VoxEU Column Global crisis

International capital flows during crises: Gross matters

How much do we really know about net capital flows? Presenting new research, this column lays out a number of new stylised facts on the dynamics of gross capital flows and their implications for policymaking. Interestingly, if we’re to learn from relatively crisis-resilient middle-income countries, policymakers may well need to monitor and perhaps regulate the separate behaviour of domestic and foreign investors to weather future crises.

The financial crises of the last three decades have spurred interest in the dynamics of international capital flows. Most of the work on the topic has focused on the behaviour of net capital flows, namely the difference between the foreign purchases of domestic assets (or capital inflows by foreigners) and the domestic purchases of foreign assets (or capital outflows by domestic agents). However, much less is known about the individual behaviour of these two components on net capital flows, which we denote as gross capital flows.

Gross capital flows

Domestic and foreign investors are likely to behave differently. And understanding their different behaviour is crucial from both a policy and a theoretical perspective. For example, it is important to understand if the reduction in net capital inflows during crises is driven mostly by a reduction in the purchases of domestic assets by foreigners – i.e. a sudden stop – or by an increase in the purchases of foreign assets by domestic agents – i.e. capital flight. This distinction can help determine the nature of crises and, thus, the appropriate policy responses both ex ante and ex post.

In this column we describe a number of new stylised facts on the dynamics of gross capital flows. This shift in attention from net to gross capital flows is part of a growing literature (see, e.g., Kraay et al. 2005, Lane and Milesi-Ferretti 2007, Cowan et al. 2008, Rothenberg and Warnock 2011, Forbes and Warnock 2012, Broner et al. 2013).

The dynamics of gross capital flows: Stylised facts

We analyse the behaviour of capital inflows by foreigners and capital outflows by domestic agents since the 1970s for a large number of high-income and middle-income countries. Figure 1 shows the raw data of capital inflows by foreigners and capital outflows by domestic agents normalised by trend GDP for a selected group of countries. A brief summary of the main results presented in Broner et al. (2013) follows:

  • Gross capital flows, capital inflows by foreigners and capital outflows by domestic agents, are large and volatile both in absolute terms and relative to the size and volatility of net capital flows.

While the size and volatility of net capital flows, capital inflows by foreigners - capital outflows by domestic agents, have remained roughly constant over the last decades, both have increased substantially over time for gross capital flows. For example, while in the 1970s and 1980s the standard deviation of capital inflows by foreigners plus capital outflows by domestic agents was similar to the standard deviation of capital inflows by foreigners minus capital outflows by domestic agents, the ratio of the two standard deviations was 2.0 in the 1990s and 2.4 in the 2000s. This reflects an increasingly positive correlation between capital inflows by foreigners and capital outflows by domestic agents, also apparent in Figure 1.

Figure 1. Capital flows in selected countries

  • Gross capital flows, capital inflows by foreigners and capital outflows by domestic agents, are procyclical;

In other words, during expansions foreigners increase their purchases of domestic assets and domestic agents increase their purchases of foreign assets. During contractions foreigners reduce their investments in domestic assets and domestic agents reduce their investments abroad. This retrenchment towards home financial markets is particularly acute during crises.

  • Crises that occur during periods of global financial turbulence are associated with particularly large retrenchments.

Moreover, retrenchments take place during banking, currency, and debt crises.

  • The retrenchments during crises take place for every type of gross capital flows, including direct investments, other investments, portfolio debt, and portfolio equity.

The behaviour of reserves differs across income groups. Reserves play an important role in the contraction of capital flows in middle-income countries and none in high-income ones.

Interpretation of the results

These new stylised facts on gross capital flows help discriminate among several existing theories of capital flows and crises. A number of papers emphasise the role of productivity shocks during economic crises and analyse the behaviour of gross capital flows during such episodes (Hnatkovska 2010; Tille and van Wincoop 2010). In these papers, the returns on domestic assets depend on productivity, but not on whether investors are foreigners or domestic agents. Other papers argue that crises affect these two sets of investors asymmetrically. One possible source of asymmetry is asymmetric information about the returns on domestic assets (Dvorak 2003; Tille and Van Wincoop 2008). In these papers, a retrenchment during crises can take place if crises are associated with an increase in information asymmetry. Another source of asymmetry comes from differential shocks to risk aversion or the relative riskiness of home and foreign assets (Milesi-Ferreti and Tille 2010). Finally, sovereign risk and, more generally, a relative deterioration of foreigners’ property rights provide another source of asymmetry between foreigners and domestic agents (Broner, Martin, and Ventura 2010). For example, if domestic agents are less likely to be defaulted on than foreigners during crises, foreigners have an incentive to sell domestic assets to domestic agents when the risk of default increases.

The evidence seems more consistent with theories in which crises affect foreigners and domestic agents asymmetrically than with theories that stress productivity shocks. Intuitively, because productivity shocks affect to the same extent foreign and domestic investors, their natural effect is to prompt investors to make similar investment decisions across countries regardless of their nationality. In contrast, shocks that affect domestic and foreign investors asymmetrically can more easily account for their different investment behaviour over the business cycle and during crises. We interpret our evidence as supportive of shocks that increase informational asymmetry and/or cause a relative deterioration of foreigners’ property and repayment rights during crises. But regardless of our own interpretation, the evidence provides a new set of empirical moments that will help judge the relevance of existing and future theories.

Policy discussion

These new stylised facts have important policy implications. In particular, policymakers have traditionally focused on net capital flows or, their counterpart, the current account to understand why crises occur and, more generally, how countries are integrating in international capital markets. Our results show that the reduction in net capital inflows during crises is in fact substantially smaller than the reduction in gross capital flows. Thus, during crises countries face a reduction in their ability not only to finance domestic investments with foreign savings, but also to share domestic idiosyncratic risks with foreigners. This became clearly apparent during the 2008-09 global financial crisis, when there was a significant reduction in financial globalisation. Therefore, the policy discussion would benefit from a shift in focus from net capital flows towards gross capital flows, as already suggested in recent discussions (Brunnermeier et al. 2012; Obstfeld 2012)

Our evidence speaks to the relative importance of foreigners and domestic agents in accounting for the reduction of net capital inflows during crises. In particular, on average foreigners are responsible for more than 100% of the reduction in net inflows, with domestic agents reducing their outflows and thereby offsetting, in part, the behaviour of foreigners. Thus, neither fire sales of domestic assets to foreigners nor domestic capital flight seem to be as important as previously thought. Instead, crises are characterised by a sudden stop of foreign-capital inflows. To reduce foreigners’ incentives to leave, policymakers might want to consider better protecting foreigners’ property rights, for example, by making it more difficult for countries to discriminate against foreigners during debt restructurings and nationalisations. Moreover, the large changes in gross capital flows during crises can lead to reallocations between foreign and domestic investors that have large effects on financial systems, as investors unwind their positions and retrench. This might be reflected in large fluctuations in asset prices and withdrawals from deposit taking institutions, generating a need for liquidity provision from domestic or international institutions.

At the prudential level, policymakers might need to monitor and perhaps regulate ex ante the separate behaviour of domestic and foreign investors. Fashionable tools include macroprudential rules and capital controls. During good times, as financial globalisation deepens, authorities might want to encourage that foreign capital inflows are channelled to assets that can cushion eventual negative shocks. In particular, capital inflows intermediated by the banking system might prove particularly dangerous given the susceptibility of banks to runs. Flows into short-term debt might also generate fragility due to the rollover risk. On the other hand, equity and direct investment seem safer ways to channel foreign-capital inflows. On the asset side, countries might be encouraged to keep accumulating large amounts of reserves to be able to withstand shocks. In fact, this type of international financial integration seems to be the strategy that middle-income countries have followed during the 2000s, which might explain their resilience during the global financial crisis.


Broner, F, T Didier, A Erce, and S Schmukler (2013), “Gross capital flows: Dynamics and crises”, Journal of Monetary Economics 60, 113-33.

Broner, F, A Martin, and J Ventura (2010), “Sovereign risk and secondary markets”, The American Economic Review 100, 1523-55.

Brunnermeier, M, J De Gregorio, P Lane, H Rey, and H Shin (2012), “Banks and cross-border capital flows: Policy challenges and regulatory responses”, VoxEU.org, 7 October.

Cowan, K, J De Gregorio, A Micco, and C Neilson (2008), “Financial diversification, sudden stops, and sudden starts”, in K Cowan, S Edwards, R Valdés, and N Loayza (eds.) Current Account and External Finance, Central Bank of Chile, 159-94.

Dvorak, T (2003), “Gross capital flows and asymmetric information”, Journal of International Money and Finance 22, 835-64.

Forbes, K, and F Warnock (2012), “Capital flow waves: Surges, stops, flight and retrenchment”, Journal of International Economics 88, 235-51.

Hnatkovska, V (2010), “Home bias and high turnover: Dynamic portfolio choice with incomplete markets”, Journal of International Economics 80, 113-28.

Kraay, A, N Loayza, L Servén, and J Ventura (2005), “Country portfolios”, Journal of the European Economic Association 3, 914-45.

Lane, P, and G Milesi-Ferretti (2007), “The external wealth of nations mark II: Revised and extended estimates of foreign assets and liabilities, 1970-2004”, Journal of International Economics 73, 223-50.

Milesi-Ferretti, G, and C Tille (2010), “The great retrenchment: International capital flows during the global financial crisis”, Economic Policy 66, 285-330.

Obstfeld, M (2012), “Financial flows, financial crises, and global imbalances”, Journal of International Money and Finance 31, 469-80.

Rothenberg, A, and F Warnock (2011), “Sudden flight and true sudden stops”, Review of International Economics 19, 509-24.

Tille, C, and E van Wincoop (2008), “International capital flows”, Journal of International Economics 80, 157-75.

Tille, C, and E van Wincoop (2010), “International capital flows under dispersed information: Theory and evidence”, NBER Working Paper 14390.

8,924 Reads