3.1 Financial Globalisation and Sources of the Crisis
Did the rise in cross-border .nancial trade contribute to the origination of the global
.nancial crisis? It is possible to identify two main channels by which .nancial globalisation contributed to the .nancial conditions that ultimately gave rise to the crisis.
First, the participation of foreign investors (especially foreign banks) fuelled the accel- erated growth of the asset-backed securities markets in the United States that were central in the original market panic in 2007-2008. As documented by Acharya and Schnabl (2009) and Bernanke et al (2011), European banks were major purchasers of asset-backed secu- rities. In large part, these banks also obtained dollar funding in the US money markets, as detailed by Shin (2011). For this reason, the role of European banks in enabling the expansion of the US ABS markets did not jump out of the balance of payments data, although the implicit risk exposure of European parent banks grew in line with these US activities. Second, .nancial globalisation permitted rapid growth in the balance sheets of many banks. This took place at two levels. In relation to globally-active banks, the size and complexity of these banks grew rapidly, making it di¢ cult for national regulators to adequately police risk pro.les. In addition, the capacity of local banks to expand lending by tapping international wholesale markets fuelled credit growth in a number of countries.
Third, the growing role of emerging markets in the world .nancial system may also have contributed to the build up of weaknesses in credit markets. In particular, the general- equilibrium impact of the demand for low-risk debt assets from emerging-market o¢ cial

8 sources and the increased supply of equity opportunities in these countries may have added fuel to the securitisation boom (see, amongst many others, Bernanke et al 2011).
In these ways, although there is no easy way to quantify its importance relative to other factors, .nancial globalisation plausibly contributed to the credit market vulnerabilities that were at the origin of the global .nancial crisis. In essence, .nancial globalisation magni.ed the impact of underlying distortions, such as inadequate regulation of credit markets and banks, by allowing the scaling-up of .nancial activities that might have faced capacity limits in autarkic .nancial systems (see also Borio and Disyatat 2011).
Going further, the variation in the cross-country incidence and propagation of the global crisis may be linked to di¤erential levels and types of participation in cross-border .nancial trade. For instance, Lane and Milesi-Ferretti (2011b) identify the rate of pre-crisis credit growth and the scale of pre-crisis current account de.cits as important correlates of the scale of the decline in output and domestic expenditure during 2009. More generally, credit growth is an important leading indicator of subsequent .nancial crises, with recent estimates provided by Jorda et al (2011) and Gourinchas and Obstfeld (2012). In related fashion, Lane and Milesi-Ferretti (2012) show that the scale of pre-crisis .excess.current account de.cits has been associated with larger current account turnarounds over 2008-
2010, which were mostly accomplished through reductions in domestic absorption.
Financial globalisation plausibly contributed to rapid domestic credit growth by oper- ating on both supply factors and demand factors. In relation to credit supply, domestic banks and the a¢ liates of foreign banks can raise various types of wholesale funding on international markets, with foreign investors also an important source of bank equity (both through portfolio equity and FDI). In relation to credit demand, capital in.ows can con- tribute to a low interest rate environment and improve the net worth of domestic borrowers by pushing up domestic asset prices.
For a sample of European countries, Lane and McQuade (2012) .nd that international capital in.ows were indeed correlated with domestic credit growth during the 2003-2008 pre-crisis period. In particular, a high level of net debt in.ows was associated with more

9 rapid domestic credit growth, as is shown in Figure 8. This is consistent with the funding channel by which domestic banks were able to expand lending capacity by raising wholesale funding on international markets. The .ip side is that high rates of net debt out.ows were associated with lower domestic credit growth in key source countries such as Germany.
As noted, it is also plausible that .nancial globalisation facilitated the emergence of large and persistent current account imbalances. Figure 9 shows the cross-country standard deviation in current account balances over 1995-2010, which captures a marked increase in dispersion during 2002-2007, followed by a compression phase during 2008-2010. On the de.cit side, a larger pool of international investors made it easier to obtain external funding; on the surplus side, the availability of foreign assets provided an alternative outlet for domestic savings.
At a global level, Blanchard and Milesi-Ferretti (2010) highlight that the factors driving current account balances varied across di¤erent phases. Still, international .nancial inte- gration surely increased the elasticity of capital .ows to these underlying forces, both in relation to welfare-increasing factors (such as di¤erences in demographics or productivity) and welfare-reducing factors (asset bubbles, overborrowing distortions).
For a European sample, Lane and Pels (2012) .nd that the covariates of current account imbalances shifted around 2003. Before 2003, external imbalances were highly correlated with di¤erences in output per capita, with richer European countries running surpluses and poorer European countries running de.cits. This pattern was interpreted benignly by Blanchard and Giavazzi (2002) as conforming to neoclassical patterns by which capital running downhill should accelerate output convergence and improve welfare.
However, Lane and Pels (2012) show that di¤erences in growth projections across coun- tries emerged as a strong correlate of current account imbalances during the 2003-2007 period in which the level of dispersion in external positions sharply increased. To the ex- tent that these growth projections were extrapolations from the unsustainable expansion of the nontraded sectors in some countries (Ireland and Spain and some Eastern European countries come to mind), these types of current account de.cits were not necessarily welfare
10
enhancing. Indeed, these authors show that a primary source of the covariation between growth projections and the current account during 2003-2007 was an increase in construc- tion investment in the de.cit countries, which was a danger signal in terms of repayment capacity (see also Blanchard 2007, Giavazzi and Spaventa 2011, Chen et al 2012).
In summary, we can consider .nancial globalisation to have contributed to the orig- ination of the crisis by enabling the scaling-up of the US securitisation boom that was the proximate trigger for the crisis .it is di¢ cult to imagine that the growth in these credit markets would have been of similar magnitude without the participation of foreign investors. In addition, .nancial globalisation also had a central role in the emergence of large and persistent di¤erences in credit growth and current account imbalances across countries .these imbalances would play a central role in determining the cross-country incidence and propagation of the original crisis.