence points to the important role of equity capital markets in the economic growth prospects of less-developed countries” (Bekaert and Harvey (2002)).

Part of our study will be dedicated to testing possible country-level economic factors on the size of more recent international diversification benefits.

(22)

The process of financial integration fostered by “policy events” such as capital account liberalizations, electronic trading, privatization programs, increasing exchanges of more precise information across borders, and reduced transaction costs led to a great increase in cross-border asset trade (Lane and Milesi-Ferretti 2003). Despite this “boom” in international investing possibilities, investors still seem to be reluctant to take full advantage of the benefits of international portfolio diversification, and end up holding portfolios concentrated in equities of their own country to a much greater extent than justified by portfolio theory, a phenomenon referred to as the “home bias in equities”. This behaviour remains even after official foreign investment restrictions are lifted by governments, with institutional investors typically keeping large amounts of their money invested in domestic securities (Driessen and Laeven (2007)). Thus, the home bias can be seen as another factor that restricts full market integration.

Evidence on home bias is well documented in the literature. Since the work of French and Poterba (1991), the home bias in stocks has puzzled academics. French and Poterba (1991) present surprising data on domestic ownership shares of the stock markets in the United States, Japan, United Kingdom, Germany and France. The results show very little cross-border diversification. The authors explore both institutional and behavioural explanations for this tendency investors have to overweight their domestic equity market. They argue that institutional constraints such as tax differences, transaction costs and explicit cross-border limits on investments can’t explain why investors specialize in their home markets, thus implying that incomplete diversification is the result of investor choices. Regarding behavioural explanations for the under-diversification phenomena, the authors argue that return expectations vary systematically across groups of investors: they are considered to be relatively more optimistic about the performance of their own markets. They also consider the fact investors tend to be better informed about investment opportunities in their own country than about foreign investments as possible explanations for home bias.

Cooper and Kaplanis (1994), Tesar and Werner (1995) and Lewis (1996) share the same view, arguing that historically investors prefer to invest in their home equity markets because they are more informed and optimistic about the future performance of the market. Cooper and Kaplanis (1994) report strong equity home bias in portfolio holdings in eight countries, including the US, the UK and Germany. Tesar and Werner (1995) report that home bias was

(23)

domestic equities, 90% in the US, 88.8% in Canada, 81.8% in Germany, and 77.5% in the UK. Lewis (1996) also reports that domestic stocks might provide better hedge for home risks than foreign stocks. Other possible explanations could be the presence of non-tradable consumption goods, diversification costs exceeding the gains, the effects of uncertainty about the economic environment, and the role of measurement errors in the data. He concludes that, overall, equity home bias remains a puzzle.

We make reference to Coeurdacier and Rey (2013) and Cooper et. al (2013) for more recent evidence on home bias.

Coeurdacier and Rey (2013) review various explanations. They find that despite better financial integration, the home bias has not decreased sizably: in 2007, U.S. investors still hold more than 80 percent of domestic equities, a much higher proportion than the share of U.S. equities in the world market portfolio. They observe home bias in equities in most countries, being higher in emerging markets, and report three broad classes of explanations: (i) hedging motives in frictionless financial markets (real exchange rate and non-tradable income risk), (ii) asset trade costs in international financial markets (such as transaction costs, differences in tax treatments between national and foreign assets or differences in legal frameworks), and (iii) informational frictions and behavioural biases.

Cooper et. al (2013) find that the equity home bias reflects a combination of factors, with information asymmetries and economic openness (including the absence of costs such as taxes on international holdings) being the most important. They also find that the bias is persistent and universal. They show that it has fallen a little over time for Organisation for Economic Co-operation and Development (OECD) countries but still remains high, and has not diminished in non-OECD countries. They divide explanations for the home bias puzzle into five broad categories: measurement issues, explicit costs and barriers, informational asymmetries, corporate governance/transparency/agency issues, and behavioural issues (investors overestimate their capabilities and are too optimistic about the familiar, and vice versa). Cooper et. al (2013) also conclude that the puzzle remains essentially unsolved.

It is also worth noting that possible explanations for the home bias effect should also be able to explain other closely related phenomena. Portes and Rey (2005) exemplifies by stating that investors not only overinvest in their home equity market but also invest most heavily in

(24)

markets that are closer to them, regardless of the potential diversification gains. By “closer”, they mean being geographically closer or related by strong information flows. This “foreign bias” is as marked and persistent as the home bias and equally presents a challenge to standard portfolio theory. Similar conclusions are presented by Coval and Moskowitz (1999), Chan et al. (2005), Leuz et al. (2010), Anderson et al. (2011), and Fedenia et al. (2013): local investors tend to overweight certain international markets in their portfolio holdings, such as those that are geographically close, or have common language or culture with the investor.

As we can see, several authors show that although over the last two or three decades investors have tended to increase the share of foreign assets in their equity portfolios, frictions in international markets still exist and the decline of the degree of home bias in international financial portfolios has been very slow, resulting in portfolios still under-diversified. In the context of our study, it is important to keep in mind the home bias effect, as it is a very important constraint to be taken into consideration when evaluating the economic benefits of international diversification. We will evaluate scenarios with different levels of home bias.

2.3. Empirical evidence on the benefits of international portfolio diversification

Modern portfolio theory has showed that diversification of portfolios reduces overall risk, especially by holding uncorrelated assets in a portfolio. The benefits of international diversification have started gaining interest and being documented in the literature in the late 1960’s, when globalization and international investing started to become important. Since then, they have been widely discussed and well documented in academic research. In this section we make a broad review on the existing literature about international diversification. This thesis relates to several different strands of the research literature on international financial markets. Initially we make an attempt to separate general branches of research related to this thesis, and then we present the main findings of some important studies in a chronological order.

To begin with, our research relates to the literature on financial market integration (e.g. Solnik (1974); Stulz (1981);Errunza, Losq, and Padmanabhan (1992); Bekaert and Harvey (1995); Shawky et. al (1997); Bekaert and Harvey (2002); Bekaert, Harvey, and Ng (2005), De Jong and De Roon (2005); Pukthuanthong and Roll (2009); Berger et. al (2011)). We review prior

(25)

for international investors and their ability to diversify their portfolios. These two broad topics are related as finance theory implies an inverse relationship between international diversification benefits and levels of world market integration.

In another strand of the literature, many time-series studies try to understand and quantify the co-movements between equity markets, as a way to determine how the diversification benefits of international investing have changed over time. Most of those studies have found that these co-movements vary over time. For instance, You and Daigler (2010) and Christoffersen et. al (2012) find that correlations have increased sharply in various markets, but this effect was much more pronounced in developed markets than in developing countries. These increasing correlations reduce potential diversification gains and thereby reduce the desirability of international equity diversification. However, there are mixed results concerning international diversification benefits, and several authors find that investors can still achieve significant diversification gains (e.g. Bekaert and Harvey (2002); Li et al. (2003); Driessen and Laeven (2007); Chiou et al. (2009)).

Besides general studies on market integration and correlation, there are studies that focus on financial crises and contagion effects. Most of these studies base their analysis on time-varying correlation as well (e.g. Johansson (2011); Nicklewsky and Rodgers (2013); Bekaert et al. (2014)). During volatile crisis periods, when demands to reduce portfolio volatility become stronger, overweighting relatively uncorrelated equity markets may become valuable.

Other researchers focus on the impact of investment constraints (such as short sales restrictions, home bias effects, transaction costs etc.), or risk factors (such as country risk, political risk, financial risk etc.), or general macroeconomic factors, on the benefits brought about by international investment diversification. Again, mixed results are found. Some authors suggest that investment constraints may not completely eliminate the benefits brought about by international diversification (Li et. al (2003); Driessen and Laeven (2007); Chiou (2008, 2009), while others states that diversification benefits to some markets have largely disappeared (e.g. De Roon et al. (2001); Niemczak (2010)).

Moreover, many studies focus on the attractiveness of emerging markets for international portfolio diversification (e.g. Errunza (1977); Bekaert and Harvey (2002, 2003); Gupta and Donleavy (2009); Christoffersen et al. (2012). There is a large discussion on this topic, as

(26)

with the increasing market integration their attractiveness is being questioned. Some authors argue that the benefits are decreasing fast, while authors state that gains from adding emerging markets to a portfolio is still large compared to benefits of adding additional developed markets to the same portfolio. More recently, in this same context, a few studies started to analyze also frontier markets (Speidell and Krohne (2007); Berger et al. (2011); Chen et. al (2014)).

As we will see, most of the research on international portfolio diversification takes a US perspective (e.g. Grubel (1968); Levy and Sarnat (1970); De Santis and Gerard (1997); Errunza et al. (1999); De Roon et al. (2001); Li et. al (2003); Chiou (2009).

Some early studies by Grubel (1968), Levy and Sarnat (1970), Agmon (1972), Lessard (1973), Solnik (1974), among others, report that there are substantial gains from investing in internationally diversified portfolios. These gains are explained by the low correlation among stock returns on international equities. These low correlations provide the basis for reducing the risk of the portfolio beyond that attainable by a domestic portfolio.

In his seminal work, Grubel (1968) shows that internationally diversified portfolios offer higher risk-return trade off when compared to undiversified portfolios, by examining common stock market averages of 11 major countries. He takes share prices indexes as proxies for each country market. He reveals the low cross-country correlations with US and shows that mean-variance efficient portfolios created with these countries help attaining portfolios which offer the same return with lower risk, or same risk with higher returns when compared with the US domestic portfolio.

Levy and Sarnat (1970) provide similar analysis by taking a sample of 28 countries for the period 1951 to 1967. They conclude that the inclusion in the efficient portfolio of countries with low correlations with US, such as developing countries, will provide higher risk-return trade-off for an American investor.

Agmon (1972) criticizes Grubel’s approach, suggesting that country indexes are incorrect measures of the potential benefits of international diversification. This is because each market has many assets and the composite market index does not capture all the possibilities

(27)

indexes are weakly correlated does not necessarily imply the superiority of international diversification over national diversification. He also arguments against the approach of segmented markets and states that share prices behave as if they were under one perfect multinational market, known as one market hypothesis. He analyses UK, German and Japanese stock prices during the period 1961-1966 and shows that they respond immediately to changes in US stock prices but the results are of poor quality (low significance). He concludes that is not possible to reject the validity of the one market hypothesis, anddespite the seeming barriers in the multinational equity market, there exists a substantial amount of relationship among the four equity markets studied.

Lessard (1973) investigates the benefits of diversification among a set of 4 Latin American countries and concludes that substantial gains can be obtained from a wide range of investment strategies regardless of the time period.

Solnik (1974) shows that significant advantages in risk reduction could be achieved by investors, both domestically and overseas. However, an internationally diversified portfolio is expected to bear a much lesser risk than a classic local portfolio. Results are provided not only for the American investor but also for seven European countries investing exclusively on their local stock market, and diversifying abroad, from 1966 to 1971. Solnik also ponders that the advantages of international investment are probably reduced by many institutional, political and psychological factors. Fluctuations in price thus represent only part of the total risk of a foreign investment, since the investor may also be concerned with the possible imposition of exchange controls and capital restrictions on foreign holdings.

Lessard (1976) builds a reasoning on the relationship between market integration and diversification benefits. He starts by explaining that a substantial part of the variance of returns on individual stocks is explained by national market indexes (systematic risk, which cannot be diversified away within the domestic market). However, correlations between the national indexes are relatively low, so only a fraction of national systematic risk elements is systematic in a world context. Therefore, international diversification should result in a substantial reduction in portfolio risk. The magnitude of these gains will depend on whether markets are segmented or integrated internationally. In segmented markets, assuming the validity of the CAPM, prices and expected returns are determined by the undiversifiable risk of each asset in the domestic portfolio. In integrated markets, prices and expected returns

(28)

are determined by the undiversifiable risk of each asset in the context of the world portfolio. Lessard notes that in fully integrated markets the advantage to diversify internationally is a pure diversification effect (reduction in the diversifiable risk of the portfolio). However, in fully segmented markets, the benefits from international diversification might be even greater, since some of the systematic (undiversifiable) risks, when measured against a national market portfolio will become unsystematic (diversifiable) due to a redefinition of the market portfolio to now include foreign assets.

Errunza (1977) uses the IMF, MSCI, and individual firm data for a group of developed and developing countries over the period 1957–1972 to show that emerging markets should be included in global portfolios and that the benefits persist over many time periods. He points out that returns available in developing countries securities compare favourably with those in sophisticated markets, even after adequate discounting for the problems and additional costs involved in such investments. The conclusion is that there is evidence that diversification appears to be desirable, and capital flows resulting from such diversification can tremendously improve the international liquidity position of the developing countries and provide a major development impact.

Jorion (1985) takes into consideration the uncertainty about the expected value of returns and covariance matrix of asset returns. He proposes a more conservative estimator for expected returns, which suggests that most of international diversification benefits are likely to come from a reduction in risk.

Eun and Resnick (1994) analyze the gains from international diversification from the Japanese and the US perspective. They use monthly return data for national bond and stock market indexes for the period between 1978 and 1989. They conclude that potential gains from international diversification are much greater for US investors than for Japanese investors. For the Japanese investors, these gains accrue in terms of lower risk, while for US investors, these gains accrue not so much in terms of lower risk but in terms of higher returns.

De Santis and Gerard (1997) use the international CAPM to estimate the expected gain from international diversification to a U.S. investor. They find that although integration may have

(29)

diversification for a U.S. investor is on average 2.11% annually, and have not declined significantly over the past two decades.

Rajan and Friedman (1997) take the viewpoint of a U.S. investor and investigate the impact of country risk on portfolio choice in the context of a Markowitz portfolio selection model. They argue that country risk assessment, in a portfolio investment decision context, should include factors that in some way limit access to a market or restrain the normal investment process. So, along with effects of political conditions and restrictions on foreign investments, factors such as discriminatory tax regulations, transaction costs, capital controls, and lack of information and liquidity differences between markets should be part of the country risk. The authors confirm that international diversification opportunities exist even after accounting for country risk (risk premiums ranging from 9.4% to 26.8% annually), and that market segmentation plays a role in generation of those benefits. But as world markets become less segmented over time, international markets will become more highly correlated and the benefits of international diversification may be reduced.

Errunza et al. (1999) also analyses the benefits of international diversification from a U.S. perspective. They examine whether portfolios of domestically traded securities can mimic foreign indexes so that investment in assets that trade only abroad is not necessary to exhaust the gains from international diversification. They study this issue within the mean-variance spanning framework of Huberman and Kandel (1987) and find that the benefits of international diversification can be achieved through investing in multinational corporates stocks, American depository receipts (ADRs) and closed-end country funds, which trade in the United States.

De Roon et al. (2001) re-examine if incorporating emerging markets into a US investor’s portfolio will improve the risk-return profile. But they consider the presence of transactions costs and short-sale constraints. They also use the mean-variance spanning tests developed by Huberman and Kandel (1987). They find that the diversification benefits to emerging markets disappear when short sales constraints or relatively higher transaction costs are imposed (they only hold with no constraints or with small transaction costs).

Bekaert and Harvey (2002) make a broad review of important research advances over the past 20 years in emerging markets finance. Previous research showed that when correlations

(30)

increase, the benefits of diversification tend to decrease. However, Bekaert and Harvey find that the correlations of emerging market returns are still sufficiently low to provide important portfolio diversification.

Li et al. (2003) use a different empirical framework than De Roon et. al (2001) to examine the international diversification benefits for U.S. equity investors when they are prohibited from short selling in emerging markets. They use Bayesian inference instead of mean-variance spanning tests. In opposition to what De Roon et. al (2001) find, Li et. al conclude there are substantial benefits of diversification in emerging markets even when investors face short sales constraints.

Speidell and Krohne (2007) examined the structural characteristics of frontier equity markets and analysed their returns and correlations with other markets. They show that correlation between frontier and developed markets is low. They see that challenges information sparsity, complex local regulations and limited research coverage by analysts as opportunities for investors, concluding that investment in these markets may be highly rewarding.

Driessen and Leaven (2007) take the viewpoint of domestic investors in different countries and investigate how the benefits of international portfolio diversification differ across countries, assuming the existence of short-sales constraints. The authors work with the same empirical framework of De Roon et al. (2001). They find that the benefits of investing abroad are largest for the investors in developing countries. They also found that a major part of the diversification benefits disappears when short-sales constraints are present, even for developing countries.

Chiou (2008) examines the relative magnitude of the international diversification benefits for the domestic investors in various countries. Taking into account the constraints of short-sales, over-weighting investments, and investing region, the empirical results suggest that local investors in less developed countries, particularly in East Asia and Latin America, comparatively benefit more from both regional and global diversifications. He collects data from 1988 to 2004, so the effects of the global financial crisis starting in 2007 are not taken into consideration.

(31)

Chiou et al. (2009) investigate the potential benefits of international diversification for the US investor with various investment constraints from both long-term and time-rolling perspectives. Their findings suggest that adding short-selling and over-weighting constraints reduce, but do not completely eliminate, the diversification benefits of international investment. The over-time analyses show that diversifying portfolios internationally is still beneficial even though financial markets are becoming more integrated. Their sample period is from 1993 to 2005.

Gupta and Donleavy (2009) analyze the benefits of international diversification towards the emerging markets for an Australian investor by calculating time-varying correlations to be used in a portfolio optimization model. They find that, despite increasing correlations, the risk-return trade off increases significantly after international diversification when compared to Australian only portfolio, concluding that there are still gains to be obtained by international diversification.

Taking recent financial crisis (2008/2009) into account, Johansson (2010) observes higher correlations and higher levels of integration during the recent financial turmoil period both in Europe and East Asia. His findings indicate that the benefits of such diversification have been of limited value during the global financial crisis.

Niemczak (2010) focus his research on the analysis of the developments in the stock markets of Eastern European countries before and during the subprime crisis, from the point of view of a US investor. He aims to evaluate the hypothesis of disappearing portfolio diversification opportunities in Eastern Europe, and concludes that they have largely disappeared.

You and Daigler (2010) find the diversification benefits are changing over time, and the issue of time variation and accurate assessment of correlations becomes more important for investment in smaller emerging markets with potentially higher volatility. They analyze data over the period 1997 to 2002 from a US viewpoint, and despite some evidence against diversification, they state that there are potential benefits depending on the benchmark (country) employed.

Berger et al. (2011) analyses 25 frontier market equities with respect to world market integration and diversification, from 1989 to 2009. They show that frontier markets have

Benefits of international equity portfolio diversification in a post-crisis scenario

University of Amsterdam - Amsterdam Business School

Master in International Finance

Master Thesis

Benefits of international equity portfolio diversification in a post-crisis scenario

Rafael Lopes de Albuquerque

Student Number: 11150874

Thesis Supervisor: Dr. Rafael Matta

Table of Contents