The United States not only continues to dominate global finance but has become even more central since the 2008 crisis. How did this happen?
Note: This article was originally published on October 7, 2013.
The great financial crisis of 2008 convinced many in the markets and policy arena that the U.S. had reached its high-water mark of dominance and that its decline was sealed. As they saw it, American financial prominence had proven so destabilizing that other countries had to insulate themselves against “profligate” U.S. behavior. Furthermore, the crisis dramatically reduced U.S. attractiveness to global capital, weakening its financial power to such an extent that the U.S. would be severely constrained in its ability to finance government debt at home and pursue geopolitical projects abroad.
As a result, many in this camp have anticipated a restructuring of the international financial system away from New York and toward China and other emerging markets. According to the World Economic Forum, Hong Kong displaced the United States as the world’s leading financial center last year. Many would agree with the economist Arvind Subramanian, who has argued that by 2030 China will be the world’s sole superpower, and that it is already the “world’s largest banker.”
These assessments see power as a result of the internal attributes of national economies: large economies with attractive financial sectors have power, while weaker ones do not. Accordingly, the U.S. decline in the share of global trade and income, and its domestic financial instability, should diminish its influence. But this focus fails to consider the ways in which the global financial network is, in fact, a complex and adaptive system. Power within this system does not depend solely on domestic attributes, but on the distribution of financial relationships that exists globally. In other words, the most well-connected economies, not just the biggest, are the most powerful. By extension, change within this structure does not follow a linear process, and economies that are initially more advantaged will continue to grow as the system develops.
The difference between these two approaches is significant. When we conceptualize the international financial system as a network, we see that the U.S. has become more central since 2007, not less. Rather than shift from West-to-East, global financial actors have responded to crisis by reorganizing around American capital to a remarkable extent. This is partially due to proactive responses to the crisis by policymakers such as the Federal Reserve, but it is also the result of factors outside the U.S. Above all, American capital markets remain attractive because complex networks contain strong path dependencies, which reinforce the core position of prominent countries while keeping potential challengers in the periphery. That is to say, policymakers and market players were limited in the decisions they could take because of factors that had already been locked in. As a result, the structure of the global financial system keeps the U.S. at the core and will continue to do so unless the entire network is fragmented, as it was during the 1930s when Great Britain lost its dominance.
Some who do see continuing U.S. financial resiliency contend that American power serves to the disadvantage of smaller countries. Indeed, they are correct that when a crisis occurs in the core – where the U.S. remains — the effects are felt throughout the system. But they miss the fact that American prominence also provides important stabilization mechanisms that can contain crises. To explain this, we need to look at what network scientists call “topology,” which refers to the organization of the components of a network, whether we are looking at a computer system or a financial system.
Once we view the international financial system in this context, we see that it is robust when facing crises in peripheral countries, but fragile when facing crises occurring in the core. This explains why the U.S. subprime crisis destabilized the global economy, while upheavals such as the 1990s East Asian crisis did not. Even the euro zone crisis has remained localized, to this point. A network perspective also explains how policy interventions by the U.S. prevented the collapse of the global system, thus ensuring that U.S. centrality persists. Finally, a network model should make us more cautious about promoting policies meant to erode U.S. financial hegemony. In fact, American centrality contained crises in peripheral countries from spreading globally, and the U.S. government demonstrated both the capacity and the willingness to pursue monetary and fiscal policies to moderate crises emanating from its own banking system. Returning to a world in which the structure of global financial relationships devolves outside the U.S. would therefore reintroduce a type of systemic risk not seen since the 1930s.
Describing the global banking network
What does the international banking network look like? First, the U.S. is strongly central, with over 70 percent of all countries placing a substantial amount of their overseas portfolio assets in the U.S., according to the Bank for International Settlements. After that, the distribution of international holdings is widely dispersed. The U.K. is the next most central, with about 35 percent of all countries significantly tied to its banking system. But most countries are only weakly tied there, even those with large financial sectors. Moving to Asia, Hong Kong — which supposedly passed New York and London as the world’s preeminent financial center — attracts a large amount of finance from fewer than 5 percent of the world’s economies. Mainland China barely exists in these networks, because the yuan is not convertible and foreign investment is tightly regulated.
Moreover, this network topology has reinforced itself over time, as ties to the U.S. have become increasingly strong. That progression paused briefly as a result of the 2007-08 shock, but quickly resumed. Perhaps most astonishingly, the U.S. has actually become more central in the aftermath of the crisis. The current international banking system is what we would call “hierarchical” – with the U.S. at its core and most other countries’ banking systems in the periphery – and displays dynamics of preferential attachment that reinforce this kind of “system hierarchy” through time.
When U.S. centrality helps, and when it hurts
What are the implications of a hierarchical financial network on international financial stability, and what does it mean for policy responses to crises? U.S. centrality, on the whole, is positive for system stability because the U.S. – with its deep, liquid banking system – is best poised to absorb and manage banking losses. When a peripheral country experiences a banking crisis it transmits its losses to foreign banks with which it has cross-border obligations, mainly U.S. financial institutions. In such cases, peripheral-country losses represent only a relatively small percentage of global banking assets, so U.S. banks are generally able to assume losses without collapsing. In this way, U.S. centrality acts as an important buffer to most interstate banking crisis contagion.
Consequently, a U.S.-centered global banking system is robust enough to withstand crises emanating from peripheral countries. This claim may seem counter-intuitive as well as controversial. But if we understand network dynamics, we can see the benefit of an international financial system centered on the U.S. And given preferential attachment, the persistence of U.S. centrality speaks to its positive effect: if U.S. centrality were in fact destabilizing, this kind of network manifestation would have quickly disintegrated after its initial formation.
That said, under certain conditions, U.S. centrality could be quite destructive to the stability of the global financial system. Because U.S. banking obligations represent a sizable percentage of bank balance sheets in most countries, most economies are vulnerable to crises that originate in America. Unlike the U.S., these peripheral countries are unable to absorb losses because U.S. obligations represent a large portion of their portfolios. Thus, a U.S.-centered global banking system is fragile to crises emanating from the core.
What this means for crisis response
This leads us to the next question: Why did the U.S. subprime crisis, which rippled throughout the international banking system, not ultimately destroy it? Indeed, as our analysis shows, the U.S. actually became more central to the system in the aftermath of the crisis. Ultimately, the answer is a combination of three factors: the lack of better alternatives, policy responses in the core, and the actions that banks took to protect themselves against risk. All of these responses strengthened ties between the U.S. and peripheral national banking systems.
First, while the subprime crisis illustrated the vulnerabilities in the U.S. banking system, financial actors continue to face limited alternatives when determining where to place assets other than in the U.S. The euro zone, deeply embroiled in its own debt crises, is not an attractive alternative. China, despite its economic growth, is an unviable center because its currency is not internationalized and its banking system is closed. Hong Kong and Switzerland are too small to provide sufficient liquidity or offer the kind of policy capacity to provide stability in the face of crisis. Other large emerging economies such as Brazil, Russia, and India have not developed large, internationally integrated financial markets. Given a lack of fit alternatives, the U.S. remains central.
Second, once the crisis unfolded, the U.S. pursued domestic and international policies that helped to stem financial losses. The Federal Reserve took swift action to unfreeze credit markets and injected massive amounts of liquidity into the global system. In an unprecedented move, the Fed also lent billions of dollars to distressed foreign firms during the crisis: 10 of the top 20 borrowers from Fed emergency lending programs were European banks; an 11th was Japanese. Fed Chairmen Ben Bernanke, an economist who understands how important global cooperation is in the face of systemic crisis, opened swap lines with almost every significant central bank in the world. Unlike Britain in the 1930s, the U.S. was able to pursue policies to reinforce the structure of the system because of its central position within that structure; it was willing to do so because it knew staying at the center confers economic and geopolitical advantages.
Finally, banks actually increased their exposure to the U.S. banking system in the aftermath of the crisis. They did so because, despite the crisis, American banks are still considered safer than others. The U.S. banking system is deep and liquid. As discussed above, the U.S. has the capacity to moderate crises by supporting systemically important banks, and its government demonstrated its willingness to pursue such stabilization policies in the face of crisis. This increased reliance on the U.S. illustrates an important feature of a network characterized by preferential attachment: once a hierarchical network is created, destruction of this system would obliterate substantial wealth, so banks have incentives to shore up the core and re-constitute the network. This dynamic is nicely demonstrated by market responses to the 2011 downgrade of the U.S. sovereign debt rating. While any other country would have seen interest rates spike after a downgrade, interest rates on U.S. government bonds actually fell.
Revising lessons learned
Many pundits claim the lesson of the Great Recession is that American financial dominance is both dangerous and waning. Our conceptualization of the international financial system as a complex network leads us to a wholly different set of conclusions.
First, the optimal policy response to crises depends on where within the network the crisis occurs. A crisis in the core requires quite different actions than does a crisis in the periphery: the latter has very little chance of generating systemic crisis and can be managed by institutions like the International Monetary Fund. While the hardship of people suffering in Greece and other countries hit by crises is very real, these peripheral calamities do not pose significant threats to the system as a whole. Crises in the U.S., by contrast, can quickly destabilize the entire financial system. Accordingly, the policy response to these crises must be designed to support the structure of the system, not change it. This requires significant global policy coordination designed to maintain liquidity and prevent systemic collapse.
Second, the actions that peripheral countries take to insulate themselves against crisis often paradoxically reinforce U.S. centrality. This dynamic is generally a net positive for network stability so long as macroeconomic imbalances are managed carefully. In Asia, for example, countries amassed foreign exchange reserves as self-insurance against negative current account shocks – a move that further entrenched U.S. centrality, since the dollar is the primary global reserve currency. As discussed above, the dollar’s primacy allows the U.S. to respond to negative shocks with large injections of liquidity into the global financial system. U.S. financial centrality, along with its demonstrated willingness to respond to shocks with the policies necessary to keep markets liquid, raises investors’ confidence in its financial system while tying their financial health to the strength of this network. This leads investors to continue to buy U.S. debt, further allowing the U.S. to borrow cheaply even after ratings agencies downgrade or threaten to downgrade U.S. debt. Indeed, our argument complements economist Barry Eichengreen’s conclusion in his book Exorbitant Privilege that the U.S. dollar will persist as the main global reserve currency due to lack of a fit alternative.
Third, our argument about “preferential attachment” suggests that the emergence of a better alternative is a rare phenomenon, since positive feedback perpetually structures the network around the center. Moreover, even if such an alternative did exist, we would not see a change to the network structure unless a shock emanating from the core was large enough to destroy the topology of international relationships. Even when shocks are large enough to destroy the system, core countries may pursue policies that prevent network collapse. Thus, preferential attachment makes the emergence of a new financial hegemon highly unlikely. At the same time, because so many countries are highly connected to the center, hierarchical networks are susceptible to hugely disruptive tail events with systemic effects. One important implication of our research is that it may be incredibly difficult to predict when we might experience a destabilizing crisis — but we do know that such a crisis must originate in the U.S.
Finally, our network model describes the necessary and sufficient conditions for the persistence and change of systemic orders. It does not mean the U.S. will forever be the center of the international financial system. But for the U.S. to be dethroned, three things need to happen jointly: a severe crisis in the U.S., the inability or unwillingness of the U.S. government to pursue policies necessary to stabilize the system, and the emergence of an alternative that is ready to move into a position at the center. Network dynamics suggest that under these three conditions, a network transformation could happen quite rapidly. Historically, such transitions have been accompanied by political instability and major power conflicts. The most recent transition – from British to American financial centrality – occurred within the context of two World Wars and a Great Depression. It appears that none of these conditions exists at present. Because of the persistence created by preferential attachment, policies aimed at eroding U.S. centrality probably will not work. But if they did, world order would be threatened.