US FINANCIAL INTERESTS AND ESF BAILOUT SELECTION
The international expansion of US bank lending that began in the late 1970s and launched the modern era of financial globalization meant that financial disturbances in foreign economies were no longer isolated to the country of origin. They now had direct channels through which they could spill over into and threaten US financial markets. While hundreds of financial institutions in the United States make loans outside of the United States, the majority of this foreign lending was concentrated in the hands of a few banks known today as systemically important financial institutions (SIFIs). As Lawrence Broz and Michael Brewster Hawes explain, SIFIs "specialize in wholesale and international banking and are located in financial centers such as New York, Chicago, and San Francisco. Their clients include governments, corporations, and other banks.” Institutions like Bank of America Corp., Citigroup Inc., JPMorgan Chase & Co., and
SIFI’s Share of US Banks’ Foreign Claims, 1982-1998
Wells Fargo & Co. fall into this category. SIFIs lie at the core of domestic banking systems and are viewed as having more "systemic importance”— defined as "the damage a bank’s failure inflicts upon the rest of the system”—than smaller banks. For instance, for the entire period under investigation, SIFI’s foreign claims have comprised more than 55 percent of total US foreign bank claims. Figure 5.2 visually displays this by plotting the total aggregate (nominal) foreign claims of these supersized banks alongside all other reporting US banks from 1982 until 1998. The figure also reports SIFI claims as a share of all US bank claims.
I argue that US economic policymakers were most likely to deploy ESF resources on behalf of foreign economies in crisis where these sys- temically important US banks were most at risk. The IMF’s unresponsiveness, due to the concerted lending strategy, during the 1980s debt crisis increased the likelihood that a sovereign would be forced to default on its debt obligations to US banks. In the 1990s, the IMF’s inability to provide sufficiently large credits meant that Fund programs alone may have left the crisis unresolved, leaving US banks and other financial institutions active in those economies at great risk. Thus, I expect that elevated exposure of SIFIs should be associated with an increase in the likelihood of a US financial rescue. Major commercial banks are a well-organized, well- financed political lobby. Banks are also keenly aware that financial crises in foreign jurisdictions threaten their profits—and even their health—if a sufficient percentage of their claims are concentrated in that location. They also know that Treasury has the capacity to provide rescue packages to these countries in crisis. Additionally, banks clearly benefit from these kinds of bailouts. Research has shown that bank stock prices rise when the IMF announces it will be bailing out countries to which the banks are exposed. Another study found that when a country facing a financial crisis receives an emergency loan, it uses the money to pay back its private creditors. When SIFIs’ interests are threatened by foreign financial crises, I anticipate that they will press top US economic policymakers to take steps that ensure the health and profitability of their institutions—steps like providing foreign ESF credits.
Yet this claim implies that decision making at Treasury was captive to Wall Street, that Treasury made foreign bailouts simply to protect the private interests ofbig banks and policymakers have no agency. Moreover, this is inconsistent with the broader argument of this book, which asserts that US ILLR actions are designed to protect vital US economic interests— not private financial interests. The likely role of the banking lobby in this story is undeniable. However, it is far too simplistic to depict top US policymakers as marionettes and big banks as holding all the strings. They are also individuals, operating inside state institutions, with their own interests in policy. I expect that policymakers prefer policy choices that increase the likelihood their institution will live up to its mandate. With respect to these cases, the key institutions are the US Treasury and, to a lesser extent, the Federal Reserve. Although these institutions have missions encompassing a number of roles, each is charged with providing a key public good for the US economy: protecting and providing for the stability of the US financial system, broadly construed. In other words, policymakers within these institutions are ordained as the guardians of
US financial stability. I expect their decisions reflect a desire to fulfill their mission.
If exposures to a potential default on the part of multiple foreign borrowers are substantial enough, policymakers may feel that a crisis poses systemic risk to the US financial system. To put it differently, policymakers have an interest in intervening when they believe the probability of a "worst-case scenario”—where the broader financial system faces the prospect of great instability as a result of foreign spillovers— is high enough to justify intervention. Thus, I expect that the effect of SIFI exposure on the likelihood of US foreign bailouts to be conditional on the level of systemic risk facing the US financial system at a given moment. Policymakers should be most concerned about commercial bank losses when the risks facing the entire financial system are elevated. The basic point is this: Individual financial crises do not happen in a vacuum. They occur in a broader international context. Some crises are relatively isolated, occurring in an otherwise stable environment. In such cases, so long as they have sufficient capital in reserve, financial institutions should be capable of weathering losses from the crisis without seriously jeopardizing the health of the US economy. However, there are also moments when individual crises occur within a far more dangerous context. Contagion can cause multiple financial fires to burn at once. Under these circumstances, systemic risk is higher as banks face the possibility of foreign losses on multiple fronts. Threats from individual crises are intensified in this environment. Concerns about the capital adequacy of banks generate fears that banks might fail, threatening the stability of the entire domestic financial system. In such cases, more than just Wall Street profits are at risk. The broader public interest is also in danger. When the health of the system is in peril, policymakers should be most likely to act defensively by providing bailouts to those countries where the risks to SIFIs are the greatest. Thus, at the domestic level, US foreign rescues reflect a joint product model where two outputs are produced by the same process: protecting the private financial interests of major banks while also protecting the stability of the national financial system.
The argument is presented visually in Figure 5.3. I expect that the United States will most likely provide ESF credits to countries where SIFIs are highly exposed when the systemic risk facing the broader US financial system from foreign crises is high. This corresponds to quadrant B in Figure 5.3. In such cases, the pleas of the bank lobby should be very intense because their survival may be in question. Additionally, US economic policymakers should be most sensitive to the pleas of the bank lobby in such times. Here, the private interests of the banks coincide with
Systemic Risk and Bank Exposure Interaction
the broader public interests that policymakers are ordained to serve and protect.
Yet, because the ESF’s resources are finite, policymakers’ actions are constrained. Consequently, under such circumstances, Treasury should prefer to deploy ESF resources where they will have the greatest effect on stabilizing US markets. Here, the banking lobby acts as an important source of information by relaying to policymakers where they face the greatest risks. All else equal, both Treasury and the big banks should prefer to deploy resources on behalf of countries where SIFIs are most exposed. Conversely, bailouts should be less likely on behalf of countries where SIFIs have little exposure and when international financial waters are calm. This corresponds to quadrant C. Here, banks will have little incentive to ask for protection and policymakers should be reluctant to offer it.
Turning to quadrant A, Treasury is unlikely to provide ESF credits to countries where SIFIs are not highly exposed—even when systemic risk is high. Because ESF resources are finite, when multiple financial fires 11
break out at once, I expect Treasury to respond to those that pose the greatest risk to the system. Moreover, in such circumstances, I anticipate that the banking lobby will press policymakers to rescue those economies where they are most exposed. Finally, I expect that policymakers are somewhat likely to provide bailouts to countries where SIFIs are highly exposed, even when systemic risk is low. This corresponds to quadrant D. In these cases, I expect the banking lobby to press Treasury to assist those countries where their claims are the greatest. Yet, in this context, policymakers should be reluctant to act, given that the system does not appear to be under threat. Bailouts in this context would essentially represent an indirect means of protecting profits of big banks. At the same time, in this context, ESF resources are less likely to be under strain from fighting multiple fires at once. Thus, policymakers may be somewhat more amenable to requests by the banks. Additionally, Treasury officials may be persuaded that such a rescue is preemptive. Inaction could result in contagion. As other economies become similarly afflicted, systemic risk will elevate, requiring additional bailouts in the future. Thus, while cases in quadrant D should be much less likely to correspond with ESF bailouts than those in quadrant B, they should be more likely to result in rescues than those in quadrants A and C. The remainder of this chapter presents an empirical model of ESF bailout selection where I test my argument.
-  In the past, these were referred to as "money center” banks. They are also sometimescalled "large financial institutions” (LFIs).
-  Broz and Hawes 2006, p. 376.
-  The total number of banks classified as SIFIs has varied from as many as nine in 1982 toas few as six in 1998.
-  Craig and von Peter 2010, p. 22.
-  Demirgu^-Kunt and Huizinga 1993; Kho, Lee, and Stulz 2000; Lau and McInish 2003;Zhang 2001.
-  Bird 1996.
-  Because of the substantial resources at their disposal and their small numbers, SIFIsshould be more likely to overcome collective action problems and successfully lobby financial authorities for protection in the form of ESF bailouts (Olson 1965). Two existing studieshave found convincing evidence that this type of lobbying influences congressional votingbehavior (Broz 2005; Broz and Hawes 2006).
-  Treasury is the “steward of US economic and financial systems” with a mission to “[protect] the integrity of the financial system.” Additionally, Treasury “works with other federalagencies, foreign governments, and international financial institutions . . . to the extent possible, predict and prevent economic and financial crises.” Similarly, one of the Fed’s primarypurposes is “maintaining the stability of the financial system and containing systemic riskthat may arise in financial markets.” For more information on the Treasury’s mission, seehttp://www.treasury.gov/about/role-of-treasury/Pages/default.aspx. For details regardingthe Fed’s mission, see http://www.federalreserve.gov/faqs/about_12594.htm.
-  Gilpin (1975, p. 142) makes a similar argument explaining the spread of US multinationals around the world, noting that “corporate interests and the ‘national interest’. . .coincided.”