Information about Europe Europa
Chapter

V Risk-Sharing Benefits of Financial Globalization: Theory and Practice

Author(s):
Giovanni Dell'Ariccia, Paolo Mauro, Andre Faria, Jonathan Ostry, Julian Di Giovanni, Martin Schindler, Ayhan Kose, and Marco Terrones
Published Date:
December 2008
Share
  • ShareShare
Information about Europe Europa
Show Summary Details

One of the key purported benefits of international financial integration relates to greater risk sharing: by making it possible for a country's residents to hold financial assets whose returns are linked to output performance abroad, financial openness provides opportunities to enjoy relatively stable consumption streams despite fluctuations in domestic output. This section considers both potential gains from risk sharing—by comparing the extent to which the volatility of domestic consumption exceeds that of foreign output—as well as actual gains measured on the basis of observed declines in the correlation between domestic consumption and domestic output. Section VI will then shift the focus to the observed effects of financial globalization on domestic consumption volatility, as well as crisis propensity and long-run economic growth.

International Risk Sharing in Theory

Economic theory suggests that under full financial integration, each country will consume a fixed share of the output produced by the group of countries with which it is integrated.1 In other words, the growth rate of consumption will be the same for all countries “in the group” and will equal the growth rate of groupwide output. Although this is unlikely to happen in practice, it is a useful benchmark for assessing the potential risk-sharing gains from financial integration. In simple terms, a practical way of measuring potential risk-sharing gains is to compare an individual country's consumption volatility with the volatility of groupwide output: if a country's individual consumption volatility is much higher than it would be under full financial integration within the group, then potential risk-sharing gains are relatively large.2 The main findings, reported in Table 5.1, are as follows:

Table 5.1.Potential Gains from Risk Pooling Among Countries
Median σ Individual Country (Consumption)σ Whole Group (Income)
(1)(2)
All countries4.450.81
Interest in risk sharing, by level of development and size of country
Advanced countries2.191.18
Emerging market countries4.011.29
Developing countries8.241.30
Interest in risk sharing, by size of country
Small countries6.721.23
Large coutries3.480.95
Current degree of international financial integration
High integration countries2.450.85
Low integration countries6.101.31
Adherence to international arrangements
Relatively strong2.311.07
Relatively weak6.111.26
Above-average institutional quality2.451.10
Below-average institutional quality6.171.32
Sources: Foreign assets are from Lane and Milesi-Ferretti (2006); GDP data in current U.S. dollars are from the IMF's World Economic Outlook. GDP and consumption data at purchasing power parity are from the World Bank's World Development Indicators.Notes: Column (1) reports the median (across countries in the indicated subsample) standard deviation (a) of individual country growth in 1975-2004. Column (2) reports the standard deviation of the growth rate of total output for the group of countries as indicated. Small (large) countries are those with a total population of less than (more than) 5.2 million in 1970. High (low) capital integration countries are those in the top (bottom) half of the sample when ranked by total foreign assets to GDP. Above- and below-average institutional quality is measured according to the index by Kaufmann, Kraay, and Mastruzzi (2005). Relatively strong adherence to international arrangements is defined as above-average institutional quality and no defaults on international debt in 1970-2004 according to Detragiache and Spilimbergo (2001) and Reinhart, Rogoff, and Savastano (2003).
Sources: Foreign assets are from Lane and Milesi-Ferretti (2006); GDP data in current U.S. dollars are from the IMF's World Economic Outlook. GDP and consumption data at purchasing power parity are from the World Bank's World Development Indicators.Notes: Column (1) reports the median (across countries in the indicated subsample) standard deviation (a) of individual country growth in 1975-2004. Column (2) reports the standard deviation of the growth rate of total output for the group of countries as indicated. Small (large) countries are those with a total population of less than (more than) 5.2 million in 1970. High (low) capital integration countries are those in the top (bottom) half of the sample when ranked by total foreign assets to GDP. Above- and below-average institutional quality is measured according to the index by Kaufmann, Kraay, and Mastruzzi (2005). Relatively strong adherence to international arrangements is defined as above-average institutional quality and no defaults on international debt in 1970-2004 according to Detragiache and Spilimbergo (2001) and Reinhart, Rogoff, and Savastano (2003).
  • The potential risk-sharing benefits (reduction in consumption volatility) from full financial integration with the rest of the world are substantial for every country: the standard deviation of worldwide output growth is 0.8 percentage point, far lower than the median standard deviation of consumption growth for individual countries (4.4 percentage points), and lower even than the standard deviation for the country with the lowest consumption volatility (1.4 percentage points).
  • The potential gains from financial globalization are larger for countries whose economies are more volatile because they are subject to more frequent and relatively damaging idiosyncratic shocks (for example, smaller, and therefore less diversified, countries) or because their ability to smooth such shocks through countercyclical policies or domestic financial markets is lower (for example, countries at a lower stage of economic and domestic financial development)
  • The potential gains are greater for countries whose international financial integration is relatively low, which to a large degree are countries whose economic cycles are less correlated with worldwide economic developments.

Against the benchmark of the potential gains from global financial integration, what can be said about “optimal” groupings of countries from a risk-sharing perspective? Empirical analysis (based on Imbs and Mauro, 2007) suggests the following:

  • The bulk of the potential risk-sharing benefits available to a country within a given sample of countries (for example, the world, a region, or countries within a given range of per capita income) can be attained in a small pool consisting of a handful of well-chosen partners. For example, consumption volatility can be reduced by more than half for a typical advanced country through full financial integration with an “optimally chosen” pool of five countries. The potential gains are even higher for optimal pools of emerging market and developing countries.
  • While regional pools can provide major benefits, risk-sharing benefits tend to be greater when countries choose partners from the worldwide sample rather than within a region. For example, median volatility of consumption growth for Latin American emerging market countries equals 6.2 percentage points and can be lowered to 1.9 percentage points by pooling with five optimally chosen Latin American emerging market countries, but to 1.3 percentage points by pooling with five optimally chosen emerging market countries in the absence of geographical constraints. Similarly, the median Asian emerging market country can reduce its volatility from 4.1 percentage points to 1.9 percentage points in a pool of five Asian emerging market countries, and to 1.4 percentage points in a pool of five emerging market countries chosen also from outside the region.3

International Risk Sharing in Practice

Even if the potential benefits of international risk sharing are large, to what extent has international risk sharing actually taken place in practice? In particular, has the increase in international financial integration over the past three decades resulted in improved risk sharing? A possible empirical proxy for high international risk sharing is a low correlation between domestic consumption and domestic output. Based on nine-year rolling window correlations between domestic consumption and domestic output, the empirical evidence suggests that international risk sharing has indeed increased somewhat for advanced countries, especially over the past two decades, but that for emerging market and developing countries there has been relatively little change (Kose, Prasad, and Terrones, 2007). Other studies confirm a favorable effect of financial integration on actual international risk sharing in advanced countries, especially among OECD, EU, and European Monetary Union countries—that is, groups where integration has increased relatively rapidly (Artis and Hoffmann, 2006a and 2006b). Relatedly, analyses of recent changes in the pattern of countries' holdings of international financial assets have found that home bias has declined in the advanced countries; and that such decline has indeed been associated with somewhat increased international risk sharing (Sørensen and others, 2007).4

The finding that actual risk-sharing benefits have been larger for advanced countries than for emerging market and developing countries, in contrast to the larger potential gains for these latter groups, may reflect faster and more substantial increases in de facto integration in the first group, as discussed in Section III. It may also result from a dependence of risk-sharing benefits on whether countries have in place certain preconditions—related for example to trade openness or domestic financial sector development (Kose, Prasad, and Terrones, 2007; and Levchenko, 2005). Looking ahead, an implication may be that a large increase in de facto financial integration and/or accompanying progress with regard to domestic fundamentals are required for emerging market and developing countries to reap significant risk-sharing benefits, and it may thus take several years for this segment of the membership to attain such benefits, unless present financial integration and reform trends accelerate significantly. This being said, recent increases in the share of equity and FDI—that is, forms of financing that facilitate international risk sharing owing to the procyclical nature of the associated payments—may suggest that actual risk sharing will be higher in the next decades than it has been in the past.

This section was prepared by André Faria and Paolo Mauro.
1The precise definition of full international financial integration in this theory involves full sharing of GDP risk, which could in principle be attained via a network of bilateral GDP swaps or the trading of claims on GDP. The analysis assumes that international financial integration does not affect GDP growth correlations across countries: see Obstfeld and Rogoff (1996, Chapter 5).
2The results are similar using an individual country's output volatility rather than consumption volatility. Some authors (for example, Lucas, 1987) have argued—based on evidence for advanced countries—that the welfare gains from reducing consumption volatility are small. However, others have shown that the welfare gains are much larger for emerging market and developing countries than for advanced countries (for example, Pallage and Robe, 2003). Potential risk-sharing benefits presented in the paper suggest sizable welfare gains.
3An approach based on the number of crises common to more than one member of a given pool yields higher costs of geographical constraints, reflecting a regional element in past emerging market country crises.
4“Home bias” refers to the observation that investors diversify across countries substantially less than would appear to be warranted based on standard portfolio theories: in other words, by increasing their holding of foreign assets, investors in most countries would be able to reduce the riskiness of their portfolios while maintaining a constant expected rate of return.

    Other Resources Citing This Publication