Microsoft Word - 2_2017.3_Awrey_Final.docx AWREY – FINAL BROTHER, CAN YOU SPARE A DOLLAR? DESIGNING AN EFFECTIVE FRAMEWORK FOR FOREIGN CURRENCY LIQUIDITY ASSISTANCE Dan Awrey* The core principles of financial crisis management call upon central banks to lend freely, against good quality collat- eral, and at a penalty rate of interest, to solvent but illiquid banks and other financial institutions during periods of wide- spread panic and instability. While often taken for granted, these principles were designed for a world in which central banks have the capacity to create money denominated in the same currency as the one in which domestic banks and other financial institutions issue deposits and other short-term lia- bilities. Unfortunately, this is not the world in which we live. The application of these principles is far from straightforward in a world where financial institutions rely on short-term foreign currency liabilities as a source of financing. This is the world of the Eurodollar market. The global financial crisis vividly illustrated the potential systemic risks arising from the exist- *Associate Professor of Law and Finance and Fellow of Linacre College, University of Oxford. The author would like to thank Weitseng Chen, Chris- tine Desan, Daniela Gabor, Anna Gelpern, Jeff Gordon, Robert Hockett, Rosa Lastra, Saule Omarova, Katharina Pistor, Morgan Ricks, Annelise Riles, Paul Lejot, Matthew Turk, John Vickers, and David Zaring, along with the participants of conferences organized by the Journal of Financial Regulation, the Columbia Center on Global Legal Transformation, the American Society of International Law, the National University of Singa- pore Centre for Banking and Financial Law, and the University of Sheffield Centre for Corporate and Commercial Law for their extremely helpful com- ments and suggestions. The author would also like to acknowledge the gen- erous support of the National University of Singapore Centre for Banking and Financial Law. All errors remain my own. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 935 ence of a large Eurodollar market. Faced with a systemic for- eign currency liquidity crisis, central banks struggled to secure access to the foreign currency reserves needed to provide emer- gency liquidity assistance to their domestic banking systems. In response, the U.S. Federal Reserve and other major central banks established a network of swap lines designed to provide foreign currency liquidity assistance to the international finan- cial system. The central bank swap lines have been hailed as one of the most important and effective policy responses to the financial crisis. However, while it may be tempting to view them as an effective prophylactic against future foreign currency liquidity crises, the current structure of the swap lines fails to establish truly credible international commitments or constrain the moral hazard problems stemming from this ambitious state- sponsored liquidity insurance. This Article examines the unique policy challenges posed by foreign currency liquidity problems, along with how to build a more effective framework for the provision of foreign currency liquidity assistance. I. Introduction ................................................................ 936 II. The Origins, Mechanics, and Growth of the Eurodollar Market ........................................................................ 948 III. Financial Crisis Management and the Problem of FCLA ..................................................................................... 955 A. The Fundamentals of Emergency Liquidity Assistance ............................................................ 956 B. The Problem of FCLA .......................................... 968 IV. The Eurodollar Market and the Financial Crisis ..... 972 V. Designing an Effective Framework for FCLA .......... 985 A. A Blanket Prohibition Against Eurodollar Liabilities ............................................................. 986 B. A Global Currency and Central Bank ................ 991 C. Reform of Existing Institutional Arrangements for FCLA .................................................................... 997 VI. Conclusion ................................................................ 1015 AWREY – FINAL 936 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 I. INTRODUCTION The massive public-sector support for banks and other fi- nancial institutions at the height of the global financial crisis has understandably attracted a maelstrom of controversy. One of the most controversial—and yet surprisingly under- scrutinized—aspects of this support was its extension to fi- nancial institutions incorporated and licensed to carry on business in other countries.1 In the United States, eighteen of the twenty-five largest borrowers under the emergency Term Auction Facility (“TAF”) established in December 2007 were European banks.2 Foreign banks also ultimately received bil- lions of dollars under Maiden Lane II and III, the vehicles that the Federal Reserve Bank of New York used to provide sup- port to global insurance giant American International Group (“AIG”).3 And perhaps most strikingly, the Federal Reserve (or, the “Fed”) extended over $USD500 billion at its peak to 1 See, e.g., Alan Grayson, Alan Grayson: “Which Foreigners Got the Fed’s $500,000,000,000?” Bernanke: “I Don’t Know.”, YOUTUBE (July 21, 2009), https://www.youtube.com/watch?v=n0NYBTkE1yQ (showing then Rep. Alan Grayson of Florida questioning then Federal Reserve Board Chairman Ben Bernanke before the House Financial Services Committee on the identity of the foreign financial institutions that received this sup- port). This controversy continued to rage on well after the acute phase of the crisis. See, e.g., Gerald P. O’Driscoll Jr., The Federal Reserve’s Covert Bailout of Europe, WALL ST. J. (Dec. 28, 2011), http://www.wsj.com/arti- cles/SB10001424052970204464404577118682763082876. 2 Efraim Benmelech, An Empirical Analysis of the Fed’s Term Auction Facility, 2 CATO PAPERS ON PUB. POL’Y 57, 78–79 (2012). In total, foreign banks received 58% of the total amount of money lent through TAF. Id. at 75. For details of borrowing under TAF, see Term Auction Facility (TAF), BD. OF GOVERNORS OF THE FED. RES. SYS., https://www.federalre- serve.gov/newsevents/reform_taf.htm [perma.cc/6PJR-43F2]. 3 See CONG. OVERSIGHT PANEL, JUNE OVERSIGHT REPORT: THE AIG RES- CUE, ITS IMPACT ON MARKETS, AND THE GOVERNMENT’S EXIT STRATEGY 87–94 (2010), https://www.gpo.gov/fdsys/pkg/CPRT-111JPRT56698/pdf/CPRT- 111JPRT56698.pdf [perma.cc/AC7G-4EKR]. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 937 foreign banks and other financial institutions under the aus- pices of temporary swap lines established with other major central banks.4 This extraordinary public support for foreign banks and other financial institutions is often defended as having been necessitated by the high level of integration and interconnect- edness within the global financial system. When the United States, Europe, or other major financial centers get a sniffle, the rest of the world stands at risk of catching a cold.5 One of the primary drivers of this integration and interconnected- ness stems from the dominant role of the U.S. dollar in inter- national trade, finance, and investment.6 At the heart of this role is an incredibly important yet relatively obscure corner of the financial system known as the “Eurodollar” market. We tend to think of money as something over which states enjoy absolute sovereignty.7 This thinking reflects the critical 4 See infra Part IV and Part V for further details regarding the opera- tion of these central bank swap lines and how they were employed during the financial crisis. 5 Indeed, this is how the Federal Reserve itself has framed this ra- tionale in various public statements, including a video designed to explain why the central bank swap lines are in the national interest of the United States. See Protecting the U.S. Economy from Strains Abroad—Why Swap Lines Are in the U.S. National Interest, FED. RES. BANK N.Y. (Sept. 19, 2017), https://www.newyorkfed.org/education/liqswap.html. 6 For a good and highly readable description of this role, see generally BARRY EICHENGREEN, EXORBITANT PRIVILEGE: THE RISE AND FALL OF THE DOLLAR AND THE FUTURE OF THE INTERNATIONAL MONETARY SYSTEM (2011). 7 Indeed, while reasonable scholars can and do disagree about its fun- damental nature and importance, there has historically been a remarkable degree of convergence around the view that the locus of monetary sover- eignty is the state. See Benjamin J. Cohen, The International Monetary Sys- tem: Diffusion and Ambiguity, 84 INT’L AFFAIRS 455, 464 (2008) (“Tradition has long assigned the primary role in monetary governance to the sovereign state.”). This state-centric view has a long and distinguished pedigree in Western political thought. Thomas Hobbes, John Stuart Mill, and Adam Smith each acknowledged the central role of the state in monetary affairs. See THOMAS HOBBES, LEVIATHAN 189 (Michael Oakeshott ed., Collier Books 1962) (1651) (“. . . money, of what matter soever coined by the sovereign of a commonwealth, is a sufficient measure of the value of all things else, be- tween the subjects of that commonwealth. By the means of which measure, all commodities, movable and immovable, are made to accompany a man to AWREY – FINAL 938 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 role that states play in establishing and maintaining the core legal and economic institutions around which modern mone- tary systems are built.8 States define which instruments qual- ify as “legal tender”, prescribe what means of payment can be used to pay taxes, and oversee the process of minting coins all places of his resort, within and without the place of his ordinary resi- dence; and the same passeth from man to man, within the commonwealth; and goes round about, nourishing, as it passeth, every part thereof; in so much as this concoction, is as it were the sanguification of the common- wealth: for natural blood is in like manner made of the fruits of the earth; and circulating, nourisheth by the way every member of the body of man.”); JOHN STUART MILL, PRINCIPLES OF POLITICAL ECONOMY WITH SOME OF THEIR APPLICATIONS TO SOCIAL PHILOSOPHY 151 (1848) (“So much of barbarism . . . still remains in the transactions of the most civilized nations, that almost all independent countries choose to assert their nationality by having, to their own inconvenience and that of their neighbours, a peculiar currency of their own.”); ADAM SMITH, AN INQUIRY INTO THE NATURE AND CAUSES OF THE WEALTH OF NATIONS 410–11 (Thomas Dobson 1796) (1776) (“A prince, who should enact that a certain proportion of his taxes should be paid in a paper money of a certain kind, might thereby give a certain value to this paper money.”). Benjamin Franklin, meanwhile, viewed the ability of a state to issue its own money as a prerequisite of sovereignty. See Pavlina R. Tcher- neva, Money, Power, and Monetary Regimes 13 (Levy Econ. Inst. of Bard Coll., Working Paper No. 861, 2016) (citing Karl Rhodes, The Counterfeiting Weapon, 16 REGION FOCUS 34, 34–35 (2012). For some insight into Frank- lin’s views on the importance of money to the economic prosperity of the American colonies, see The Nature and Necessity of a Paper-Currency, 3 April 1729, NAT’L ARCHIVES (last updated June 29, 2017), http://founders.ar- chives.gov/documents/Franklin/01-01-02-0041 [perma.cc/GTT8-3Y2C]. For more recent treatises reflecting this state-centric view, see ROSA MARÍA LASTRA, INTERNATIONAL FINANCIAL AND MONETARY LAW 3–4 (2d ed. 2015); CHARLES PROCTOR, MANN ON THE LEGAL ASPECT OF MONEY 500–01 (6th ed. 2005). 8 This view is often described as the “chartalist” theory of money. See 1 JOHN MAYNARD KEYNES, A TREATISE ON MONEY: THE PURE THEORY OF MONEY 4–7 (reprinted 1953) (1930); GEORG FRIEDRICH KNAPP, THE STATE THEORY OF MONEY 31–32 (reprinted 1973) (1924); Abba P. Lerner, Money as a Creature of the State, 37 AM. ECON. REV. 312 (1947). For a detailed historical study of the role of the state in the development of money and monetary institutions in the United Kingdom, see CHRISTINE DESAN, MAKING MONEY (2014). The “chartalist” view of money is often contrasted with the more market-ori- ented or “metallist” view. See Charles A.E. Goodhart, The Two Concepts of Money: Implications for the Analysis of Optimal Currency Areas, 14 EUR. J. POL. ECON. 407 (1998). AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 939 and printing paper currency. States also establish central banks such as the Federal Reserve and endow them with the power and responsibility to manage the money supply in pur- suit of price stability, full employment, and other policy objec- tives.9 Viewed solely from this perspective, states appear to exercise almost complete control over the legal, fiscal, and even physical machinery necessary to create money and en- sure its widespread use amongst its citizenry. This state-cen- 9 The monetary policy objectives of the Board of Governors of the Fed- eral Reserve System and the Federal Open Market Committee, for example, are to “maintain long run growth of the monetary and credit aggregates commensurate with the economy’s long run potential to increase production, so as to promote effectively the goals of maximum employment, stable prices, and moderate long-term interest rates.” Federal Reserve. Act, 12 U.S.C. § 225a (2000). See also LASTRA, supra note 7, at 55–63, for a useful introduction to the different objectives central banks pursue. AWREY – FINAL 940 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 tric view has long been recognized in common law jurispru- dence10 and as a matter of public international law.11 It is also deeply embedded in the everyday language of money: when we talk about money, we talk about the U.S. dollar, British pound, and Icelandic króna.12 10 As the English Privy Council observed as far back as 1605, money “inheres in the bones of princes.” DESAN, supra note 8, at 170 (citing The Case of Mixed Money (1605), 118 (“Monetandi jus principum ossibus in- haeret, Jus monetae comprehenditur in regalibus, quae nunquam a region sceptro abdicantur.”)). Over the centuries, the concept of monetary sover- eignty has evolved in response to fundamental changes in the political land- scape and the corresponding shift in the identity of the sovereign from ab- solute monarchs to democratically elected governments. See BENN STEIL & MANUEL HINDS, MONEY, MARKETS, AND SOVEREIGNTY 240 (2009) (“The trans- formation of the concept of sovereignty reflects the evolution over centuries of the dominant system of government from monarchy to democracy, with the mythology of the king as the embodiment of the popular will being sub- stituted for by the elected executive or legislature.”); Claus D. Zimmerman, The Concept of Monetary Sovereignty Revisited, 24 EUR. J. INT’L L. 797, 808 (2008) (“At the time when the concept of monetary sovereignty first ap- peared, in order to support the exercise of the royal prerogative to coin money as exercised by absolutist monarchs, the locus of both sovereignty and the power to exercise it might still have been identical, but times have obviously changed. The contemporary mainstream view of states being in- struments at the service of their peoples as true holders of sovereignty may be regarded as a corollary of the fundamental idea of popular sovereignty or sovereignty of the people.”) (footnote omitted). Yet despite these profound changes, the concept has retained its state-centric orientation and remains deeply rooted in the “one state, one currency” model that is observed in most of the world today. For a comprehensive overview of this jurisprudence, see PROCTOR, supra note 7, at chs. 3–8, 13 (describing, inter alia, the private law of monetary obligations, related conflicts of law issues, and the lex mon- etae principle). 11 For a general overview, see PROCTOR, supra note 7, at chs. 19–23 (de- scribing various aspects of public international law in relation to money). The recognition in public international law of a state’s sovereignty over its monetary affairs dates at least as far back as a 1929 decision of the Perma- nent International Court of Justice. Payment of Various Serbian Loans Is- sued in France (Fr. v. Kingdom of the Serbs, Croats, and Slovenes), Judg- ment, 1929 P.C.I.J. (ser. A) No. 20, at 25 (July 12, 1929) (“It is indeed a generally accepted principle that a State is entitled to regulate its own cur- rency.”). 12 The euro and, more recently, crypto-currencies, such as bitcoin, are obvious exceptions. While this Article generally uses the terms “country” or AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 941 While often far less appreciated, states also play an im- portant role in licensing private money creation.13 States au- thorize the establishment of deposit-taking banks and pro- hibit all firms other than banks from issuing deposit liabilities.14 States then subject banks to portfolio restrictions, capital and liquidity requirements, and other forms of pruden- tial regulation and intensive prudential supervision.15 States also provide banks with various forms of support not generally available to other commercial enterprises. This support in- cludes deposit guarantee schemes, emergency liquidity assis- “state” to describe the political units that issue currencies, these terms should be interpreted as encompassing, mutatis mutandis, political units such as the Euro Area that issue “supranational” currencies. 13 See generally Robert C. Hockett & Saule T. Omarova, The Finance Franchise, 102 CORNELL L. REV. 1143 (2017) (describing the modern finan- cial system as a public-private partnership characterized by the public ac- commodation and monetization of private liabilities). 14 The unique privilege bestowed upon banks to issue deposit liabilities is reflected in the legal definition of a “bank” in most jurisdictions. In the European Union, for example, a “credit institution [bank] means an under- taking the business of which is to take deposits or other payable funds from the public and to grant credits for its own account . . .” (emphasis added) (internal quotation marks omitted). Regulation (EU) No 575/2013 of the Eu- ropean Parliament and of the Council of 26 June 2013 on Prudential Re- quirements for Credit Institutions and Investment Firms and Amending Regulation (EU) No 648/2012, art. 4.1(1), 2013 O.J. (L 176) 18. There are several definitions of a “bank” under U.S. law. For example, a bank is de- fined as “an institution organized under the laws of the United States . . . which both—(i) accepts demand deposits or deposits that the depositor may withdraw by check or similar means for payment to third parties or others; and (ii) is engaged in the business of making commercial loans.” U.S. Bank Holding Company Act of 1956, 12 U.S.C. § 1841(2)(c)(1) (2012) (emphasis added). Section 3(a)(2)(A) of the Federal Deposit Insurance Act, in contrast, defines a state bank as “engaged in the business of receiving deposits, other than trust funds” without a corresponding requirement that they also be engaged in the business of making commercial loans. 12 U.S.C. § 1813(a)(2)(A) (2012). These definitions are typically accompanied by entry restrictions prohibiting all firms other than banks from issuing deposit lia- bilities. See, e.g., 12 U.S.C. § 378(a)(2) (2012). 15 For a more detailed description of this prudential regulation and su- pervision, see JOHN ARMOUR ET AL., PRINCIPLES OF FINANCIAL REGULATION chs. 14, 15, 26 (2016). AWREY – FINAL 942 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 tance or “lender of last resort” facilities, and special bank- ruptcy or “resolution” regimes for failing banks.16 Collectively, this support enhances the credibility of a bank’s commitment to honor its deposit and other short-term liabilities. This cred- ibility is further enhanced by the willingness of the state to accept deposit liabilities in satisfaction of tax and other obli- gations. The role of the state in licensing, regulating, and sup- porting banks thus serves to entrench deposit liabilities as close substitutes for coins, paper currency, and other forms of state money. Once again, it is the state—this time in its ca- pacity as the guardian of the conventional banking system— that exercises control over the machinery of money creation. If only it were so simple. Recent decades have witnessed the gradual erosion of the privileged and mutually beneficial relationship between the state, banks, and private money cre- ation. In retail banking, money market mutual funds have emerged as a substitute for conventional deposit and checking accounts.17 In commercial and wholesale banking, money market funds, short-term commercial paper, and repurchase (or “repo”) agreements have grown to rival the more tradi- tional savings products offered by banks.18 Equally important, although again often far less appreciated, banks and other fi- nancial institutions incorporated and licensed in one country have increasingly adopted the practice of issuing deposits and other short-term liabilities denominated in the currency of an- other country. Banks in London offer U.S. dollar deposit ac- 16 For an in-depth look at this support, see id. at chs. 15, 16. For a dis- cussion of how such support enhances the credibility of the commitments underpinning these liabilities, see Anna Gelpern & Erik F. Gerding, Inside Safe Assets, 33 YALE J. ON REG. 363 (2016). 17 For a good summary of the emergence of money market mutual funds as substitutes for bank deposits, see Jonathan Macey, Reducing Sys- temic Risk: The Role of Money Market Mutual Funds as Substitutes for Fed- erally Insured Bank Deposits, 17 STAN. J.L. BUS. & FIN. 131 (2011). 18 For a description of how and why commercial paper and repurchase agreements serve as substitutes for more traditional banking products, see ARMOUR ET AL., supra note 15, ch. 21. For a description of how money market mutual funds serve as substitutes for bank deposits, see Macey, supra note 17. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 943 counts; investment banks in Shanghai provide euro denomi- nated trade financing. The unfettered ability of financial in- stitutions to create these short-term foreign currency liabili- ties with little more than the stroke of a bookkeeper’s pen—or a few keys—effectively severs the cord between states and money creation. This is the world of the Eurodollar market. The Eurodollar market has been around for a very long time. Writing in 1873, the English essayist and father of mod- ern central banking, Walter Bagehot, described the predeces- sor to the modern Eurodollar market as an important compo- nent of London’s money market.19 Yet it is only in the past twenty years or so that this market has blossomed into a multi-trillion dollar behemoth, the health and stability of which has a significant impact on global economic activity and financial stability. The best-known Eurodollar market is the market for U.S. dollar denominated deposits held with foreign banks in countries like the United Kingdom. However, there are also significant Eurodollar markets for commercial paper, repo agreements, derivatives, and trade financing across a range of different countries and currencies.20 The most striking feature of the Eurodollar market is that it revolves around the issuance of short-term liabilities by fi- nancial institutions that are not licensed or subject to pruden- tial regulatory oversight in the country that issues the cur- rency in which these liabilities are denominated. Most importantly for the present purposes, these institutions do not enjoy access to the emergency liquidity assistance (“ELA”) fa- cilities provided by the central bank in the country that issues 19 See WALTER BAGEHOT, LOMBARD STREET: A DESCRIPTION OF THE MONEY MARKET 33 (reprinted 1999) (1873) (“Now that London is the clear- ing-house to foreign countries, London has a new liability to foreign coun- tries. At whatever place many people have to make payments, at that place those people must keep money. A large deposit of foreign money in London is now necessary for the business of the world.”). This nascent Eurodollar market differed from its modern counterpart in one very important respect: instead of national currencies such as the pound or dollar, it was gold bul- lion that served as the de facto international reserve currency. Id. at 44. 20 See infra Part II for further details of the origins, basic mechanics, and recent growth of these markets. AWREY – FINAL 944 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 the relevant currency. These ELA facilities are an essential tool of financial crisis management: providing collateralized loans to solvent but illiquid banks and other financial institu- tions in the event that private sources of financing become un- available.21 These facilities are thus often and understanda- bly viewed as the last line of defense against financial panic and instability before states are forced to resort to emergency bailouts. While financial institutions issuing Eurodollar lia- bilities may have access to the ELA facilities provided by the central bank in their home country, there is no guarantee that this central bank will hold or have access to sufficient reserves of the required foreign currency.22 This leaves these financial institutions heavily reliant on private markets as a source of foreign currency financing. When these markets break down, these institutions run the risk of finding themselves without a port in the middle of a dangerous storm. This fault line in the global monetary, regulatory, and cri- sis management architecture leaves the financial institutions that issue Eurodollar liabilities particularly vulnerable to li- quidity shocks of the variety experienced during the global fi- nancial crisis. Indeed, one of the most pernicious episodes of the financial crisis was the international U.S. dollar liquidity shortage that began in December 2007 and reached its apex in the autumn of 2008.23 This shortage forced banks and other financial institutions that relied on Eurodollar markets as a 21 See infra Section III.A for further information regarding the func- tions and institutional characteristics of these ELA facilities. See also Thomas M. Humphrey, Lender of Last Resort: The Concept in History, FED. RES. BANK RICHMOND ECON. REV., Mar./Apr. 1989, at 8, 8. For a brief history of the development of these facilities, see LASTRA, supra note 7, at 150–60; Mark A. Carlson & David C. Wheelock, The Lender of Last Resort: Lessons from the Fed’s First 100 Years (Fed. Res. Bank of St. Louis, Working Paper No. 2012-056B, 2013). 22 Indeed, as described infra Section III.B, central banks may incur sig- nificant costs in holding these foreign currency reserves. They may also be unable to acquire sufficient quantities of foreign currency on the open mar- ket during periods of widespread market disruption. 23 See infra Part IV for a more detailed description of the U.S. dollar liquidity shortage during the financial crisis. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 945 source of financing to either cut back on their U.S. dollar lend- ing activities or sell dollar denominated assets. The shortage thus contributed to the broader contraction of lending and fire sale dynamics that fanned the flames of the crisis.24 Ulti- mately, it was the desire to extinguish these flames that mo- tivated the Federal Reserve and other central banks to pro- vide such extraordinary support to foreign financial institutions. This Article examines the two deeply intertwined policy challenges arising from the widespread issuance of Eurodollar liabilities by banks and other financial institutions. The first challenge stems from the prospect of destabilizing cross-bor- der capital flows as foreign currency liquidity is withdrawn from the international financial system during periods of widespread financial instability. The second challenge stems from the significant practical constraints on the ability of cen- tral banks to respond to this withdrawal of liquidity by provid- ing financial institutions with ELA denominated in the re- quired foreign currency.25 As this Article will show, these 24 For a more detailed description of these fire sale dynamics and how they played out during the financial crisis, see Markus K. Brunnermeier, Deciphering the Liquidity and Credit Crunch 2007–2008, 23 J. ECON. PER- SPECTIVES 77 (2009). 25 While not widely acknowledged or studied in the academic literature, at least four scholars have previously identified aspects of this problem. See LASTRA, supra note 7, at 25 (“The growth in Eurocurrency markets entails a partial loss of monetary sovereignty.”) (footnote omitted); Tullio Treves, Monetary Sovereignty Today, in INTERNATIONAL MONETARY LAW: ISSUES FOR THE NEW MILLENNIUM 117 (Mario Giovanoli ed., 2000) (“The lack of the state of the currency’s coercive power [over Eurodollar markets] . . . means that such a state does not have the power effectively to regulate credit in its cur- rency . . . .”); Milton Friedman, The Euro-Dollar Market: Some First Princi- ples, FED. RES. BANK OF ST. LOUIS R., July 1971, at 16, 17 (“The location of the banks [that issue Eurodollar liabilities] is important primarily because it affects the regulations under which the banks operate and hence the way that they can do business.”); Edward J. Frydl, The Eurodollar Conundrum, FED. RES. BANK OF N.Y. Q. REV., Spring 1982, at 11 (“. . . the Euromarket could in the future become an important snag in monetary control . . .”). However, as this Article will show, even these prescient scholars failed to foresee the challenges posed by a large Eurodollar market in terms of effec- tive financial crisis management. AWREY – FINAL 946 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 challenges generate enormous risks for the financial institu- tions that issue Eurodollar liabilities, for the countries in which these institutions are incorporated and licensed and carry on business, and, ultimately, for global financial stabil- ity. Having framed these important challenges, this Article goes on to examine a range of policy alternatives that might be used to address them. At one end of the spectrum is a blan- ket prohibition against banks and other financial institutions issuing short-term foreign currency liabilities. At the other end of the spectrum is the creation of a global currency and central bank. In between these extremes reside a number of important but largely incremental reforms to the emerging in- stitutional architecture for the provision of foreign currency liquidity assistance (“FCLA”). This architecture includes the central bank swap lines used to provide U.S. dollar liquidity to foreign banks in the thick of the financial crisis, along with the new and still untested emergency lending facilities re- cently introduced by the International Monetary Fund (the “IMF”). In the case of the central bank swap lines, these re- forms include the introduction of more observable and objec- tive conditions for the provision of FCLA: ex ante qualifica- tions for eligible countries, regulatory constraints for eligible financial institutions, and a fee structure that better reflects the role of this liquidity assistance as a form of insurance. In the case of the IMF’s new emergency lending facilities, these reforms include a significant expansion of the available finan- cial resources and the imposition of targeted ex post monitor- ing on countries accessing FCLA. Collectively, these reforms would help constrain the build-up of risk within the Eurodol- lar market and ameliorate the potential moral hazard prob- lems associated with this state-sponsored liquidity support for private markets and institutions. This Article makes two distinct contributions to the rapidly expanding literature examining the sources of financial insta- bility and the theory and practice of international financial crisis management. First, it describes the important and var- ied roles that the modern Eurodollar market plays within the international financial system. While a great deal has already AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 947 been written about the Eurodollar market, the preponderance of the existing literature explores questions around its role as a vehicle for regulatory arbitrage and whether its enormous growth contributes to the aggregate money supply and, thus, inflation. With some notable exceptions, this literature also has a tendency to describe the Eurodollar market at a high level of abstraction, thereby contributing to the considerable degree of confusion around what the Eurodollar market is, how it works, and why it is so important. In sharp contrast, this Article provides a more detailed description of the me- chanics of various components of the Eurodollar market and examines how they can contribute to financial instability. Se- cond, this Article identifies the challenges that a large Euro- dollar market poses in terms of effective international finan- cial crisis management. It then identifies and evaluates the relative merits of various alternatives for how one might go about addressing these challenges with the objective of pro- moting greater global financial stability. This Article is organized into six parts. Part II begins by tracing the origins of the Eurodollar market and documenting its tremendous growth in recent decades. It also describes the basic mechanics of this market and the important roles that it plays in supporting international trade, finance, and invest- ment. Part III describes the functions of conventional ELA fa- cilities. It also describes the unique challenges posed by the existence of a large Eurodollar market and the corresponding need for a credible framework for the provision of FCLA. Part IV chronicles the important role of the Eurodollar market in the thick of the global financial crisis, along with the unprec- edented measures taken by the Federal Reserve and other major central banks to inject U.S. dollar liquidity into the in- ternational financial system. Part V then examines the vari- ous ways that one might seek to strengthen this emerging in- ternational financial crisis management architecture. Part VI concludes. AWREY – FINAL 948 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 II. THE ORIGINS, MECHANICS, AND GROWTH OF THE EURODOLLAR MARKET The Eurodollar market consists of liabilities—promises to pay—issued by banks and other financial institutions that are denominated in the currency of a country other than the one in which the institution is incorporated and licensed to carry on business.26 These liabilities are issued on a short-term ba- sis, with the issuer typically obligated to deliver the relevant currency on demand, upon the occurrence of a specified con- tingency, or at the expiry of a fixed period, ranging from over- night up to one year.27 The term “Eurodollar” is derived from the oldest, largest, and arguably best-known segment of this market: the market for U.S. dollar time deposits offered by banks incorporated and licensed in the United Kingdom and Continental Europe.28 In practice, however, the term has long been used more expansively to describe short-term foreign currency liabilities of any denomination issued by financial in- stitutions incorporated and licensed in any country.29 Like many other financial innovations, the emergence and early growth of the modern Eurodollar market in the 1960s and 1970s was the product of regulatory arbitrage.30 The pri- 26 Alexander K. Swoboda, The Euro-Dollar Market: An Interpretation, in 64 ESSAYS IN INT’L FIN. 1, 1 (1968). 27 The Eurodollar market can be distinguished from the market for longer-term foreign currency liabilities, generally known as the Eurobond market. 28 See generally Ronald I. McKinnon, The Eurocurrency Market, in 125 ESSAYS IN INT’L FIN. 1, 2 (1977); Swoboda, supra note 26, at 1; Friedman, supra note 25, at 17. 29 McKinnon, supra note 28, at 2. 30 See, e.g., PERRY MEHRLING, THE NEW LOMBARD STREET: HOW THE FED BECAME THE DEALER OF LAST RESORT 71–75 (2011) (describing the emer- gence of parallel loans—the precursor to modern swap markets—in re- sponse to capital controls imposed under the Bretton Woods system of fixed exchange rates); Timothy Q. Cook & Jeremy G. Duffield, Money Market Mu- tual Funds: A Reaction to Government Regulation or a Lasting Innovation?, FED. RES. BANK OF RICHMOND ECON. REV., July/Aug. 1979, at 15, 27–29 (de- scribing the emergence of money market funds in response to the con- straints imposed under Federal Reserve Regulation Q). AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 949 mary impetus for this arbitrage was the impact of U.S. Fed- eral Reserve Regulations D and Q.31 Regulation D establishes a minimum fraction of customer deposits that federally li- censed banks are required to hold in the form of non-interest bearing central bank reserves.32 These reserve requirements thus impose a constraint on the amount of deposit funding that banks can use to make loans and other investments. Be- fore being phased out in 1986 (and ultimately repealed in 2011), Regulation Q prohibited banks from paying interest on demand deposits and imposed a ceiling on the interest rates that banks were permitted to pay on time deposits and sav- ings deposits.33 The combined effect of Regulations D and Q was to under- mine the competitiveness of U.S. banks in relation to their in- ternational peers. Where market interest rates on deposits ex- ceeded the ceiling under Regulation Q—as they did continuously between 1966 and 198634—depositors could shift their savings to banks in countries such as the United King- 31 For early work examining this impact, see McKinnon, supra note 28, at 5–10; Friedman, supra note 25, at 17–18, 24; Lawrence L. Kreicher, Eu- rodollar Arbitrage, FED. RES. BANK OF N.Y. Q. REV., Summer 1982, at 10. Secondary drivers of this arbitrage were the direct and indirect capital con- trols imposed under the Bretton Woods monetary and exchange rate frame- work. McKinnon, supra note 28, at 8; Friedman, supra note 25, at 17–18. The desire of market participants in some jurisdictions—perhaps most no- tably the former Soviet Union—to hold U.S. dollars outside the United States in order to avoid potential seizure also likely contributed to early demand for Eurodollar deposits. See PAUL EINZIG, THE EURO-DOLLAR SYSTEM 30 (5th ed. 1973). 32 See Reserve Requirements of Depositary Institutions (Regulation D), 12 C.F.R. § 204 (2012). Today, Regulation D also applies to “euro-currency” liabilities of federally licensed banks. Id. 33 See Payment of Interest on Demand Deposits, 12 U.S.C. § 371(a) (1979) (repealed 2011). See also 76 Fed. Reg. 42015 (July 18, 2011) (repeal- ing Regulation Q effective July 21, 2011). 34 See R. Alton Gilbert, Requiem for Regulation Q: What It Did and Why It Passed Away, FED. RES. BANK OF ST. LOUIS REV., Feb. 1986, at 22, 29. AWREY – FINAL 950 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 dom that adopted a more laissez faire approach toward inter- est rates.35 Similarly, insofar as the Fed’s reserve require- ments were more onerous than those imposed on U.S. dollar deposits in other countries, borrowers in the United States could often lower their cost of capital by borrowing U.S. dol- lars from foreign banks.36 The extension of these loans would then create new Eurodollar deposits.37 Predictably, the com- petitive distortions created by these differences in regulatory treatment precipitated a shift in U.S. dollar deposits and loans from banks in the United States to those in the United Kingdom, Europe, and elsewhere. Given its origins, there is a tendency to view the Eurodol- lar market purely as a creature of regulatory arbitrage. Today, however, banks and other financial institutions issue Euro- dollar liabilities for a variety of different reasons. Many of these reasons reflect the day-to-day needs of commercial firms operating within the global economy. These firms make and receive payments in foreign currencies in the normal course of business. When Apple Inc. purchases camera components from Sony for use in the iPhone 7, for example, it might pay Sony in Japanese yen.38 When Apple then sells the fully as- 35 See McKinnon, supra note 28, at 2, 5–10; Friedman, supra note 25, at 17–18. See also Catherine R. Schenk, The Origins of the Eurodollar Mar- ket in London: 1955–1963, 35 EXPLS. IN ECON. HIST. 221 (1998) 36 See Dong He & Robert N. McCauley, Offshore Markets for the Domes- tic Currency: Monetary and Financial Stability Issues 14 (Bank for Int’l Set- tlements, Working Paper No. 320, 2010) (citing Robert N. McCauley & Rama Seth, Foreign Bank Credit to U.S. Corporations: The Implications of Offshore Loans, 17 FED. RES. BANK OF N.Y. Q. REV., Spring 1992, at 52). 37 While the reason for this may not seem immediately obvious to the casual observer, it is sufficient for the present purposes simply to observe that the extension of bank loans typically involves crediting the borrower’s account with the bank. In this way, new loans serve to create new deposit liabilities. For a more detailed description, see MICHAEL MCLEAY ET AL., BANK OF ENG., QUARTERLY BULLETIN 2014 Q1: MONEY CREATION IN THE MOD- ERN ECONOMY 14 (2014), http://www.bankofengland.co.uk/publications/Doc- uments/quarterlybulletin/2014/qb14q1prereleasemoneycrea- tion.pdf#page=1 [perma.cc/ZU6A-HZ78]; supra Section III.A. 38 Alternatively, of course, Apple can pay Sony in U.S. dollars. Either way, one of the two firms is dealing in a foreign currency. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 951 sembled iPhone at its store in Covent Garden, London, it re- ceives payment in Pound Sterling. As a U.S. firm incorporated and headquartered in California, Apple may then seek to hedge the foreign exchange risk stemming from both its global supply chain and international sales.39 Similarly, the prices of some goods (e.g., crude oil) are only quoted in a single currency (e.g., U.S. dollars), thus necessitating that all prospective pur- chasers hold reserves of, and all sellers accept payments in, this currency. Understandably, firms that make and receive payments in foreign currencies often look to domestic banks and other financial institutions to provide the commercial ac- counts, wholesale funding, and risk management services nec- essary to support their international operations.40 For this reason, economic globalization and a robust Eurodollar mar- ket are often viewed as going hand in hand. A second and related reason why banks and other financial institutions issue Eurodollar liabilities is to support interna- tional trade. While reliable data is scarce, it has been esti- mated that approximately half of all international trade is in- voiced and settled in U.S. dollars.41 Understandably, the dominance of the dollar in international trade incentivizes im- porters and exporters outside the United States to maintain dollar denominated bank accounts in order to accept and make payments in U.S. dollars.42 The majority of these ac- counts are held with banks incorporated and licensed outside 39 As it happens, one of the ways in which Apple in particular hedges the foreign exchange risk stemming from its international sales is to fre- quently update its domestic prices in response to changes in the exchange rate between the domestic currency and the U.S. dollar. However, most firms do not enjoy this degree of market power and thus rely on financial intermediaries to help them manage their foreign exchange risk. 40 Firms may do so for several reasons, ranging from pre-existing rela- tionships, to higher interest rates on domestic deposits, to preferential tax treatment. 41 Benjamin J. Cohen, Currency and State Power, in BACK TO BASICS: STATE POWER IN A CONTEMPORARY WORLD 159, 167 (Martha Finnemore & Judith Goldstein eds., 2013). See Linda S. Goldberg & Cédric Tille, Vehicle Currency Use in International Trade, J. INT’L ECON., Dec. 2008, at 177, 184– 85 (providing a more detailed country-by-country breakdown). 42 See Swoboda, supra note 26, at 6. AWREY – FINAL 952 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 the United States.43 Foreign banks are also an important source of trade financing. One of the most widely used forms of trade financing are letters of credit. Letters of credit repre- sent a commitment by a bank to an exporter (seller) on behalf of an importer (buyer) to make or guarantee payment for goods upon receipt of documentary evidence that they have been delivered. Letters of credit can thus be understood as a form of insurance against the risk of non-payment by the im- porter. These letters of credit possess two important features for the present purposes. First, most letters of credit are is- sued on a relatively short-term basis: with maturities typi- cally ranging from seventy to ninety days.44 Second, and re- flecting the currency composition of international trade more generally, the International Chamber of Commerce estimates that over eighty percent of all letters of credit are denomi- nated in U.S. dollars.45 A third reason why financial institutions issue Eurodollar liabilities is to execute what is known as a foreign currency ‘carry trade.’ Foreign currency carry trades seek to exploit dif- ferences between the costs of borrowing in one currency rela- tive to the expected return on financial assets denominated in another currency. Thus, for example, a German investment bank might borrow cheaply in Japanese yen and then invest the proceeds in higher-yielding Australian dollar-denomi- nated assets. Where successful, the bank stands to capture the difference—or ‘spread’—between its yen-borrowing costs and its Australian dollar investment returns. Simultaneously, of course, the trade exposes the investment bank to exchange rate volatility and, more specifically, the risk that the yen might appreciate against the Australian dollar over the dura- tion of the trade.46 Importantly, the foreign currency borrow- 43 He & McCauley, supra note 36, at 2–4. 44 See COMMITTEE ON THE GLOBAL FIN. SYS. BANK FOR INT’L SETTLE- MENTS, TRADE FINANCE: DEVELOPMENTS AND ISSUES 14 (2014). 45 See INT’L CHAMBER OF COMMERCE, 2016 RETHINKING TRADE & FI- NANCE: AN ICC PRIVATE SECTOR DEVELOPMENT PERSPECTIVE 58 (2016). 46 At this point, the bank will either have to rollover its (now more ex- pensive) yen borrowing, or simply repay its yen-denominated liabilities. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 953 ing necessary to execute this trade often takes the form of for- eign exchange (or “f/x”) swaps, commercial paper, repo agree- ments, and other short-term wholesale liabilities. Finally, banks and other financial institutions create Eu- rodollar liabilities in the context of their general financing and market-making activities. Many investment banks, for exam- ple, borrow funds within f/x swap, foreign currency repo, and other wholesale funding markets in order to raise capital and manage mismatches between the currency composition of their assets and liabilities.47 Investment banks also create and market structured finance products, f/x swaps, and other derivatives to their clients.48 Structured finance products and derivatives are not typically viewed as creating Eurodollar li- abilities. Nevertheless, where these instruments are denomi- nated in a currency not issued by a country in which one or both parties carry on business, they ultimately share a num- ber of important similarities. Most importantly for the present purposes, these instruments typically include state-contin- gent contractual mechanisms such as variation margin, close- out netting, and novation requirements designed to protect the contracting parties against changes in counterparty credit and market risks.49 When these mechanisms are triggered, they require parties to deliver the agreed foreign currency, of- ten within a relatively short time frame.50 These ostensibly long-term contracts thus include embedded short-term foreign currency liabilities. This risk can of course be hedged, but at the expense of the expected profit- ability of the trade. 47 See ARMOUR ET AL., supra note 15, at 452–54. 48 See id. at 460–65, 468–69. 49 See id. at 469–72. 50 See generally Darrell Duffie, The Failure Mechanics of Dealer Banks, 24 J. ECON. PERSP. 51, 55–58 (2010) (describing the vulnerability of dealers to runs by derivatives and repo counterparties and other creditors); Mark J. Roe, The Derivatives Market’s Payment Priorities as Financial Crisis Accel- erator, 63 STAN. L. REV. 539, 542–43 (2011) (describing the runs by deriva- tives and repo counterparties on AIG, Bear Stearns, and Lehman Brothers). AWREY – FINAL 954 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 FIGURE 1: BANKING SECTOR FOREIGN CURRENCY LIABILI- TIES (1980–2015)51 Reflecting their wide range of potential uses in interna- tional trade, finance, and investment, the outstanding stock of Eurodollar liabilities has increased significantly over the past several decades. While limitations on available data have long made measuring the size of the Eurodollar market diffi- cult, figures recently released by the Bank for International Settlements (the “BIS”) shed some light on the enormous growth of the global banking system’s foreign currency liabil- ities between 1980 and 2015 (Figure 1). These figures show foreign currency liabilities peaking at approximately $USD25 trillion on the eve of the global financial crisis, with more than half of this amount made up of liabilities denominated in U.S. dollars.52 To put these figures into perspective, the total stock 51 EDD DENBEE ET AL, BANK OF ENG., FINANCIAL STABILITY PAPER NO. 36: STITCHING TOGETHER THE GLOBAL FINANCIAL SAFETY NET 6 fig.3 (2016), http://www.bankofengland.co.uk/financialstability/Documents/fpc/fspa- pers/fs_paper36.pdf [perma.cc/X4DN-PS4A]. 52 See id. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 955 of domestic U.S. dollar deposits held by commercial banks in- sured by the U.S. Federal Deposit Insurance Corporation (the “FDIC”) was approximately $USD11.7 trillion at the end of 2016.53 Ultimately, of course, these figures only provide us with an impressionistic sense of the size of the global Eurodollar mar- ket. Specifically, insofar as these figures include longer-term foreign currency liabilities of banks (and exclude short-term foreign currency liabilities of non-bank financial intermediar- ies and contingent foreign currency liabilities embedded within structured finance products and other derivatives), these figures overstate (understate) the true size of this mar- ket. Nevertheless, these figures underscore both the im- portant roles that the Eurodollar market plays within the global economic and financial system and its dramatic growth in recent decades. III. FINANCIAL CRISIS MANAGEMENT AND THE PROBLEM OF FCLA The growth and importance of the Eurodollar market poses a number of challenges for central banks charged with the re- sponsibility for financial crisis management. As articulated most famously by Walter Bagehot, the core principles of finan- cial crisis management call upon central banks to lend freely, against good quality collateral, and at a penalty rate of inter- est, to solvent but illiquid banks and other financial institu- tions during periods of widespread panic and instability.54 This prescription, often referred to as “Bagehot’s Dictum,” represents the animating spirit, if not always the strict legal 53 See FED. DEPOSIT INS. CORP., LATEST INDUSTRY TRENDS: STATISTICS AT A GLANCE (2016), https://www.fdic.gov/bank/statistical/stats/2016dec/indus- try.pdf [perma.cc/X4DN-PS4A]. 54 See BAGEHOT, supra note 19, at 21–22. While Bagehot is often cred- ited with having first articulated these principles, that honour largely falls to fellow Englishman Henry Thornton. See generally HENRY THORNTON, AN ENQUIRY INTO THE NATURE AND EFFECTS OF THE PAPER CREDIT OF GREAT BRIT- AIN (F. A. v. Hayek ed., Farrar & Rinehart 1939) (1802). AWREY – FINAL 956 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 letter, underpinning the diverse array of institutional ar- rangements through which central banks provide ‘lender of last resort’ or ELA facilities.55 Historically, the core principles of financial crisis manage- ment have been developed and applied predominantly at the domestic level, with national central banks providing ELA to the domestic banking system, denominated in the domestic currency.56 As a consequence, these principles have not fully evolved to reflect the emergence and growth of the Eurodollar market. This Part describes these principles. It then identifies the significant practical constraints in applying these princi- ples in a world where banks and other financial institutions rely heavily on the issuance of Eurodollar liabilities and where, accordingly, these institutions are likely to require FCLA during periods of widespread panic and instability. A. The Fundamentals of Emergency Liquidity Assistance In order to understand the fundamentals of ELA, one must first understand the peculiar business of banking. The defin- ing feature of banks is that they raise a significant proportion of their capital through the issuance of short-term, liquid lia- bilities. These liabilities include demand, time, and other types of deposits, along with short-term wholesale borrowing from other banks and financial institutions through the com- mercial paper and repo markets.57 Banks then invest this cap- 55 See Xavier Freixas et al., Lender of Last Resort: A Review of the Lit- erature, FIN. STABILITY REV. 151 (1999). 56 This domestic focus is reflected in the academic and policy literature. See generally id. at 158–59, 162. FCLA discussions have often been in ex- tremely vague terms. See Bank for Int’l Settlements, Statement of Central Bankers, N.Y. TIMES, Sept. 11, 1974, at 64 (“The governors also had an ex- change of views on the problem of the lender of last resort in the Euromar- kets. They recognized that it would not be practical to lay down in advance detailed rules and procedures for the provision of temporary liquidity.”). 57 The mix of deposits and wholesale funding as a proportion of total bank financing has notably changed over time. A recent study covering sev- AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 957 ital in longer term, illiquid loans, mortgages, and other finan- cial assets. In the language of financial economics, banks thus engage in significant levels of credit, liquidity, and maturity transformation by taking the short-term, liquid liabilities owed to depositors and converting them into longer-term, il- liquid investments in households, businesses, and govern- ments.58 Importantly, the credit, liquidity, and maturity transfor- mation performed by conventional deposit-taking banks in- volves considerably more than the simple intermediation of savings from depositors to borrowers.59 It also represents an important source of private money creation. When a bank makes a loan, the proceeds of the loan are credited to the bor- rower’s bank account.60 The extension of a new loan can thus be understood as mechanically resulting in the creation of new deposits.61 These new deposits can then be used to purchase goods, pay for services, or make investments. In short, these deposits can be used as money—money that would not have existed but for the extension of the loan. Rather than merely acting as a conduit for the accumulation and allocation of cap- ital, the business of banking thus necessarily involves a sig- nificant level of credit and, ultimately, money creation.62 The important role played by banks in creating money is reflected enteen countries, for example, reports a shift in deposit funding from ap- proximately 80% of total finding following World War II to approximately 50% by the early 2000s. Oscar Jordà et al., Bank Capital Redux: Solvency, Liquidity, and Crisis 11–12 (Fed. Reserve Bank of San Francisco, Working Paper No. 2017–06, 2017). 58 See ARMOUR ET AL., supra note 15, at 277–78 (describing banks as performing credit, maturity, and liquidity transformation). 59 See Hockett & Omarova, supra note 13, at 1148 (delineating between the ‘credit-intermediation,’ ‘credit-multiplication,’ and ‘credit-creation’ mod- els of finance). 60 See generally McLeay et al., supra note 37, at 3–5. 61 See id. at 1. 62 See id. at 2–4. AWREY – FINAL 958 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 in the value of outstanding bank deposits relative to other components of the aggregate money supply (Figure 2).63 FIGURE 2: COMPONENTS OF THE U.S. MONEY SUPPLY ($USD AS OF MARCH 2017) Component of Money Supply Amount (in billions) Percentage of Total Currency in circulation $1,513,317 9.8% Reserves64 $2,314,729 15.0% Commercial bank deposits65 $11,594,200 75.2% TOTAL $15,422,246 100% The business of banking works well so long as large num- bers of depositors or other short-term creditors do not demand their money back at the same time.66 Where liquidity de- mands remain uncorrelated, banks need only maintain a small reserve of cash or other highly liquid assets to satisfy 63 See FED. RESERVE SYS., H.3 AGGREGATE RESERVES OF DEPOSITARY IN- STITUTIONS AND THE MONETARY BASE (2017), https://www.federalre- serve.gov/releases/h3/current/h3.pdf [perma.cc/VNB8-VV5Z]; FED. RESERVE SYS., H.8 ASSETS AND LIABILITIES OF COMMERCIAL BANKS IN THE UNITED STATES (2017), https://www.federalreserve.gov/releases/h8/current/h8.pdf [perma.cc/LJ47-XEDW]. The Federal Reserve also publishes weekly “Money Stock Measures” (H.6) that report slightly different figures. 64 Includes total reserves maintained with the Federal Reserve System and vault cash used to satisfy required reserves. FED. RESERVE SYS., H.3 AGGREGATE RESERVES OF DEPOSITARY INSTITUTIONS AND THE MONETARY BASE (2017), https://www.federalreserve.gov/releases/h3/current/h3.pdf [perma.cc/96VC-TQQS]. 65 The Federal Reserve’s Money Stock Measures report a slightly smaller $USD11,289,500 billion in demand, time, and other checkable de- posits at commercial banks and thrifts as of February 2017. Id. 66 See Douglas W. Diamond & Philip H. Dybvig, Bank Runs, Deposit Insurance, and Liquidity, 91 J. POL’Y ECON. 401, 402 (1983) (describing how the demand deposit system permits efficient risk sharing where depositors demand liquidity at random times). See generally Franklin Allen et al., Moral Hazard and Government Guarantees in the Banking Industry, 1 J. FIN. REG. 30 (2015) (providing a recent survey of the literature on the fra- gility of banks). AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 959 idiosyncratic withdrawals.67 When combined with the fact that new loans create new deposits, this ‘fractional reserve’ model enables banks to extend a large volume of loans on the foundations of a relatively small deposit base.68 By the same token, however, where the liquidity demands of short-term creditors become correlated, the reliance by banks on short- term, liquid liabilities to finance investments in longer term, illiquid assets renders these institutions vulnerable to desta- bilizing runs by depositors and other short-term creditors.69 The fragility of bank balance sheets is typically framed in one of two ways. The first views runs as a multiple equilib- rium (or coordination) problem amongst short-term credi- tors.70 Pursuant to this account, each creditor’s decision about whether to run is a function not only of their demand for li- quidity and evaluation of the issuer’s creditworthiness, but also—and crucially—their subjective perception of whether other creditors are likely run.71 The second account views runs as a product of the realization by short-term creditors that as- sets that they previously believed to represent reliable stores of nominal value—or “moneyness”72—are in fact sensitive to the revelation of new information about the creditworthiness of the issuer, the quality of any underlying collateral, or other 67 The term “withdrawals” in this context includes the failure to rollo- ver short-term wholesale funding. 68 An example may help illustrate this point. Imagine a bank extends a loan of $100 to a borrower. This loan results in $100 of new deposits. As- suming a 10% reserve ratio, the bank is able to extend an additional $90 of loans on the basis of these new deposits. These new loans will then result in $90 of new deposits, enabling the bank to extend $USD81 in new loans. This process could theoretically continue ad infinitum. 69 See Diamond & Dybvig, supra note 66, at 402. 70 See id. at 402–03, 417–18. 71 See id. 72 See MILTON FRIEDMAN & ANNA JACOBSON SCHWARTZ, MONETARY STA- TISTICS OF THE UNITED STATES: ESTIMATES, SOURCES, METHODS, 151–52 (1970); J. R. HICKS, VALUE AND CAPITAL, 163 (2nd ed., Clarendon Press 1946) (1939). AWREY – FINAL 960 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 variables.73 Pursuant to this second account, rather than in- vesting the time and energy necessary to incorporate this new information into the price of these assets, creditors may simply prefer to shift to less informationally-sensitive substi- tutes that, in effect, possess a higher degree of moneyness.74 Where the liquidity demands of depositors and other short- term creditors exceed available reserves, banks are left with two possible options. The first is to seek funding from other financial institutions within wholesale funding markets. The availability of this option will depend on market perceptions of the borrower’s solvency,75 along with whether or not pro- spective lenders within this market are themselves facing—or anticipate facing—similar liquidity problems.76 The second option is to sell loans or other financial assets and to use the proceeds to satisfy the bank’s obligations to short-term credi- tors. Crucially, however, where banks are forced to sell longer- term, opaque, and illiquid assets within a short time frame, this will reduce the probability that these institutions will be able to sell these assets at full market value. As a result, banks may be forced to accept losses on long-term assets in order to honor their liabilities to short-term creditors. 73 See generally GARY B. GORTON, SLAPPED BY THE INVISIBLE HAND: THE PANIC OF 2007 (2010); Bengt Holmstrom, Understanding the Role of Debt in the Financial System (Bank of Int’l Settlements, Working Paper No. 479, Jan. 2015), https://www.bis.org/publ/work479.pdf [perma.cc/H44L-C4PV]. The key difference between this account and multiple equilibrium views of run-like behavior is the absence of a coordination problem, i.e., creditors in this second account do not switch in anticipation of the decisions of other creditors. 74 See generally GORTON, supra note 73; Holmstrom, supra note 73. 75 See Freixas et al., supra note 55, at 153. 76 See Mark J. Flannery, Financial Crises, Payment System Problems, and Discount Window Lending, 28 J. MONEY, CREDIT & BANKING 804, 805– 06 (1996) (suggesting that banks may become more reluctant to lend in the wholesale market during periods of crisis); Xavier Freixas et al., Systemic Risk, Interbank Relations, and Liquidity Provision by the Central Bank, 32 J. OF MONEY, CREDIT & BANKING 611, 611, 613 (2000) (presenting a model in which banks are reluctant to lend in the wholesale market owing to uncer- tainty about their own ability to borrow in the market in the future). AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 961 Through this channel, what began as a short-term liquidity problem can quickly evolve into a full-blown solvency problem. However one elects to frame the source of this fragility, the destabilizing run-like behavior at the heart of bank runs can escalate into more generalized panics characterized by the widespread withdrawal of financing by short-term creditors and a resulting “scramble for high-powered money.”77 These generalized panics can be understood as throwing the frac- tional reserve model into reverse: with the flight of short-term creditors reducing the deposit base and precipitating fire sales of loans and other assets, and the sale of loans mechanically resulting in further reductions in the deposit base.78 Morgan Ricks has characterized this relationship among runs, finan- cial panics, and private money creation as “the money prob- lem”79 and views it as “far and away the biggest threat the financial system poses to the broader economy.”80 At the core of this problem is the contraction in the supply of privately created money during financial panics and the re- sulting impact on the real economy.81 These contractions force financial institutions that rely on short-term funding to sell 77 Anna J. Schwartz, Real and Pseudo-Financial Crises, in MONEY IN HISTORICAL PERSPECTIVE 271 (1987). 78 Although at a systemic level, the reduction in deposits will only serve to reduce the aggregate money supply where the withdrawn funds flow out of the conventional banking system. See George G. Kaufman, Lender of Last Resort: A Contemporary Perspective, in FINANCIAL CRISES, CONTAGION, AND THE LENDER OF LAST RESORT 169, 174 (Charles Goodhart & Gerhard Illing eds., 2002). 79 MORGAN RICKS, THE MONEY PROBLEM: RETHINKING FINANCIAL REGU- LATION (2016). 80 Id. at 103. 81 In this respect, Ricks draws on Friedman and Schwartz’s analysis of the monetary causes of the Great Depression. See generally MILTON FRIED- MAN & ANNA JACOBSON SCHWARTZ, A MONETARY HISTORY OF THE UNITED STATES, 1867–1960 (1971). Ricks also relies on the empirical evidence of Vic- toria Ivashina and David Scharfstein relating to the financial crisis of 2007– 09. See Victoria Ivashina & David Scharfstein, Bank Lending During the Financial Crisis of 2008, 97 J. FIN. ECON. 319 (2010). AWREY – FINAL 962 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 assets in order to cover their short-term liabilities.82 The re- sulting fire sales lead to a fall in asset prices and a correspond- ing increase in the yields on these assets. The increase in yields then changes the opportunity cost calculus for financial institutions looking to make new investments—precipitating an increase in borrowing costs for households, businesses, and governments, and a corresponding decrease in levels of invest- ment and economic activity.83 Viewed from this perspective, panics drive contractions in the money supply, which in turn threaten to drive contractions in economic growth. It is against the backdrop of this threat that the functions of state support for the conventional banking system can be most clearly understood.84 The first—microprudential—func- tion is to reduce the risk of idiosyncratic bank runs.85 The se- cond—macroprudential—function is to prevent bank runs and other shocks from escalating into more widespread panics and, ultimately, contractions in the money supply and eco- nomic growth.86 While early writers such as Bagehot and Henry Thornton arguably viewed ELA as performing solely a macroprudential function, more recent observers have empha- sized the largely complementary relationship between these two functions insofar as the failure to support individual in- stitutions can, in some cases, precipitate more widespread panic and financial instability.87 82 RICKS, supra note 79, at 110. 83 See id. at 110–11. 84 Although, as Thomas M. Humphrey and Robert E. Keleher have ob- served, these functions are rarely spelled out, leading to significantly diver- gent understandings in practice. Thomas M. Humphrey & Robert E. Kele- her, The Lender of Last Resort: A Historical Perspective, in FINANCIAL CRISES, CONTAGION, AND THE LENDER OF LAST RESORT 73 (2002). 85 See Freixas et al., supra note 55. 86 See Freixas et al., supra note 55, at 153; Humphrey, supra note 21, at 8-9; Humphrey & Keleher, supra note 84, at 74–75; Kaufman, supra note 78, at 173–74. 87 In effect, this second perspective reflects the idea that some financial institutions are “too big to fail” in the sense that their failure is expected to trigger more widespread panic and market disruption. See Freixas et al., supra note 55, at 152; Kaufman, supra note 78, at 169. See generally Marvin Goodfriend & Robert G. King, Financial Deregulation, Monetary Policy, and AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 963 There are two basic ways that the state can perform these functions. The first is deposit insurance. While deposit insur- ance schemes vary across jurisdictions, the basic strategy is to have a third-party guarantee depositors’ funds in the event of bank failure. This third party effectively steps into the shoes of the bank: honoring the bank’s commitments to depositors during periods of institutional distress.88 In order to make this commitment credible, the government typically provides this third-party guarantee through institutions such as the FDIC. The introduction of deposit insurance beginning in the United States in the 1930’s is often credited with having significantly enhanced the level of consumer protection enjoyed by bank de- positors.89 However, insofar as these schemes generally only protect a subset of a bank’s short-term creditors—typically re- tail depositors with accounts below a specified threshold—it is at best an incomplete strategy for counterbalancing the inher- ent fragility of bank balance sheets. The second, far older, and far broader strategy is ELA. Dis- tilled to its essence, ELA is “the discretionary provision of li- quidity to a financial institution (or the market as a whole) by the central bank in reaction to an adverse shock which causes an abnormal increase in demand for liquidity which cannot be met by an alternative source.”90 In effect, ELA is designed to pump money into the financial system, thereby offsetting any contraction in the private money supply and signaling to de- positors and other short-term creditors that the central bank will provide whatever support is necessary in order to restore market confidence and financial stability. This liquidity typi- Central Banking, in RESTRUCTURING BANKING & FINANCIAL SERVICES IN AMERICA 216 (William S. Haraf & Rose Marie Kushmeider eds., 1988). 88 The third party then steps into the shoes of the depositors in order to seek redress from the estate of the bank. 89 See, e.g., Andrew Campbell & Peter Cartwright, Deposit Insurance: Consumer Protection, Bank Safety and Moral Hazard, 10 EUR. BUS. L. REV. 96, 96–97, 99 (1999) (describing the competing views of deposit insurance as a mechanism for either protecting consumers or minimizing the risk of bank runs). 90 Freixas et al., supra note 55, at 152. AWREY – FINAL 964 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 cally takes the form of short-term loans to banks or other fi- nancial institutions experiencing temporary liquidity prob- lems. These loans are fully secured by a pledge of eligible col- lateral assets by the financial institutions receiving ELA. Eligible collateral assets generally include commercial loans, residential and commercial mortgages, government securi- ties, investment grade bonds, asset-backed securities, and other fixed income instruments. These collateral assets are then subject to a discount—or “haircut”—designed to protect the central bank against any decrease in their market value.91 The institutional frameworks through which central banks provide ELA vary from jurisdiction to jurisdiction. In the United States, the Federal Reserve has historically provided ELA to eligible member banks through its discount window, a standing collateralized loan facility.92 Pursuant to section 13(3) of the Federal Reserve Act, the Fed can also extend ELA to non-bank financial institutions under certain conditions in the event of “unusual and exigent circumstances.”93 In the 91 This discount takes the form of a decrease in the amount of the loan that can be secured against any given collateral asset. 92 For the framework implementing the Federal Reserve’s discount window, see generally Extensions of Credit by Federal Reserve Banks (Reg- ulation A), 12 C.F.R. § 201 (2017). See also FED. RES. DISCOUNT WINDOW, https://www.frbdiscountwindow.org/ (Oct. 15, 2017) [perma.cc/K2AR- GRM9]. The Fed can also extend ELA through its open market operations (i.e., the purchase and sale of securities on the open market). Kaufman, su- pra note 78, at 180–81. For a discussion of the conceptual problems of con- flating ELA with open market operations, see generally Charles Goodhart, Myths About the Lender of Last Resort, in FINANCIAL CRISES, CONTAGION, AND THE LENDER OF LAST RESORT 227, 231 (Charles Goodhart & Gerhard Illing eds., 2002). 93 Federal Reserve Act, 12 U.S.C. § 343 (2010). The provision of ELA under section 13(3) must be pursuant to a program or facility with “broad- based” eligibility, thereby prohibiting ad hoc support to individual institu- tions. Unlike discount window lending, institutions in receipt of ELA under section 13(3) must also demonstrate that they cannot secure credit accom- modation from other financial institutions. As Kathryn Judge observes, there are also a number of other government facilities that can be viewed as performing a similar function. Kathryn Judge, Three Discount Windows, 99 CORNELL L. REV. 795 (2014) (describing, for example, the operations of AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 965 United Kingdom, meanwhile, the Bank of England operates its discount window, ELA, and other liquidity facilities under separate frameworks, each of which are available to banks, building societies, securities dealers, and central counterpar- ties.94 Predictably, the strategy of having central banks pump money into the financial system—to “lend freely” in Bagehot’s terms—gives rise to potentially significant moral hazard prob- lems.95 Specifically, where banks and other financial institu- tions anticipate that they will receive ELA, this may incentiv- ize them to rely more heavily on short-term liabilities or invest in more risky and illiquid assets. The other components of Bagehot’s dictum can each be understood as designed to con- strain this problem.96 The requirement that central banks lend only against good quality collateral, for example, limits the ability of banks to invest in risky and illiquid assets while still retaining a sufficient portfolio of eligible collateral that they can pledge for the purposes of obtaining ELA.97 For the same reason, the fact that a bank is able to pledge good quality the Federal Home Loan Bank in providing support to banks during the fi- nancial crisis). 94 For a detailed description of the discount window and contingent term repo facilities, see generally BANK OF ENG., THE BANK OF ENGLAND’S STERLING MONETARY FRAMEWORK (2015). ELA is provided under a separate, largely unpublished, framework and can be provided to any entity (includ- ing but not limited to financial institutions). The Bank also administers an index long-term repo facility designed for institutions with predictable needs for liquid assets. 95 Once again, Henry Thornton seems to have been the first observer to remark on this problem. THORNTON, supra note 54, at 188. 96 Ultimately, the capital, liquidity, and other forms of prudential reg- ulation and intensive prudential supervision to which banks are subject also play important roles in constraining potential moral hazard problems. For a description of this regulation and supervision, see ARMOUR ET AL., supra note 15, at chs. 14–16. 97 This requirement also serves to insulate central banks from counter- party credit risk, along with any market risk stemming from volatility in the price of pledged collateral. AWREY – FINAL 966 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 collateral serves as a credible signal of its solvency.98 The re- quirement to lend at a penalty rate of interest, meanwhile, deters banks from relying on ELA where private market- based sources of short-term financing are available.99 Finally, as banks and other financial institutions approach insolvency, they have a natural incentive to take on more risk. The re- quirement that ELA be extended only to financial institutions experiencing temporary liquidity problems thus ensures that fundamentally insolvent institutions do not use it in an at- tempt to gamble for resurrection. Ultimately, while it is de- batable whether Bagehot’s dictum is always observed in prac- tice100—and while it may be a challenge to calibrate penalty interest rates, define what constitutes good quality collateral, and distinguish between illiquid and insolvent institu- tions101—the components of this dictum coalesce to provide a relatively coherent framework for understanding when and how central banks should provide ELA to banks and other fi- nancial institutions. Over the past decade, the effectiveness of conventional ELA frameworks in many jurisdictions has been tested by the emergence and growth of the so-called “shadow” banking sys- tem102: non-bank financial intermediaries that perform credit, liquidity, and maturity transformation without explicit access 98 Stanley Fischer, On the Need for an International Lender of Last Re- sort, 13 J. ECON. PERSP. 85, 90 (1999). 99 It also gives banks receiving ELA a strong incentive to pay back the loans as soon as possible. 100 Michael D. Bordo, for example, points out several instances where it appears Bagehot’s dictum was not strictly observed, including the bailouts of Franklin National (1974), First Pennsylvania (1980), Continental Illinois (1984), and Johnson Matthey Bankers (1982). Michael D. Bordo, The Lender of Last Resort: Alternative Views and Historical Experience, in FINANCIAL CRISES, CONTAGION, AND THE LENDER OF LAST RESORT 109, 122 (Charles Goodhart & Gerhard Illing eds., 2002). 101 For a discussion of these challenges, see Goodhart, supra note 92, at 229–34; Kaufman, supra note 78, at 178–81. 102 RICKS, supra note 79, at 160–63, 200–19. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 967 to ELA facilities or other forms of state support.103 During the global financial crisis of 2007–09, for example, the Federal Re- serve extended a relatively small proportion of its emergency lending through its discount window.104 Instead, the Fed was forced to introduce several new emergency lending facilities designed to provide liquidity support to securities dealers, money market mutual funds, and the tri-party repo, asset- backed security, and asset-backed commercial paper mar- kets.105 It also introduced bespoke facilities designed to pro- vide support to systemically important institutions such as Bear Stearns, AIG, Citigroup, and Bank of America.106 Im- portantly, conventional ELA frameworks were also exposed by the crisis as not fit for purpose in a globally integrated and interconnected financial system in which many banks and other financial institutions required access to FCLA during periods of widespread panic and financial instability. 103 See Zoltan Pozsar et al., Shadow Banking, FED. RES. BANK N.Y. ECON. POL’Y REV., Dec. 2013, at 1 (defining shadow banks as non-bank fi- nancial intermediaries that perform credit, liquidity, and maturity trans- formation outside the regulated banking system). 104 See Stephen G. Cecchetti, Crisis and Responses: The Federal Re- serve in the Early Stages of the Financial Crisis, 23 J. ECON. PERSP. 51, 55 (2009) (reporting average borrowing through the discount window of ap- proximately one billion dollars during the first seven months of the financial crisis). 105 These facilities included the Term Securities Lending Facility and Primary Dealer Credit Facility (designed to support securities dealers); the Asset-Backed Commercial Paper Money Market Mutual Fund Facility, Commercial Paper Funding Facility, and Money Market Investor Funding Facility (designed to support money market mutual funds and the asset- backed commercial paper market); and the Term Asset-Backed Securities Loan Facility (designed to support the asset-backed security market). See also Judge, supra note 93 (describing the operations of the Federal Home Loan Bank in providing support to banks during the financial crisis). 106 For an overview of these facilities, see Dietrich Domanski et al., Central Banks as Lender of Last Resort: Experiences During the 2007–2010 Crisis and Lessons for the Future 6-9 (Fed. Reserve Bd., Washington, DC, Divs. of Research & Statistics and Monetary Affairs, Fin. and Econ. Discus- sion Series No. 2014-110, 2014). AWREY – FINAL 968 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 B. The Problem of FCLA The core principles of financial crisis management envision that the central banks responsible for providing ELA will pos- sess several key attributes. The first and most important at- tribute is the capacity to issue an unlimited stock of liabilities that are universally accepted as a form of money within the domestic economy.107 The defining feature of the Federal Re- serve is that it can print U.S. dollars, the Bank of England that it can print Pound Sterling. These central bank liabilities are implicitly backed by the government’s authority to raise and collect taxes, and to dictate that these taxes be paid in the domestic currency. Second, central banks must have the legal authority to use these central bank liabilities to extend loans or purchase assets denominated in the domestic currency for the purpose of providing ELA. Third, central banks must have operational frameworks in place that enable them to extend ELA extremely rapidly—typically within a matter of days, if not hours, from the moment at which it has been determined that ELA is necessary in order to maintain or restore financial stability. These key attributes highlight the fact that conventional ELA frameworks are designed to provide liquidity support to banks and other financial institutions that raise capital and make investments denominated in the domestic currency. But what about financial institutions that raise capital and make investments in foreign currencies? In short: what about the Eurodollar market? Where financial institutions that rely heavily on the Eurodollar market as a source of financing also perform significant levels of credit, liquidity, and maturity transformation, these institutions will be susceptible to desta- bilizing runs by depositors and other short-term creditors. As we have seen, these runs can escalate into broader contrac- tions in the money supply within the domestic financial sys- tem. The key difference in this case is that these runs will take 107 Anna J. Schwartz, Earmarks of a Lender of Last Resort, in FINAN- CIAL CRISES, CONTAGION, AND THE LENDER OF LAST RESORT 449, 454–55 (Charles Goodhart & Gerhard Illing eds., 2002). AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 969 place in foreign currency.108 This fairly benign observation ex- poses the fundamental problem of FCLA: central banks only have the capacity to print money denominated in their own currency. The Bank of England cannot print new U.S. dollars any more than the Federal Reserve can print Pound Ster- ling.109 Nor can these central banks force foreign currency creditors to accept payment in the domestic currency.110 Ac- cordingly, where financial institutions rely extensively on short-term foreign currency financing, central banks will lack the ability to create the critical resource needed to provide ELA to the domestic financial system. There are several ways that central banks can theoreti- cally secure access to foreign currency for the purpose of providing FCLA. The first option is to accumulate foreign cur- rency reserves in denominations and amounts that reflect the composition of the domestic financial system’s short-term for- eign currency liabilities. This option is essentially a form of self-insurance against the prospect of eventually having to provide FCLA. However, while this option may seem rela- tively straightforward, it is also potentially costly and distor- tive. As a preliminary matter, central banks pursuing this op- tion will be exposed to fluctuations in the exchange rate between the domestic and foreign currencies.111 In order to limit the effects of accumulating foreign reserves on the do- mestic money supply, central banks may also need to under- take so-called “sterilization” measures: for example, selling government securities in order to neutralize the monetary im- pact of using domestic currency to purchase the required for- eign currency.112 These sterilization measures will be costly 108 See Roberto Chang & Andrés Velasco, The Asian Liquidity Crisis (Nat’l Bureau of Econ. Research, Working Paper No. 6796, 1998) (describing this problem in the context of the Asian liquidity crisis of 1997–98). 109 See Goodhart, supra note 92, at 235. 110 At least where this is not explicitly contemplated by the relevant contracts. See id. 111 Denbee et al., supra note 51, at 9. 112 Id. The monetary effects of accumulating foreign exchange reserves stem from the fact that these reserves need to be purchased with domestic AWREY – FINAL 970 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 where the yields on the government securities exceed the yield on the central bank’s foreign reserve holdings.113 More broadly, where a significant number of central banks simulta- neously pursue the option of accumulating foreign currency reserves, the resulting increase in demand for reserve curren- cies such as the U.S. dollar may compress yields, thereby dis- torting prices and potentially driving a search for yield among investors.114 The second option is to negotiate an ex ante contractual right to borrow foreign currency in the event that it is required in order to provide FCLA. The central bank of Argentina, for example, has reportedly entered a series of repo agreements with a syndicate of international banks enabling it to swap government securities and other collateral in exchange for U.S. dollars.115 Once again, however, this is likely to be an extremely costly option as the borrower pays a fee for the op- tion itself, along with interest on any borrowed funds.116 The borrower would also be exposed to the risk that the U.S. dollar might appreciate over the term of the agreement. More im- portantly, this option would only be as effective as the lender’s ability to acquire sufficient foreign currency to meet its obli- gations under the type of distressed market conditions in which the borrowing central bank would be required to pro- vide FCLA. currency, thereby increasing the aggregate money supply. Selling govern- ment securities in exchange for domestic currency can thus be used to offset this increase. 113 Id. 114 Id. 115 Schwartz, supra note 107, at 457. See also Carolina Millan & Katia Porzecanski, Argentina Agrees to Borrow $5 Billion from Wall Street Banks, BLOOMBERG (Jan. 29, 2016, 3:45 PM), https://www.bloomberg.com/news/ar- ticles/2016-01-29/argentina-agrees-to-5-billion-of-loans-from-wall-street- banks. 116 Schwartz reports that, under one such agreement entered into in 1996, the Central Bank of Argentina was required to pay a thirty-three ba- sis point (0.33%) insurance premium plus an interest rate on borrowed funds of LIBOR, 205 basis points (2.05%). Schwartz, supra note 107, at 457. Borrowing costs will, of course, vary with both the currency in question and the creditworthiness of the borrower. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 971 The third option is to purchase foreign currency on the open market at the point in time at which it is actually needed to provide FCLA. This option may be expedient and cost effec- tive where central banks require foreign currency in order to provide FCLA in response to idiosyncratic liquidity problems or other localized shocks. However, where the catalyst for the provision of FCLA is a more systemic shock—especially one characterized by the breakdown of international wholesale funding and foreign currency markets—central banks may be left without a reliable private source of foreign currency li- quidity. Under these circumstances, central banks that have neglected to secure some form of ex ante (self-)insurance will have no other option but to seek liquidity support from the central bank that prints the relevant currency. Central banks that pursue this option will thus find themselves in the vul- nerable position of having to ask foreign policymakers for a garden hose whenever their own house catches fire. As we shall see in Part IV, this is precisely the position in which many central banks found themselves in the thick of the global financial crisis. The extension of foreign currency loans from one central bank to another for purposes of providing FCLA poses a host of technical challenges. How can the central bank lending the relevant currency ensure that the borrowing central bank only provides FCLA against good quality collateral, at a penalty rate of interest, and to solvent but illiquid banks and other financial institutions? What forms of collateral should the lender itself take against the obligations of the borrower? How should it calculate any ex ante insurance premium, along with the rate of interest on borrowed funds? Should there be limits on the amount that central banks can borrow? And perhaps most importantly, what mechanisms can be used to make the commitments of both the lender and borrower as credible as possible? The answers to these questions will understandably have an impact on both the effectiveness of FCLA and the ex- tent to which it gives rise to potential moral hazard problems. Until relatively recently, the challenges of providing FCLA were largely theoretical. Where FCLA was provided, it was extended almost exclusively on an ad hoc basis in response to AWREY – FINAL 972 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 localized foreign currency liquidity shocks.117 Over the course of the past several decades, however, the emergence and growth of the Eurodollar market has dramatically increased the risk of more systemic foreign currency liquidity problems and, with them, the need for an effective framework for the provision of FCLA. Just how urgently this framework was re- quired would be exposed by the tumultuous events of 2007– 2009. IV. THE EURODOLLAR MARKET AND THE FINANCIAL CRISIS The global financial crisis of 2007–2009 vividly illustrated the potential systemic risks arising from the existence of a large Eurodollar market.118 Between 2000 and 2007, the boom in U.S. real estate markets, the demand for ‘safe’ assets such as U.S. treasury securities, and the globalization of banking and institutional asset management combined with other fac- tors to spur demand for U.S. dollar-denominated financial as- sets by foreign banks and other financial institutions.119 The BIS estimates that the outstanding stock of foreign claims held by banks grew by roughly three times during this period: from approximately $USD11 trillion at the end of 2000 to 117 For example, the $USD6 billion swap line extended by the United States to Mexico in March 1994 in order to stabilize the peso following the assassination of presidential candidate Luis Donaldo Colosio. See MICHAEL D. BORDO ET AL., STRAINED RELATIONS: U.S. FOREIGN EXCHANGE OPERATIONS AND MONETARY POLICY IN THE TWENTIETH CENTURY 327 (2015). 118 For one of the first legal accounts of this aspect of the financial crisis and the response of the Federal Reserve, see Colleen Baker, The Federal Reserve’s Use of International Swap Lines, 55 ARIZ. L. REV. 603 (2013). 119 See Patrick McGuire & Goetz von Peter, The US Dollar Shortage in Global Banking, BIS Q. REV., Mar. 2009, at 47; Ben S. Bernanke et al., In- ternational Capital Flows and the Returns to Safe Assets in the United States, 2003–2007 (Bd. of Governors of the Fed. Res., Int’l Fin Discussion Papers, No. 1014, 2011); Naohiko Baba et al., US Dollar Money Market Funds and Non-US Banks, BIS Q. REV., Mar. 2009, at 65. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 973 $USD31 trillion by the summer of 2007.120 A significant pro- portion of these claims were denominated in foreign curren- cies.121 Many of the foreign banks and other financial institutions that purchased U.S. dollar-denominated assets financed them using one or a combination of U.S. dollar deposits, overnight repo, and short-term commercial paper.122 Many institutions also relied on f/x swap markets to convert their domestic cur- rency holdings into U.S. dollars.123 As a result, the longer term U.S. dollar-denominated assets held by foreign banks and other financial institutions were often supported by very short-term liabilities that, in effect, required these institu- tions to rollover their U.S. dollar funding on a more or less continuous basis.124 This funding model gave rise to an acute form of funding (or ‘rollover’) risk: the risk that the amount of new short-term dollar funding that these institutions would be able to raise in any given period would be insufficient to meet its obligations to holders of maturing short-term dollar liabilities.125 Most importantly, this funding model left these institutions highly vulnerable to the breakdown of repo, com- mercial paper, and f/x swap markets. In August 2007, what began as a localized disruption within the U.S. subprime mortgage market metastasized into a full-blown crisis within international wholesale funding markets. As described by Gary Gorton and Andrew Metrick, 120 McGuire & von Peter, supra note 119, at 50. 121 Id. 122 Id. 123 See WILLIAM A. ALLEN, INTERNATIONAL LIQUIDITY AND THE FINANCIAL CRISIS 33 (2013). This funding model involved selling the domestic currency spot (for dollars) and buying it forward (with dollars). This, in turn, required the institution to have access to the U.S. dollars necessary to repurchase the domestic currency. As the crisis unfolded, the withdrawal of liquidity from wholesale funding markets generated significant upward pressure on prices with f/x swap markets. Id. at 50–51. 124 McGuire & von Peter, supra note 119, at 48, 53. 125 Id. AWREY – FINAL 974 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 concerns about counterparty credit risk led to a dramatic in- crease in the costs of repo financing.126 These concerns also precipitated the withdrawal of liquidity from both the com- mercial paper127 and f/x swap markets.128 As this liquidity evaporated, financial institutions in the United States began to accumulate dollars for the purposes of meeting their domes- tic liabilities, resulting in a corresponding decrease in the availability of U.S. dollar funding for financial institutions in other jurisdictions.129 As a result, financial institutions out- side the United States found it increasingly expensive—and in some cases impossible—to roll over their short-term U.S dollar liabilities.130 The BIS estimates that this U.S. dollar funding gap for European banks was somewhere between $USD1.1 and 1.3 trillion as of the summer of 2007.131 Signifi- cant funding gaps also emerged for Japanese yen in the United Kingdom, for euros in the United States, and for Swiss francs in the euro area.132 Financial institutions seeking to plug this U.S. dollar fund- ing gap were confronted with a limited range of options. The first option was to sell U.S. dollar-denominated assets. Gener- ally speaking, the assets available for sale included U.S. treas- ury securities, as well as dollar-denominated corporate bonds, 126 See Gary Gorton & Andrew Metrick, Securitized Banking and the Run on Repo, 104 J. FIN ECON. 425 (2012). 127 See Marcin Kacperczyk & Philipp Schnabl, When Safe Proved Risky: Commercial Paper During the Financial Crisis of 2007–2009, 24 J. ECON. PERSP. 29 (2010). 128 See Naohiko Baba et al., The Spillover of Money Market Turbulence to FX Swap and Cross-Currency Swap Markets, BIS Q. REV., Mar. 2008, at 73. 129 Rosalind Z. Wiggins & Andrew Metrick, The Federal Reserve’s Fi- nancial Crisis Response C: Providing US Dollars to Foreign Central Banks (Yale Program on Fin. Stability Case Study 2015-1C-V1, 2016); Brun- nermeier, supra note 24, at 95; ALLEN, supra note 123, at 32. 130 Wiggins & Metrick, supra note 129, at 8. Notably, the domestic (for- eign) liabilities of U.S. (foreign) financial institutions included their contin- gent liabilities to provide backstop liquidity to sponsored structured finance vehicles. See ARMOUR ET AL., supra note 15, ch. 21. 131 McGuire & von Peter, supra note 119, at 48. 132 ALLEN, supra note 123, at 38. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 975 asset-backed securities, structured finance products, and other privately-issued debt securities.133 While the market for U.S. treasury securities is the most liquid in the world, the tightening of credit conditions within repo markets during the crisis made these securities amongst the most sought-after collateral assets.134 In many cases, financial institutions would have therefore preferred to use these government secu- rities to obtain U.S. dollar repo financing rather than sell them outright.135 This left privately-issued debt securities as the most likely candidates for sale. The problem here was that prevailing market conditions during the crisis meant that fi- nancial institutions would have often been forced to sell these assets into illiquid and volatile markets, thus exposing them- selves to potentially significant losses. Viewed from this per- spective, correlated selling of dollar-denominated assets by fi- nancial institutions seeking to plug the U.S. dollar funding gap would have contributed to the pernicious fire sale dynam- ics observed during the financial crisis. The second option was for these financial institutions to reduce their U.S. dollar lending activities.136 For many finan- cial institutions outside the United States, this involved cut- ting back on their trade financing and related currency hedg- ing businesses. Data from the Organization for Economic Cooperation and Development suggests that the value of global merchandise trade decreased by forty-two percent be- tween July 2008 and February 2009.137 While a considerable 133 McGuire & von Peter, supra note 119, at 58. 134 See Peter Hördahl & Michael R. King, Developments in Repo Mar- kets During the Financial Turmoil, BIS Q. REV., Dec. 2008, at 37, 46. 135 Id. 136 Brunnermeier, supra note 24, at 95. For empirical evidence relating to the reduction in dollar lending by European banks during the financial crisis, see Victoria Ivashina et al., Dollar Funding and the Lending Behavior of Global Banks (Nat’l Bureau of Econ. Research, Working Paper No. 18528, 2012). 137 Robert C. Shelburne, The Global Financial Crisis and its Impact on Trade: The World and the Emerging European Economies 2 (U.N. Econ. Comm. for Eur., Discussion Paper No. 2010.2, 2010). AWREY – FINAL 976 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 proportion of this decrease is no doubt attributable to a reduc- tion in demand and production during the crisis, evidence sug- gests that the U.S. dollar funding gap also contributed to a reduction in credit supply and corresponding increase in the costs of trade financing.138 The WTO has estimated that spreads on short-term trade credit facilities peaked at be- tween 300 to 600 basis points (or 3–6%) over LIBOR during the crisis, compared with an average of between 10 to 20 basis points under normal market conditions.139 This dramatic in- crease in the costs of trade financing was compounded by the illiquidity within the f/x swap market, which increased the costs incurred by exporters in order to hedge against volatility in the exchange rate between the U.S dollar and their domes- tic currencies.140 Through these channels, the global dollar shortage at the height of the financial crisis likely contributed to the generalized contraction of credit, international trade, and economic growth. The breakdown of private wholesale funding markets dur- ing the financial crisis forced many banks and other eligible financial institutions to seek ELA and other forms of govern- ment support.141 As we have already seen, however, the cen- tral banks that administered conventional ELA facilities were typically not in a position to provide the U.S. dollars and other foreign currencies that these financial institutions re- quired.142 It was in response to this problem that the Federal Reserve, in coordination with the central banks in several 138 See Daniel McDowell, Emergent International Liquidity Agree- ments: Central Bank Cooperation after the Global Financial Crisis, J. INT’L REL. & DEV. (forthcoming 2017). 139 The Challenges of Trade Financing, WTO (2009), https://www.wto.org/english/thewto_e/coher_e/challenges_e.htm [perma.cc/7NN6-A67Z]. 140 McDowell, supra note 138. 141 ALLEN, supra note 123, at 49. 142 A notable exception were domestic subsidiaries of foreign banks, which typically had access to domestic ELA facilities and could lend on the borrowed currencies to their foreign parents. Wiggins & Metrick, supra note 129, at 7. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 977 other key jurisdictions, undertook one of the most ambitious and important policy interventions of the financial crisis. On December 12, 2007, the Federal Reserve announced that it had entered into temporary reciprocal foreign currency swap lines with the European Central Bank (the “ECB”) and the Swiss National Bank (the “SNB”).143 These swap lines per- mitted the ECB and SNB to exchange their domestic curren- cies for U.S. dollars that could then be lent to banks and other domestic financial institutions in order to plug their U.S. dol- lar funding gaps. At the time, the Fed characterized the ex- tension of these swap lines as designed to “address elevated pressures in short-term funding markets.”144 In effect, the ob- jective was to mitigate the risk that foreign financial institu- tions would be unable to roll over their short-term U.S. dollar liabilities, thereby reducing the pressure on these institutions to engage in destabilizing fire sales of U.S. dollar-denomi- nated assets or cut back on their trade financing and other lending activities.145 These central bank swap lines worked as follows. When a foreign central bank wished to draw on the swap line, it would sell a given quantity of its own currency to the Federal Re- serve at the prevailing exchange rate.146 Contemporaneously with this sale, the central bank would agree to repurchase its currency from the Fed at the same exchange rate at an agreed- upon maturity date.147 The foreign central bank would then 143 Press Release, Bd. of Governors of the Fed. Reserve Sys., Federal Reserve and Other Central Banks Announce Measures Designed to Address Elevated Pressures in Short-term Funding Markets (Dec. 12, 2007), http://www.federalreserve.gov/newsevents/press/monetary/20071212a.htm [perma.cc/BP2C-DCWR]. 144 Id. 145 Michael J. Fleming & Nicolas J. Klagge, The Federal Reserve’s For- eign Exchange Swap Lines, FED. RES. BANK N.Y. CURRENT ISSUES IN ECON. & FIN., April 2010, at 1, 3. 146 Id. at 2. While these swap lines were technically reciprocal, for ease of exposition this description focuses on the mechanics of how foreign cen- tral banks purchased U.S. dollars from the Federal Reserve, lent them on to domestic financial institutions, and subsequently returned them to the Fed. 147 Id. AWREY – FINAL 978 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 establish the eligibility requirements, range of acceptable col- lateral, and other terms and conditions under which it would lend the newly acquired dollars to domestic banks and other financial institutions.148 It would then extend U.S. dollar loans to eligible institutions, typically by way of fixed or vari- able-rate auctions.149 The duration of these loans ranged from overnight, to one week, one month, and three month terms. At the conclusion of the swap, the foreign central bank would re- purchase its currency with U.S. dollars and pay the Fed inter- est equal to the interest rate charged on the dollar loans ex- tended to eligible institutions.150 Importantly, the fact that the foreign central bank stood between the Fed and eligible institutions left it exposed to both credit risk in connection with its U.S. dollar loans and, in the event that these loans were not repaid before the expiry of the swap, foreign ex- change risk in connection with its obligations to repurchase its domestic currency.151 The first swap lines established by the Federal Reserve and other central banks were relatively modest in both their scale and scope. The swap lines between the Fed and the ECB and the SNB were initially capped at $USD20 billion and $USD4 billion, respectively, and only approved for a period of up to 6 months.152 However, as the financial crisis continued 148 See, e.g., News Release, Bank of Eng., U.S. Dollar Overnight Repo Operations: Bank of England Market Notice (Sept. 22, 2008), http://www.bankofengland.co.uk/markets/Documents/market- notice080918.pdf [perma.cc/2ZWU-7K9N]. 149 Fleming & Klagge, supra note 145, at 2. 150 Id. While the Fed did not pay interest on the foreign currency it received under the swap, it did commit to keep the currency in an account in its name at the foreign central bank. Id. at 2–3. 151 Foreign exchange risk arises in this context by virtue of the neces- sity of acquiring U.S. dollars on the open market in order to repurchase the domestic currency. Specifically, were the U.S. dollar to appreciate against the relevant currency during the term of the swap, the fact that the spot and forward rates under the swap were identical would mean that the costs of repurchasing the domestic currency in the second leg of the swap would exceed the costs of acquiring dollars in the first leg. 152 Fed. Res. Press Release, supra note 143. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 979 to unfold, the Fed was compelled to expand this nascent net- work of central bank swap lines and to pump significantly more U.S. dollar liquidity into this network. Following the failure of Lehman Brothers on September 15, 2008, the Fed announced a significant increase in the level of U.S. dollar funding available under its swap lines with the ECB and SNB.153 It also announced new swap lines with the Bank of Japan ($USD60 billion), the Bank of England ($USD40 bil- lion), and the Bank of Canada ($USD10 billion).154 The follow- ing week, the Fed extended swap lines to the Reserve Bank of Australia ($USD10 billion), Svergis Riksbank ($USD10 bil- lion), Norges Bank ($USD5 billion), and Danmarks National- bank ($USD5 billion).155 In response to the continued deterioration of global credit markets, the Fed announced on September 29 that it was more than doubling the aggregate amount of U.S. dollars available under these swap lines from $USD290 billion to $USD620 billion.156 This was followed by the announcement on October 13 that the Fed was removing the cap on its swap lines with the ECB, SNB, and the Bank of England, thus ef- fectively giving these central banks unlimited access to U.S. dollar liquidity.157 The aggregate amount outstanding under 153 Press Release, Bd. of Governors of the Fed. Reserve Sys., Federal Reserve and Other Central Banks Announce Further Measures to Address Elevated Pressures in Funding Markets (Sept. 18, 2008), http://www.feder- alreserve.gov/newsevents/press/monetary/20080918a.htm [perma.cc/ZUW5-2YMG]. 154 Id. 155 Press Release, Bd. of Governors of the Fed. Reserve Sys., Federal Reserve and Other Central Banks Announce Additional Measures to Ad- dress Elevated Pressures in Funding Markets (Sept. 24, 2008), https://www.federalreserve.gov/newsevents/press/mone- tary/20080924a.htm [perma.cc/4VCR-7PV2]. 156 Press Release, Bd. of Governors of the Fed. Reserve Sys., Federal Reserve and Other Central Banks Announce Further Coordinated Actions to Expand Significantly the Capacity to Provide U.S. Dollar Liquidity (Sept. 29, 2008), https://www.federalreserve.gov/newsevents/press/mone- tary/20080929a.htm [perma.cc/2M9T-F7WR]. 157 Press Release, Bd. of Governors of the Fed. Reserve Sys., Federal Reserve and Other Central Banks Announce Further Measures to Provide AWREY – FINAL 980 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 these swap lines would peak at $US586 billion in December 2008 (see Figure 3).158 In total, the Federal Reserve would make over $USD10 trillion available to foreign central banks under these swap lines over the course of the financial cri- sis.159 Similar central bank swap lines were established dur- ing the crisis to provide euros, yen, Swiss francs, and Chinese renminbi.160 FIGURE 3: DRAWINGS UNDER FEDERAL RESERVE SWAP LINES (2007–2009)161 ($USD MILLIONS AS AT END OF QUARTER) 2007 Q 4 2008 Q 1 2008 Q 2 2008 Q 3 2008 Q 4 2009 Q 1 2009 Q 2 2009 Q 3 2009 Q 4 Canada 0 0 0 0 0 0 0 0 0 ECB 20,000 15,000 50,000 174,742 291,352 165,717 59,899 43,662 6,506 Sw itzerland 4,000 6,000 12,000 28,900 25,175 7,318 369 0 0 Japan 0 0 0 29,622 122,716 61,025 17,923 1,530 545 Broad Access to Liquidity and Funding to Financial Institutions (Oct. 13, 2008), https://www.federalreserve.gov/newsevents/press/mone- tary/20081013a.htm [perma.cc/7GGG-G92T]. 158 ALLEN, supra note 123, at 113. 159 James Felkerson, $29,000,000,000,000: A Detailed Look at the Fed’s Bailout by Funding Facility and Recipient 11 (Levy Econ. Inst., Working Paper No. 698, 2011), http://www.levyinstitute.org/pubs/wp_698.pdf [perma.cc/X6XW-ENM2]. 160 For an overview of these other central bank swap networks, see AL- LEN, supra note 123, at 86–98. 161 Id. at 114. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 981 U K 0 0 0 39,999 33,080 14,963 2,503 13 0 D enm ark 0 0 0 5,000 15,000 5,270 3,930 580 0 Australia 0 0 0 10,000 22,830 9,575 240 0 0 Sw eden 0 0 0 0 25,000 23,000 11,500 2,700 0 N orw ay 0 0 0 0 8,225 7,050 5,000 1,000 0 N . Zealand 0 0 0 0 0 0 0 0 0 K orea 0 0 0 0 10,350 16,000 10,000 4,050 0 Brazil 0 0 0 0 0 0 0 0 0 M exico 0 0 0 0 0 0 3,221 3,221 3,221 Singapore 0 0 0 0 0 0 0 0 0 TO TA L 24,000 21,000 62,000 288,263 552,728 309,918 114,585 56,576 10,272 The decision to establish a network of central bank swap lines for the purposes of providing U.S. dollar liquidity to for- eign banks and other financial institutions has been hailed by AWREY – FINAL 982 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 many observers as one of the most important and effective pol- icy responses to the financial crisis.162 Yet it would be a mis- take to conclude from this success that this bilateral swap net- work represents a comprehensive and durable solution to foreign currency liquidity problems. As described in the next section, the design of these swap lines fails to counterbalance the potentially significant distortions generated by this ambi- tious form of state-sponsored liquidity insurance. The swap lines also fail to create truly credible international commit- ments, exposing the jurisdictions that rely on them to unpre- dictable and self-interested decisions by foreign policymakers. These shortcomings beg an important question: how might policymakers design a more effective framework for the provi- sion of FCLA? This question has taken on an added degree of urgency in light of recent market developments. Since the be- ginning of 2014, Eurodollar markets have experienced a num- ber of episodes broadly resembling the U.S. dollar liquidity shortage at the height of the financial crisis.163 These episodes have occurred in the context of a more general trend in terms of the deterioration of conditions within U.S. dollar wholesale funding markets. Figure 4 shows the cross-currency basis swap spreads over 3-month LIBOR between the U.S. dollar and each of the euro, pound sterling, and Japanese yen from 2005 to 2016. These spreads, which reflect the cost of borrow- ing U.S. dollars using f/x swaps, are a useful barometer for potential liquidity problems within wholesale funding mar- kets.164 162 See ALLEN, supra note 123, at 117; Fleming & Klagge, supra note 145, at 7. 163 See Izabella Kaminska, All About the Eurodollars, Redux, FIN. TIMES: ALPHAVILLE (Sept. 24, 2015, 3:18 AM), https://ftal- phaville.ft.com/2015/09/24/2140580/all-about-the-eurodollars-redux/. See also Izabella Kaminska, Petrodollars are Eurodollars, and Eurodollar Base Money is Shrinking, FIN. TIMES: ALPHAVILLE (Jan. 25, 2016, 7:09 AM), https://ftalphaville.ft.com/2016/01/25/2151037/petrodollars-are-eurodol- lars-and-eurodollar-base-money-is-shrinking/. 164 See Michael Chang & Carl Lantz, Credit Suisse Basis Points: Cross- Currency Basis Swaps, CREDIT SUISSE (Apr. 19, 2013), https://credit- suisse.com/researchandanalytics [perma.cc/P2FV-ETPK]. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 983 FIGURE 4: CROSS-CURRENCY BASIS SWAP SPREADS165 As Figure 4 clearly illustrates, basis swap spreads between the U.S. dollar and other major currencies have recently wid- ened to levels last observed during the global financial crisis and subsequent European sovereign debt crisis.166 The grow- ing stress within these markets has been attributed to several factors, including: the sustained increase in the price of the U.S. dollar relative to the euro, pound sterling, and other ma- jor currencies over this period167; the ECB’s negative interest 165 Thomson Reuters (2016). 166 Although, the more gradual deterioration of market conditions from 2014 onward will have undoubtedly provided market participants with the opportunity to more proactively address the resulting funding pressures by, for example, reducing any f/x imbalances on their balance sheets. 167 See Hyun Song Shin, Econ. Advisor & Head of Research, Bank of Int’l Settlements, Speech Presented at the World Bank Conference, The State of Economics, the State of the World, Washington, DC: Global Liquid- ity and Procyclicality, 2–5 (June 8, 2016), http://www.bis.org/speeches/sp160608.htm [perma.cc/Q376-HFXB]. See also Izabella Kaminska, Textbook Defying Global Dollar Shortages, FIN. TIMES: ALPHAVILLE (June 9, 2016, 7:45 AM), https://ftal- phaville.ft.com/2016/06/09/2165690/textbook-defying-global-dollar-short- ages/. AWREY – FINAL 984 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 rate policy168; the introduction of comprehensive reforms tar- geting U.S. money market funds163; and programs such as the Federal Reserve’s new Term Deposit Facility that have had the effect of incentivizing U.S. banks to repatriate short-term dollar liquidity from abroad.169 Looking forward, this stress is likely to intensify as the Fed raises interest rates from their current historic lows and ramps up policies such as reverse repo facilities designed to unwind quantitative easing and thus absorb dollar liquidity from the international financial system.170 The resulting withdrawal of short-term U.S. dollar liquidity will inevitably put pressure on foreign banks, other financial institutions, and commercial firms that rely on Eu- rodollar markets.171 The existing central bank swap lines do little to constrain the build-up of these foreign currency funding pressures. There is also no guarantee, especially in light of the shifting political winds in the United States, that the Federal Reserve will intervene to support global U.S. dollar funding markets during the next financial crisis—or that it will take into ac- count the global repercussions of its own policy decisions in potentially triggering this crisis. President Trump’s vision for ‘Making America Great Again’ may not include the provision of FCLA to foreign banks and other financial institutions. These developments suggest that the foreign currency liquid- ity crisis of 2007–2009 may provide us with a window into the catalysts of future crises. They also frame why designing a more effective framework for the provision of FCLA should be 168 See Izabella Kaminska, Eurodollars, FX Reserve Managers and the Offshore RRP Issue, FIN. TIMES: ALPHAVILLE (Sept. 1, 2015, 5:39 AM), https://ftalphaville.ft.com/2015/09/01/2139085/eurodollars-fx-reserve-man- agers-and-the-offshore-rrp-issue/. 169 See Izabella Kaminksa, All About the Eurodollars, FIN. TIMES: AL- PHAVILLE (Sept. 5, 2014, 11:14 AM), https://ftal- phaville.ft.com/2014/09/05/1957231/all-about-the-eurodollars/; Kaminska, supra note 168. 170 See Kaminska, supra note 168. 171 Especially where these firms’ balance sheets are characterized by significant maturity and currency mismatches. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 985 viewed by policymakers as an important and pressing prior- ity. V. DESIGNING AN EFFECTIVE FRAMEWORK FOR FCLA The turmoil of 2007–09 revealed an important fault line in the structure of the global financial and economic system. On one side of the divide is a global network of financial markets and institutions characterized by high levels of integration and interconnectedness, where financial institutions located in one country can, with the stroke of a few keys create liabil- ities denominated in the currency of another country, and where these liabilities can grease the wheels of trade, finance, and investment across every corner of the globe. On the other side of the divide is an emerging, but still highly fragmented, international monetary, regulatory, and crisis management architecture, vulnerable to regulatory arbitrage, acute coordi- nation problems, and conflicting domestic political prefer- ences. As Benn Steil and Manuel Hinds observe, this divide represents a significant threat to economic globalization: “If anything is likely to throw globalization into reverse, it is not trade itself, but the money that facilitates it. National monies and global markets simply do not mix; together they make a deadly brew of currency crises and geopolitical tension and create ready pretexts for damaging protectionism.”172 This Part examines some of the ways that one might begin to bridge this divide by designing a more effective framework for the provision of FCLA: one that reflects the imperatives of Bagehot’s Dictum while simultaneously constraining the moral hazard and other problems generated by this ambitious state-supported liquidity insurance. Broadly speaking, the available policy alternatives include the introduction of a blanket prohibition against the issuance of Eurodollar liabili- ties, replacing the Eurodollar market with a global currency and central bank, and strengthening the existing interna- tional architecture for the provision of FCLA. 172 STEIL & HINDS, supra note 10, at 8. AWREY – FINAL 986 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 A. A Blanket Prohibition Against Eurodollar Liabilities As this Article has shown, scholars such as Morgan Ricks see a clear relationship between private money creation, runs, and financial instability.173 As Ricks observes, while deposit guarantee schemes and ELA facilities have proven to be rela- tively effective mechanisms for dampening the destabilizing impact of private money creation within the conventional banking system, the effectiveness of these mechanisms has been eroded by the emergence and growth of a vibrant and diverse shadow banking system.174 In response, Ricks pro- poses a public-private partnership pursuant to which banks would be licensed as the exclusive issuers of all deposits, short- term liabilities, and other “money-claims.”175 These money- claims would be explicitly backed by a state guarantee, thereby eliminating the distinction between public and pri- vate money and, with it, the incentives of short-term creditors to engage in destabilizing run-like behavior.176 In exchange, the state would receive a periodic fee from each bank based on the returns from its investment portfolio and the quantity of its issued and outstanding money-claims.177 Banks would also be subject to portfolio, capital, and other prudential regulatory requirements designed to ameliorate any potential moral haz- ard problems generated by the state guarantee.178 Ricks’ proposal envisions that financial institutions other than licensed banks would be prohibited from issuing deposit 173 See the discussion of Ricks’ “money problem” in Part III, supra. 174 RICKS, supra note 79, at 215–19. As Ricks observes, deposit guaran- tee schemes and ELA facilities also introduce potentially significant distor- tions: encouraging growth in the size of financial institutions, the use of leverage, and the oversupply of credit to the economy. Id. at 3. 175 Id. at 226, 235. Under Ricks proposal, banks would be licensed to issue a specific quantity of money-claims. The system would then employ a cap and trade system, enabling banks to sell excess capacity to other banks with more attractive investment opportunities. Id. at 228. 176 Id. at 227, 241. 177 Id. at 226. 178 Id. at 225. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 987 liabilities and other money-claims.179 Given the domestic lo- cus of bank licensing regimes, one of the logical consequences of this proposal would thus be the complete abolition of the Eurodollar market. This is no accident. As Ricks explains: “Simply put, the Eurodollar markets—and the Eurocurrency markets more generally—are incompatible with financial sta- bility”.180 Writing from the vantage point of the United States, Ricks goes on to identify two possible avenues for abolishing the Eurodollar market. The first avenue would be to impose a blanket prohibition on the issuance of short-term U.S. dollar liabilities by financial institutions incorporated and licensed outside the United States. In pursuing this avenue, the United States would find little support in public international law. The IMF Articles of Agreement do not prohibit the use of a currency issued by one member state as legal tender in an- other member state.181 Similarly, while customary interna- tional law recognizes the sovereignty of states over monetary affairs, it does not provide a clear basis for incursions into the monetary sovereignty of another state for the purpose of re- stricting the use of foreign currencies.182 What is more, nei- ther the IMF Articles of Agreement nor customary interna- tional law explicitly address the issue of private money creation.183 As a result, any prohibition would almost cer- 179 Id. at 226. 180 Id. at 239. 181 For the analysis of one former IMF general counsel on this point, see François Gianviti, Use of a Foreign Currency Under the IMF’s Articles of Agreement, in 3 CURRENT DEVELOPMENTS IN MONETARY AND FINANCIAL LAW 817 (2005). 182 LASTRA, supra note 7, at 17 (“With regard to the question of whether the issuer of a currency may prohibit the use of its currency by another country, there is at present no basis in international law to support the view that the issuer may object to the currency being used by other countries.”). Consistent with this position, courts in the United States have explicitly acknowledged the right of the U.S. government to designate foreign curren- cies as legal tender without the consent of the issuing country. See Tyson v. United States, 285 F.2d 19, 22 (10th Cir. 1960). 183 In this respect, the threat that a contract creating a Eurodollar lia- bility may be invalidated under private international law—e.g., that the AWREY – FINAL 988 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 tainly need to take the form of unilateral restrictions on mar- ket access: for example, denying foreign banks and other fi- nancial institutions that issue Eurodollar liabilities (and their affiliates) access to both the Federal Reserve’s ELA and other liquidity facilities, as well as the U.S. dollar clearing services provided by domestic banks.184 In theory, a unilateral prohibition would serve to shift the risk of foreign currency liquidity problems to the financial in- stitutions that issued Eurodollar liabilities and, ultimately, to the jurisdictions that permitted them to do so. In practice, however, the impact of this unilateral prohibition is likely to be relatively modest insofar as foreign institutions still en- joyed access to U.S. dollar liquidity through international wholesale funding and f/x swap markets.185 Somewhat per- versely, then, by increasing the reliance of foreign institutions on these markets, restrictions on market access may only serve to amplify the instability associated with widespread market breakdown. The second avenue would be to negotiate a multilateral agreement pursuant to which signatory states agreed to pro- hibit domestic financial institutions from issuing Eurodollar liabilities.186 Where the signatories to this agreement include the governments of other major international financial cen- ters, this avenue would have the effect of bringing interna- U.S. prohibition would be applied on an extraterritorial basis—may prove more effective in disrupting the Eurodollar market. For a discussion of the use of private international law as a mechanism for constraining interna- tional regulatory arbitrage, see Annelise Riles, Managing Regulatory Arbi- trage: A Conflicts of Law Approach, 47 CORNELL INT’L L. J. 63 (2014). 184 See RICKS, supra note 79, at 239. In theory, of course, these market access restrictions could be even broader: encompassing not only U.S. dollar clearing services, but also restrictions on the ability to conduct any business in U.S. markets or with U.S. clients. 185 Although the fact that these markets are prone to breakdown dur- ing periods of widespread panic and instability would almost certainly con- tribute to the build-up of risk within the remaining Eurodollar market. 186 RICKS, supra note 79, at 239–40. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 989 tional wholesale funding and f/x swap markets within the pe- rimeter of this prohibition.187 Simultaneously, of course, a multilateral agreement constraining the issuance of Eurodol- lar liabilities may simply serve to drive these markets to other (non-signatory) states. While Ricks views this multilateral av- enue as more promising, the likelihood of reaching an agree- ment—to say nothing of its effectiveness—would ultimately hinge on the potential for this type of regulatory arbitrage and the corresponding threat that the issuance of Eurodollar lia- bilities outside the perimeter of the prohibition would under- mine financial stability within signatory states.188 There is compelling logic in Ricks’ argument that strength- ening sovereign authority over private money creation would yield significant benefits from the perspective of domestic fi- nancial stability. A blanket prohibition against the issuance of Eurodollar liabilities would shift responsibility for the pro- vision of FCLA to the countries and financial institutions that failed to comply with the prohibition, thereby strengthening their incentives to more proactively address potential foreign currency liquidity problems, and reducing potential moral hazard problems. Restrictions on market access for foreign banks and other institutions issuing Eurodollar liabilities would also potentially help reduce—although almost certainly not eliminate—the risk of cross-border contagion stemming from the materialization of foreign currency liquidity prob- lems. Ultimately, however, these domestic benefits must be weighed against the wider implications of abolishing the Eu- rodollar market. As this Article has described, a significant 187 An agreement covering the United States, European Union, United Kingdom, Singapore, Hong Kong, and Japan, for example, would cover the vast majority of transactions within the f/x swap market. See Bank of Int’l Settlements, Turnover of OTC Foreign Exchange Instruments, by Country (Sept. 1, 2016), http://www.bis.org/statistics/d11_2.pdf [perma.cc/EK3Q- JEHY]. 188 As discussed in greater detail in Section V.C, any agreement would also likely need to be accompanied by credible restrictions on market access targeting financial institutions incorporated and licensed to carry on busi- ness in non-signatory states. AWREY – FINAL 990 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 proportion of international trade outside the United States is invoiced and settled in U.S. dollars. This trade is often sup- ported by short-term U.S. dollar trade financing provided by foreign banks and other financial institutions.189 Foreign fi- nancial institutions are also important participants in inter- national U.S. dollar-denominated wholesale funding and f/x swap markets. The funding obtained in these markets is then often used to support U.S. dollar loans and other invest- ments.190 The elimination of the Eurodollar market would thus necessitate a significant restructuring of many of the markets and institutions at the heart of the global financial and economic system. While this restructuring may ulti- mately represent a transitional problem, it is a problem that may nevertheless trigger significant economic disruption and financial instability. While it is obviously difficult to predict precisely how this restructuring would unfold, two possible scenarios stand out. These scenarios are not mutually exclusive, and may even play out in parallel with one another.191 In the first scenario, large numbers of foreign banks might seek to obtain banking licenses in the United States in order to ensure their contin- ued ability to issue U.S. dollar denominated money-claims. The net effect of this migration would thus be to bring the Eu- rodollar market back ‘onshore’. It would also create an enor- mous concentration of financial and economic activity in one country, further entrenching the United States as the de facto 189 While Ricks excludes “trade credit” from his definition of a money- claim, this must be distinguished from trade financing arrangements such as letters of credit, supply chain finance, or factoring. Whereas trade credit effectively refers to the ability to purchase goods on credit and the resulting creation of short-term inter-firm liabilities generally referred to as “ac- counts payable,” trade financing generally involves the extension of loans, guarantees, or other short-term liabilities by third party financial institu- tions for the purpose of facilitating international trade. That Ricks’ excep- tion only extends to these narrow inter-firm liabilities is apparent from his proposed statutory language. Id. at 244. 190 Without access to these markets, foreign financial institutions hold- ing U.S. dollar assets would face significant foreign exchange risk. 191 Although it is likely that gravitation toward either extreme would be the most likely to engender pronounced instability. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 991 global lender of last resort and increasing the exposure of the rest of the world to its fiscal, monetary, and regulatory policy decisions.192 Strengthening the monetary sovereignty of the United States would thus come at the expense of the policy sovereignty of other countries. In the second scenario, firms might switch from the U.S. dollar to other currencies for the purposes of international trade, finance, and investment. The resulting fragmentation of the currency composition of these business activities would likely precipitate an increase in financing and trading costs and create an impediment to global economic integration. Im- portantly, the existence of winners and losers under both of these scenarios renders the prospects of a binding multilateral agreement to dismantle the Eurodollar market decidedly bleak. At the same time, any unilateral action would risk pro- voking retaliatory action from other states. Moreover, under either scenario there is a significant risk that the restructur- ing process may unfold in a disorderly fashion. Viewed from this perspective, the practical effect of Ricks’ proposal may simply be to undermine financial and economic globalization and export financial instability from the United States to the rest of the world. B. A Global Currency and Central Bank Ricks’ proposal to dismantle the Eurodollar market envi- sions the vigorous assertion of the state’s sovereign authority over private money creation. Other, even more radical, pro- posals envision the transfer of this authority to international institutions. Proposals for an international monetary author- ity and currency have a long and colorful history. Perhaps the most (in)famous proposal is John Maynard Keynes’ scheme for 192 Thereby further intensifying the global financial cycle. See Hélène Rey, Dilemma not Trilemma: The Global Financial Cycle and Monetary Pol- icy Independence, FED. RES. BANK KANSAS CITY (2013), https://www.kansas- cityfed.org/hgiSo/publicat/sympos/2013/2013Rey.pdf [perma.cc/47G7- Y6BT]. AWREY – FINAL 992 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 the creation of an ‘international clearing union.’193 Advanced by Keynes at the Bretton Woods conference in July 1944, the international clearing union was designed to prevent the build-up of destabilizing international trade imbalances through a multilateral system of debits and credits denomi- nated in a new unit of account known as ‘bancor.’194 Recently, scholars such as Robert Hockett have returned to Keynes’ original proposal and advocated it as the basis for the creation of a global currency and central bank.195 While the institu- tional details of these proposals vary, they generally involve the creation of a new currency, issued and managed by an in- ternational monetary authority, and intended for use along- side existing domestic and supranational currencies.196 The benefits of a global currency are typically framed as flowing from the potential to create a global ‘risk-free’ asset, 193 See JOHN MAYNARD KEYNES, ACTIVITIES 1940–1944: SHAPING THE POST-WAR WORLD: THE CLEARING UNION (1980), reprinted in 25 THE COL- LECTED WRITINGS OF JOHN MAYNARD KEYNES (Donald Moggridge ed., Cam- bridge Univ. Press, 1980). 194 See id. for further details regarding Keynes’ scheme as originally proposed. The scheme encountered significant resistance at the Bretton Woods conference, especially from the United States, and was abandoned in favor of what would eventually become the IMF. 195 Robert Hockett, Bretton Woods 1.0: A Constructive Retrieval for Sus- tainable Finance, 16 N.Y.U. J. LEGIS. & PUB. POL’Y. 401 (2013). See also EM- MANUEL FARHI ET AL., REFORMING THE INTERNATIONAL MONETARY SYSTEM 31–49 (2011). 196 At the same time, not all proposals envision the creation of a new currency. See FARHI ET AL., supra note 195. A related, if somewhat more modest, proposal is to promote the use of the Special Drawing Rights (“SDRs”) issued by the IMF in international commercial transactions. For a recent synthesis of the main arguments for and against the expanded use of SDRs in international commercial transactions, see IMF, Enhancing Inter- national Monetary Stability—A Role for the SDR? (Jan. 7, 2011), https://www.imf.org/external/np/pp/eng/2011/010711.pdf [perma.cc/8GAV- J9UH]. More ambitious proposals, meanwhile, envision that an interna- tional currency could eventually replace existing national currencies in the same way that the euro did for euro area member states. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 993 reduce exchange rate volatility, and moderate the destabiliz- ing impact of global trade imbalances.197 In theory, however, the creation of a global currency and central bank could also help address foreign currency liquidity problems. The promise of an international currency—let’s call it ‘bancor’ in honor of Keynes198—from this perspective resides in the ability of a new international monetary authority (“IMA”) to effectively consolidate authority for monetary policy, the licensing and prudential regulation of financial institutions engaged in pri- vate money creation, and the access of these institutions to FCLA. The logical starting point would be for signatory states to confer upon the IMA clear authority over all matters relat- ing to the issuance of short-term liabilities denominated in bancor. Banks and other financial institutions licensed to carry on business in these jurisdictions would then be re- quired to apply to the IMA for an additional license to be able to issue short-term bancor denominated liabilities.199 This li- cense would then be granted on the condition that financial institutions comply with a limited range of targeted pruden- tial requirements—e.g., currency-specific portfolio, capital, and liquidity constraints—designed to eliminate the prospect 197 See, e.g., Mark Carney, then Governor of the Bank of Can., Remarks at the Spruce Meadows Changing Fortunes Round Table, Calgary, Alberta: Restoring Faith in the International Monetary System (Sept. 10, 2010), http://www.bis.org/review/r100916a.pdf [perma.cc/ZG7U-EWM7]. 198 And in honor of the tradition of naming hypothetical international currencies ‘bancor’ in honor of Keynes. See e.g., Hockett, supra note 195, at 480. 199 Notably, direct supervisory authority over domestic financial insti- tutions is not a feature of most proposals for the creation of an IMA and currency. One possible alternative would be for signatory states to retain supervisory authority under the oversight of an IMA. This oversight author- ity would then need to be coupled with a credible mechanism for imposing costs on signatory states that failed to (i) adopt adequate prudential regu- lation and supervision in connection with bancor denominated assets and short-term liabilities or (ii) take action to prevent large foreign exchange imbalances within their domestic banking systems. AWREY – FINAL 994 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 of regulatory arbitrage and other distortions.200 Financial in- stitutions would also be required to provide the IMA with in- formation about their bancor denominated assets and liabili- ties on an ongoing basis.201 Where these institutions experienced temporary liquidity problems, the IMA would then stand ready to provide bancor loans collateralized against high quality bancor denominated assets on terms that—reflecting Bagehot’s dictum—minimized potential moral hazard problems. By more firmly tethering monetary and regulatory authority, the creation of an international cur- rency and central bank could thus eliminate some of the fric- tions that currently undermine the ability of domestic policy- makers to monitor the build-up of the risks associated with private money creation, impose consistent and effective regu- latory constraints across jurisdictions and institutions, and respond in a more timely and comprehensive fashion to inter- national financial crises. It would be an understatement to say that there are a num- ber of significant obstacles to the creation and successful adop- tion of an international currency and central bank. The first obstacle would likely be to satisfy policymakers that the global economy represented what economists refer to as an “optimal currency area.”202 Whether a particular economic region con- stitutes an optimal currency area is a function of a number of factors. These factors include: the volume of cross-border trade between the countries within the region; existing barri- ers to the free movement of goods, labour, and other factors of production between these countries; the degree of homogene- 200 In order to minimize unnecessary divergence or duplication, these requirements could be linked to the capital adequacy and liquidity rules produced by the Basel Committee on Banking Supervision (BCBS). See BA- SEL COMM. ON BANKING SUPERVISION, BASEL III: A GLOBAL REGULATORY FRAMEWORK FOR MORE RESILIENT BANKS AND BANKING SYS. (2011), http://www.bis.org/publ/bcbs189.pdf [perma.cc/382V-BRCH]. 201 The type of information that could be provided, and the uses to which it might be put, are examined in greater detail in Section V.C. 202 See Robert A. Mundell, A Theory of Optimal Currency Areas, 51 AM. ECON. REV. 657 (1961). AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 995 ity in the economic structure and business cycles of the indus- tries within these countries and; any differences in the vola- tility of prices across the region.203 The higher the level of eco- nomic integration and homogeneity, the more likely it is that a region will represent an optimal currency area. At the heart of this determination is then a tradeoff between the antici- pated reduction in transaction costs and uncertainty associ- ated with a currency union versus the constraints that this union imposes on the ability of domestic policymakers to ad- just to idiosyncratic shocks through changes in domestic in- terest and exchange rates.204 The concept of an optimal currency area is far from perfect. Amongst other problems, the type of static and homogeneous economic regions that are most likely to represent optimal currency areas are—for the very same reasons—likely to suf- fer from sclerotic growth and innovation and inefficient and potentially destabilizing under-diversification.205 Neverthe- less, even a cursory examination of the factors that determine whether a region constitutes an optimal currency area sug- gests that—as a threshold matter—the economic case for the creation of an international currency is far from clear cut. The second challenge would be the naked political one that Keynes himself encountered at Bretton Woods.206 How does one persuade sovereign countries—and especially countries 203 For a synthesis of these variables, see PAUL R. KRUGMAN & MAURICE OBSTFELD, INTERNATIONAL ECONOMICS: THEORY & POLICY, 558–72 (7th ed. 2006). See also STEIL & HINDS, supra note 10, at 138–45. 204 Maurice Obstfeld, Professor of Econ., U. of Cal. Berkeley, Paul Mas- son, Senior Advisor, IMF Res. Dep’t, Robert Mundell, 1999 Nobel Laureate and Professor of Econ., Colum. U., Remarks at the IMF Economic Forum— One World, One Currency: Destination or Delusion? (Nov. 8, 2000), https://www.imf.org/en/News/Articles/2015/09/28/04/54/tr001108 [perma.cc/PG5P-E6C5]. 205 See STEIL & HINDS, supra note 10, at 141–45 for a more detailed overview of some of the theoretical problems associated with the concept of an optimal currency area. 206 For an excellent account of Keynes’ experience at the Bretton Woods conference, see BENN STEIL, THE BATTLE OF BRETTON WOODS: JOHN MAYNARD KEYNES, HARRY DEXTER WHITE, AND THE MAKING OF THE NEW WORLD ORDER (2013). AWREY – FINAL 996 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 that derive substantial benefits from the existing interna- tional system—to transfer authority over an important area of economic affairs to an international institution? This chal- lenge would inevitably be compounded by thorny issues of in- stitutional design. What sort of constraints would need to be imposed on domestic trade, fiscal, and monetary policy in or- der to make the new currency credible? How would the bur- dens of ELA be distributed between countries? And how would senior positions within the IMA be filled? These and other questions would all need to be answered in a way that per- suaded the largest number of countries—representing the largest possible proportion of global trade, finance, and invest- ment—to agree to the creation of a new international cur- rency. As demonstrated by the European Union’s ongoing ex- periment with the single currency, failing to properly address these issues at the outset could undermine the new currency’s long-term credibility. The third and fundamentally intertwined challenge would be how to make a new international currency attractive to a critical mass of financial and commercial counterparties. Cur- rencies are often viewed as classic ‘network’ goods.207 In a nut- shell, where a currency reaches a threshold level of use within a given market, it starts to develop its own gravity: with users attracted not only by the currency’s intrinsic value, but also by the fact that it is widely used. The path dependency created by these network effects is reflected in a form of ‘stickiness’— entrenching incumbent currencies and erecting a significant barrier to the emergence and widespread adoption of potential 207 See Paul Krugman, Vehicle Currencies and the Structure of Interna- tional Exchange, 12 J. MONEY, CREDIT & BANKING 513 (1980); Paul Krugman, The International Role of the Dollar: Theory and Prospects, in EXCHANGE RATE THEORY AND PRACTICE (1984); Kiminori Matsuyama et al., Toward a Theory of International Currency, 60 REV. ECON. STUD. 283 (1993); PHILIPP HARTMANN, CURRENCY COMPETITION AND FOREIGN EXCHANGE MAR- KETS: THE DOLLAR, THE YEN AND THE EURO (Cambridge Univ. Press, 1998); Hélène Rey, International Trade and Currency Exchange, 68 REV. ECON. STUD. 443 (2001); Swoboda, supra note 26. For an alternative perspective, see generally Livia Chitu et al., When Did the Dollar Overtake Sterling as the Leading International Currency? Evidence from the Bond Markets (Nat’l Bureau of Econ. Research, Working Paper No. 18097, 2012). AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 997 substitutes. We might expect these network effects to be par- ticularly pronounced in the case of an international currency given that one of the most frequently cited intrinsic benefits— relatively low exchange rate volatility—is itself partly a func- tion of the depth and volume of the market for the currency itself. Accordingly, even if public policymakers were able to surmount the political obstacles to the creation of an interna- tional currency, there would remain the sticky problem of how to attract private market participants away from the U.S. dol- lar and other major international currencies.208 Finally, the introduction of an international currency would throw up a host of technocratic policy challenges. One of the most important of these challenges would be how to co- ordinate the monetary policy of the IMA with the policies of signatory states. There would also be the challenge of ensur- ing that the introduction of an international currency was not itself a source of financial instability. As the ECB has warned, there may be significant “unintended consequences”209 associ- ated with “replacing international currencies that have estab- lished themselves as a result of the autonomous decisions of private and official agents with a synthetic, policy-imposed, international currency.”210 Given these challenges, the crea- tion and widespread adoption of a new international currency seems both highly unlikely and potentially undesirable. C. Reform of Existing Institutional Arrangements for FCLA The third and final strategy is to strengthen existing inter- national arrangements for the provision of FCLA. Two sets of 208 Thereafter, there would also be the problem of maintaining the cur- rency’s attractiveness—with any failure to do so potentially precipitating a flight (back) to other currencies. 209 Ettore Dorrucci & Julie McKay, The International Monetary System After the Financial Crisis, OCCASIONAL PAPER SERIES (Eur. Cent. Bank, Frankfurt, Ger.), Feb. 2011, at 35. 210 Id. AWREY – FINAL 998 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 arrangements stand out as particularly promising and, simul- taneously, ripe for potential reform.211 The first is the network of bilateral central bank swap lines established in response to the financial crisis. The first generation of swap lines estab- lished between the Federal Reserve and other central banks expired in February 2010.212 Within a matter of months, how- ever, the deterioration of wholesale funding markets precipi- tated by the European sovereign debt crisis prompted the Fed to re-establish temporary swap lines with the ECB, SNB, Bank of England, Bank of Japan, and Bank of Canada.213 In October 2013, the Fed announced that these temporary swap lines would be converted into standing facilities enabling any of the six central banks to borrow an unlimited amount of any of the currencies issued by the other participating central banks. As the Fed explained in its announcement: The conversion of these liquidity lines with pre-set ex- piration dates to standing lines further supports fi- nancial stability by reducing uncertainties among 211 There are a number of other—predominantly regional—arrange- ments for the provision of FCLA. These include the Chiang Mai Multilateral Mechanism administered by the ASEAN group of countries. These arrange- ments are also potentially important and, arguably, in need of reform. The focus of this paper on the network of central bank swap lines and IMF emer- gency lending facilities reflects the potential impact of these arrangements at this point in time. The swap lines and IMF lending facilities are also unique insofar as they benefit from the direct participation of the United States as issuer of the world’s most widely used reserve currency. 212 Press Release, Bd. of Governors of the Fed. Res. Sys., FOMC State- ment (Jan. 27, 2010), http://www.federalreserve.gov/newsevents/press/mon- etary/20100127a.htm [perma.cc/G8K4-T2RN]. 213 Press Release, Bd. of Governors of the Fed. Res. Sys., FOMC State- ment: Federal Reserve, European Central Bank, Bank of Canada, Bank of England, and Swiss National Bank Announce Reestablishment of Tempo- rary U.S. Dollar Liquidity Swap Facilities (May 9, 2010), https://www.fed- eralreserve.gov/newsevents/press/monetary/20100509a.htm [perma.cc/HDN8-6SJZ]; Press Release, Bd. of Governors of the Fed. Res. Sys., FOMC Statement: FOMC Authorizes Re-establishment of Temporary U.S. Dollar Liquidity Swap Arrangement with the Bank of Japan (May 10, 2010), https://www.federalreserve.gov/newsevents/pressreleases/mone- tary20100510a.htm [perma.cc/W562-9V73]. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 999 market participants as to whether and when these ar- rangements would be renewed. This action results from the ongoing cooperation among these central banks to help maintain financial stability and confi- dence in global funding markets.214 What began as a somewhat ad hoc response to the financial crisis has thus rather quietly metamorphosized into an im- portant—and seemingly permanent—feature of the global fi- nancial crisis management architecture. The standing swap facilities announced in October 2013 are enshrined in a series of bilateral agreements between each of the participating central banks.215 These agreements share several important features. The first is that the extension of FCLA under these agreements is highly contingent on domes- tic legal and political considerations. The wording of these agreements strongly suggests that swap transactions will only be approved and executed upon the mutual agreement of both parties at the time one of them actually requests FCLA.216 The extension of FCLA must also be consistent with the stat- utory mandates of the relevant central banks. These require- ments are in all likelihood linked; the contingent nature of the commitment is ultimately a reflection of the need to ensure that any given swap transaction falls squarely within the scope of the central bank’s legal authority. Perhaps not sur- 214 Frequently Asked Questions: Standing Lines for U.S. Dollar and Foreign Currency Liquidity Swaps, BD. OF GOVERNORS OF THE FED. RES. SYS., (Oct. 2013), https://www.federalreserve.gov/monetarypolicy/bst_swap- faqs.htm. 215 The agreements among the Fed and participating central banks are available at Central Bank Liquidity Swaps, FED. RES. BANK OF N.Y., https://www.newyorkfed.org/markets/liquidity_swap.html [perma.cc/8JKM-L5GY]. 216 See e.g., U.S. Dollar-British Pounds Swap Agreement, FED. RES. BANK OF N.Y. (Jan. 16, 2014), https://www.newyorkfed.org/mediali- brary/media/markets/USD_Pound_swap_agreement.pdf [perma.cc/2533- UWW3]. Section 1(c) of this agreement begins: “The Parties agree, at such times as they mutually agree, to enter into Swap Transactions . . .” (empha- sis added). AWREY – FINAL 1000 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 prisingly, however, these mandates rarely make explicit ref- erence to the promotion of monetary or financial stability in other countries. In the United States, for example, the statutory authority of the Fed to enter into these swap transactions is provided by section 14 of the Federal Reserve Act.217 Section 14 authorizes each of the twelve regional Federal Reserve banks to purchase and sell foreign currencies as a part of the Fed’s open market operations.218 While this might appear to give the Fed a wide berth, these transactions must still be consistent with its wider mandate to promote maximum employment, stable prices, and moderate long-term interest rates.219 This man- date has been interpreted by senior Fed officials as implicitly encompassing the promotion of financial stability.220 Con- sistent with this interpretation, the Fed has signaled that 217 Federal Reserve Act, 12 U.S.C. § 226 (1913). 218 Section 14 of the Federal Reserve Act states: “Any Federal reserve bank may, under rules and regulations prescribed by the Board of Gover- nors of the Federal Reserve System, purchase and sell in the open market, at home or abroad, either from or to domestic or foreign banks, firms, cor- porations, or individuals, cable transfers” (emphasis added). Written in 1913, the reference to “cable transfers” is in fact a reference to foreign cur- rencies. Section 14(e), meanwhile, permits any Federal reserve bank to: [O]pen and maintain accounts in foreign countries, appoint corre- spondents, and establish agencies in such countries wheresoever [sic] it may be deemed best for the purpose of purchasing, selling, and collecting bills of exchange, and to buy and sell, with or with- out its indorsement [sic], through such correspondents or agen- cies, bills of exchange (or acceptances) arising out of actual com- mercial transactions which have not more than ninety days to run… and… to open and maintain banking accounts for such for- eign correspondents or agencies, or for foreign banks or bankers, or for foreign states. . . . Federal Reserve Act, 12 U.S.C. § 358 (1913). 219 See Federal Reserve Act, 12 U.S.C. § 2A (1913). 220 See e.g., Tom Baxter, Exec. Vice President and Gen. Counsel of the Fed. Res. Bank of N.Y., Remarks at the Future of Banking Regulation and Supervision in the EU Conference, Frankfurt, Germany: Financial Stabil- ity: The Role of the Federal Reserve System (Nov. 15, 2013), https://www.newyorkfed.org/newsevents/speeches/2013/bax131120 [perma.cc/J6TU-AL4X]. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 1001 swap transactions under these facilities will generally be ap- proved only where they are “directed at countering disorderly market conditions.”221 The Fed has also produced a video, bluntly titled “Why Swap Lines are in the U.S. National In- terest,” explaining how the facilities support investment, em- ployment, and economic growth in the United States.222 In light of these and other similar statements, it seems likely that swap transactions will only be approved where there is a clear and present danger to the financial stability of the juris- diction that has received a request for FCLA.223 Where there exists no such threat, or where providing liquidity assistance to foreign financial institutions might prove politically costly, the structure of these agreements—to say nothing of the ab- sence of a generally agreed upon definition of ‘financial stabil- ity’—gives participating central banks significant discretion to decline a request for FCLA. The second important feature of these swap agreements is the complete absence of both ex ante qualifications for partic- ipating jurisdictions and ex post constraints on the extension of FCLA to eligible financial institutions. The Federal Reserve has ostensibly selected participating jurisdictions on the basis of its own evaluation of the potential threat that these juris- dictions pose to the financial stability of the United States.224 221 Minutes of the Federal Open Market Committee Meeting, FED. RES. BANK N.Y. 6 (Jan. 28–29, 2014), https://www.federalreserve.gov/monetary- policy/fomcminutes20140129.htm [perma.cc/CJ9J-6LXD]. 222 See Protecting the U.S. Economy from Strains Abroad—Why Swap Lines Are in the U.S. National Interest, FED. RES. BANK N.Y. https://www.newyorkfed.org/education/liqswap.html [perma.cc/WN9K- PXB8]. 223 See Denbee et al., supra note 51, at 11 (observing, “the scope of swap line liquidity provision is very narrow, and typically limited to facilitating market-wide liquidity provision to cross-border banks for the purposes of domestic financial stability”). 224 See e.g., Meeting of the Federal Open Market Committee on October 28–29, 2008, FED. RES. BANK N.Y. 10–11, 32–35 (Oct, 28–29 2008), https://www.federalreserve.gov/monetarypolicy/files/FOMC20081029meet- ing.pdf [perma.cc/4NZV-H736] (emphasizing the size of a country’s econ- omy, its structural interconnectedness with the global financial system, and whether it is a major financial center as important factors in determining AWREY – FINAL 1002 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 While this approach is broadly understandable, the absence of more formal and objective ex ante qualifications has also re- sulted in at least one questionable inclusion (Canada) and one glaring omission (China). For participating jurisdictions, meanwhile, the swap agreements do not impose any limits around the types of financial institutions to which recipient central banks can supply foreign currency funding. Nor do these agreements in any way dictate the terms upon which this funding will be made available to these institutions. Per- haps most importantly, the agreements do not prescribe that FCLA be provided against good quality collateral, at a penalty rate of interest, and only to fundamentally solvent but tempo- rarily illiquid banks and other financial institutions. Central banks in receipt of FCLA are also under no obliga- tion to ensure that the prudential regulatory requirements imposed on eligible financial institutions are functionally equivalent to those imposed on financial institutions in the ju- risdiction that provides this assistance. While some observers have argued that the BCBS capital and liquidity standards ensure a sufficient degree of regulatory and supervisory har- monization, the implementation of these standards varies from country to country. There are also signs that the Basel consensus is beginning to fray around the edges, with the Eu- ropean Union having recently been determined by the BIS to be in material non-compliance with Basel standards.225 The absence of ex ante qualifications and ex post constraints thus introduces the prospect that FCLA may in some cases be available to a wider range of financial institutions, on more whether to extend swap lines to Brazil, Singapore, Mexico, Korea, and Ice- land). 225 See RCAP on Consistency: Jurisdictional Assessments, BANK FOR INT’L SETTLEMENTS, http://www.bis.org/bcbs/implementation/rcap_jurisdic- tional.htm [perma.cc/Q8SQ-M3G7]. See also Silla Brush et al., Bank Regu- lators Face Santiago Showdown on Capital Reform, BLOOMBERG (Nov. 17, 2016, 7:01 PM), https://www.bloomberg.com/news/articles/2016-11- 18/bank-regulators-stare-across-capital-divide-in-santiago-showdown [perma.cc/P5XE-AQCV]; Alex Barker, Jim Brunsden, & Martin Arnold, EU to Retaliate Against US Bank Capital Rules, FIN. TIMES (Nov. 21, 2016), https://www.ft.com/content/26078750-b003-11e6-a37c-f4a01f1b0fa1. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 1003 favorable terms, and subject to less onerous regulatory con- straints, than in the jurisdiction that actually issues the rele- vant currency. Closely tied to the absence of qualifications and conditions is the contemplated fee structure of the swap transactions. There are at least two ways to view the economic function of the standing facilities. The first is as a line of credit. Unlike most privately negotiated lines of credit,226 however, central banks with access to the swap lines are not charged a ‘com- mitment fee’ in exchange for the option of obtaining FCLA. Instead, central banks accessing this liquidity support are simply charged interest on any outstanding balance. This in- terest is calculated on the basis of a spread above a benchmark short-term interest rate such as the overnight index swap (“OIS”) rate.227 While comprehensive data is not available, the reported spreads on swap transactions under the temporary swap lines in operation between 2007 and 2013 ranged from 50–100 basis points (0.5–1.0%) above the OIS rate.228 The second way to view the facilities is as a form of insur- ance against currency-specific liquidity problems. Notably, however, there is no evidence that the methodology for calcu- lating interest reflects the value of this insurance to partici- pating central banks on the basis of, for example, the out- standing stock of the relevant currency, its use by financial institutions within the relevant jurisdiction, or the probability of a market wide currency-specific liquidity shortage. As the Bank of England has observed, the swap transactions are in- stead priced “at a cost which is not attractive during normal 226 And, as this Article shall discuss, the IMF’s new emergency lending facilities. 227 See e.g., U.S. Dollar-British Pounds Swap Agreement, supra note 216, § 4(c). 228 Experience with Foreign Currency Liquidity Providing Central Bank Swaps, MONTHLY BULL. (Eur. Cent. Bank, Frankfurt, Ger.), Aug. 2014, at 73, 65–82; Denbee et al., supra note 51, at 12. AWREY – FINAL 1004 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 periods, but is not prohibitively expensive during crisis peri- ods.”229 Whichever view one adopts, it thus seems highly prob- able that the current fee structure of the standing facilities will result in significant underpricing of FCLA. The final feature relates to the contemplated term struc- ture of the swap transactions. The agreements state that each swap will have a maximum maturity of eighty-eight days, or such other period as may be mutually agreed by the parties.230 However, in the event that one of the parties fails to repur- chase its currency on or before the relevant maturity date, the swap transaction automatically rolls over into a new over- night swap transaction.231 The amount owing under the ma- turing swap is then cancelled and replaced with this rollover swap transaction, which for the purposes of the agreements is expressly deemed to be equivalent to a newly entered-into swap transaction.232 As a result, should the party also fail to repurchase its currency upon the maturity of a rollover swap transaction, the swap will once again automatically rollover. In theory, therefore, these swap transactions could roll over in perpetuity.233 These features reveal two critical flaws in the design of the standing facilities. The first is the complete absence of mech- anisms designed to constrain the potentially significant moral hazard problems generated by the theoretically open-ended, underpriced provision of FCLA. In the absence of such mech- anisms, participating central banks may come to expect that they will be able to acquire foreign currency under these facil- ities at any time and on favorable terms—thereby undercut- ting their incentives to strictly regulate the issuance of short- 229 Denbee et al., supra note 51, at 11. 230 See, e.g., U.S. Dollar-British Pounds Swap Agreement, supra note 216, § 1(c). 231 Id. § 6(a). 232 Id. 233 While the party providing FCLA under the swap is authorized to set off any obligations it owes against the corresponding currency purchased from the other party, id. § 6(b), one might expect this right to be exercised in only the most extraordinary of circumstances given the potential political ramifications of exercising this option. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 1005 term foreign currency liabilities by domestic financial institu- tions or proactively manage the risk of foreign currency liquid- ity problems. These financial institutions, in turn, may main- tain—or even increase—their short-term foreign currency borrowing with the expectation that the central bank will step in to provide FCLA during periods of widespread market dis- ruption. In this important respect, the incomplete institution- alization of FCLA under the standing facilities may actually serve to foment potential financial instability. The second flaw resides in the politically contingent nature of the standing facilities. Some view this flaw as necessary to ensure that participating central banks are not ex ante com- mitted to provide FCLA in circumstances that would fall out- side the scope of their statutory authority.234 Some even view it as desirable insofar as it introduces a degree of constructive ambiguity, thereby helping to ameliorate potential moral haz- ard problems.235 Ultimately, however—and irrespective of whether we view it as a ‘feature’ or a ‘bug’—the contingent nature of the facilities generates two potentially significant distortions. First, one might expect both participating central banks and the financial institutions under their oversight to respond to any uncertainty around their access to the facilities by expanding their foreign currency assets and liabilities and, more generally, by increasing their levels of integration and interconnectedness with the financial systems and real econ- omies in other participating jurisdictions. By doing so, these central banks and financial institutions would seek to in- crease the likelihood that any domestic instability would spill over into other participating jurisdictions, thereby maximiz- ing the probability that these jurisdictions would deem it in their economic best interests to provide FCLA. Second, the po- litically contingent nature of the facilities opens the door to decisions regarding access to foreign currency liquidity that are motivated by domestic political and other considerations. As the history of financial crises clearly demonstrates, these 234 See, e.g., Eur. Cent. Bank, supra note 228, at 37. 235 Id. at 82. AWREY – FINAL 1006 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 decisions often come into conflict with the objective of promot- ing global financial stability.236 The question thus becomes how to make the standing fa- cilities more credible and effective sources of FCLA. The first step would be to place access to the standing facilities on more certain legal footing. Some observers have gone as far as sug- gesting that the mandate of the Federal Reserve should be for- mally amended to reflect its de facto role as global lender of last resort.237 A related, and ultimately complementary, pro- posal would be to introduce formal, binding, ex ante qualifica- tions for jurisdictions seeking access to FCLA. These qualifi- cations could reflect the size of a jurisdiction’s domestic economy, its structural interconnectedness with the global fi- nancial system, whether it was a major international financial center, and whether it otherwise posed a threat to the finan- cial stability of the jurisdiction providing FCLA.238 Once qual- ified, the extension of FCLA to these jurisdictions could then be made contingent on observable and objective metrics of market disruption or financial instability. Potentially useful metrics include basis swap spreads (which, as we have seen, are a barometer for foreign currency funding pressures) and the LIBOR-OIS spread (a key barometer of perceptions of counterparty credit risk within wholesale funding markets). Where these spreads exceed specified thresholds, this could be 236 For examples from the 1920s and 1930s, see LIAQUAT AHAMED, LORDS OF FINANCE: THE BANKERS WHO BROKE THE WORLD (2009). Over its history, even the Federal Reserve has not been immune to oscillations in its approach towards the maintenance of international financial stability. See Peter Conti-Brown & David Zaring, Foreign Affairs and the Federal Reserve, (American Soc’y of Int’l Law Int’l Econ. Law, Working Paper, 2016) (on file with author) (describing the Fed’s oscillation between a “cosmopolitan” and “isolationist” approach to foreign affairs). 237 See STEIL & HINDS, supra note 10, at 246 (“The best hope for salvag- ing financial globalization, then, is a renewed statutory framework for the Fed, one which explicitly acknowledges the global role of the dollar and the dependence of the U.S. economy on foreign confidence in it.”). 238 These factors were notably taken into consideration by the Federal Open Market Committee in determining which jurisdictions would receive central bank swap lines during the crisis. See Minutes of the Federal Open Market Committee Meeting, supra note 224, at 10. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 1007 deemed to constitute evidence of financial instability, thereby enabling the Fed and other participating central banks to ex- tend FCLA in accordance with their statutory mandates. In exchange for this more certain commitment to provide FCLA, the second step would be to require qualifying jurisdic- tions to agree to a binding framework governing the terms upon which FCLA would be extended to domestic banks and other financial institutions. The framework could articulate, for example, which types of financial institutions would be el- igible to receive FCLA, the categories of collateral that could be pledged, and the haircuts imposed on each category of col- lateral. Qualifying jurisdictions could also undertake to en- sure that short-term foreign currency liabilities issued by do- mestic financial institutions were subject to the same reserve ratios as the equivalent liabilities in the jurisdiction that is- sued the relevant currency. They could also agree to impose currency-specific liquidity coverage ratios designed to ensure that domestic financial institutions held a sufficient stock of high quality liquid assets denominated in the relevant cur- rency to be able to meet expected outflows in the event of a hypothetical thirty-day stress scenario.239 The primary thrust of these requirements would be to eliminate the distortions created by substantive differences in ELA and prudential reg- ulatory regimes across jurisdictions, constrain the build-up of foreign currency mismatches, and alleviate the pressure on fi- nancial institutions to sell foreign currency assets during pe- riods of market disruption.240 The third step would be to introduce a fee structure that better reflected the role of the standing facilities as a form of liquidity insurance. The role of FCLA as a form of insurance 239 Indeed, the European Systemic Risk Board has encouraged the com- petent authorities within EU member states to monitor liquidity risks asso- ciated with foreign currency assets and liabilities. See Recommendation of the European Systemic Risk Board of 22 December 2011 on US Dollar De- nominated Funding of Credit Institutions, 2011 O.J. (C 72) 1, 2. 240 See Jeremy C. Stein, Member, Bd. of Governors of the Fed. Reserve Sys., Remarks at the Global Research Forum, Int’l Finance and Macroeco- nomics: Dollar Funding and Global Banks (Dec. 17, 2012); Denbee et al., supra note 51, at 25. AWREY – FINAL 1008 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 suggests that all qualifying jurisdictions should be required to pay an ongoing fee—a premium—irrespective of whether they actually draw down on the facilities. This premium would the- oretically need to reflect the use of the relevant currency within the domestic economy of a qualifying jurisdiction, the level and proportion of short-term foreign currency funding within its domestic financial system, and any pronounced for- eign currency mismatches on the balance sheets of domestic banks and other financial institutions. It would also need to reflect any other forms of (self-)insurance—e.g., the accumu- lation of foreign exchange reserves—undertaken by a qualify- ing jurisdiction for the purpose of insulating the domestic fi- nancial system from foreign currency liquidity problems. While accurately pricing this type of insurance is notoriously difficult, some reasonable attempt at doing so is ultimately necessary in order to constrain the moral hazard problems stemming from the underpriced provision of FCLA under the current standing facilities. Finally, the standing facilities would need to be augmented by binding commitments on the part of qualifying jurisdic- tions to share data relating to foreign currency assets and lia- bilities within their domestic financial systems. This data sharing would be vital for the purposes of both monitoring compliance with ex ante qualifications and ex post constraints and calculating the requisite liquidity insurance premiums. More generally, enhanced data sharing would also enable more effective macroprudential surveillance—especially with respect to cross-border capital flows and the build-up of poten- tially destabilizing foreign exchange imbalances on the bal- ance sheets of banks and other financial institutions.241 The second set of arrangements for the provision of FCLA are two emergency lending facilities recently introduced by the IMF. Established in 1945, the IMF is an international re- serve fund originally designed to provide short-term loans to assist member countries facing actual or potential balance of 241 FARHI ET AL., supra note 195, at 48. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 1009 payments problems.242 IMF member countries are required to make contributions to the fund based on a quota formula that takes into account each country’s gross domestic product (“GDP”), the openness and variability of its economy, and ac- cumulated foreign exchange reserves.243 The IMF then uses these funds to make loans to member countries under various lending facilities.244 The IMF also conducts surveillance of the international monetary system, along with the fiscal, mone- tary, and financial sector policies of each of its 189 member countries.245 Following the collapse of the Bretton Woods fixed exchange rate regime in the early 1970s, the IMF has at- tempted to reinvent itself as an international financial crisis “firefighter.”246 Perhaps most notably, the IMF has played an important role in coordinating the international response to the emerging market debt crisis of the early 1980s247, the 242 For an overview of the IMF’s role, see The IMF at a Glance, IMF (Apr. 20, 2017), http://www.imf.org/en/About/Factsheets/IMF-at-a-Glance [perma.cc/DH9D-ETRB]. For a more detailed description of its activities, see Annual Report of the Executive Board, IMF, http://www.imf.org/exter- nal/pubs/ft/ar/index.htm [perma.cc/8CU5-WXEV]. 243 For further details, see Report of the Executive Board to the Board of Governors on the Reform of Quota and Voice in the International Monetary Fund, IMF (Mar. 28, 2008), https://www.imf.org/exter- nal/np/pp/eng/2008/032108.pdf [perma.cc/Q8SY-S67T]. 244 For a complete list of IMF lending facilities, see IMF Lending, IMF (Oct. 11, 2017), http://www.imf.org/en/About/Factsheets/IMF-Lending [perma.cc/8FM7-8RFH]. 245 Alongside this surveillance, the IMF also provides technical assis- tance to member countries in relation to various aspects of economic policy. Id. For a useful description of the IMF’s role within the broader context of the global financial and regulatory architecture, see generally Chris Brum- mer, How International Financial Law Works (And How It Doesn’t), 99 GEO. L.J. 257 (2011). 246 See Wanted: Chief Firefighter, THE ECONOMIST (June 2, 2011), http://www.economist.com/node/18772058 [perma.cc/5GYY-SBXF]. 247 See James M. Boughton, From Suez to Tequila: The IMF as Crisis Manager 17–18 (IMF, Working Paper No. 97/90, 1997), https://www.imf.org/external/pubs/ft/wp/wp9790.pdf [perma.cc/9NQL- KLBB]. AWREY – FINAL 1010 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 Asian financial crisis of 1997–1998248 and, more recently, the rescues of Ireland, Greece, Portugal, and Cyprus during the European sovereign debt crisis.249 Historically, the IMF has not been an important source of FCLA. However, spurred by the global financial crisis, the IMF has recently introduced two new emergency lending fa- cilities: the Flexible Credit Line (the “FCL”) and the Precau- tionary and Liquidity Line (the “PLL”).250 Introduced in March 2009, the FCL is designed to provide emergency liquid- ity assistance to member countries exhibiting strong economic fundamentals and policy frameworks.251 A member country’s access to the FCL is subject to certain ex ante qualifications based on its external and capital account position, access to international sovereign debt markets, the health of its public finances, the effectiveness of financial sector supervision, and other factors.252 Once qualified, member countries are entitled 248 See IMF, RECOVERY FROM THE ASIAN CRISIS AND THE ROLE OF THE IMF (2000). 249 For a summary of the IMF’s recent activities in Europe, see The IMF and Europe, IMF (Mar. 23, 2016), http://www.imf.org/exter- nal/np/exr/facts/europe.htm [perma.cc/6EJD-2RGQ]. 250 The FCL and PLL are not the IMF’s first attempt to introduce emer- gency liquidity facilities. Previous (failed) attempts have included the Con- tingent Credit Line (launched in 1999) and the Short-Term Liquidity Facil- ity (launched in 2008). For a summary of the development of, and debates surrounding, these earlier initiatives, see generally Roberto Marino & Ul- rich Volz, A Critical Review of the IMF’s Tools for Crisis Prevention (Ger. Dev. Inst., Paper No. 4/2012, 2012). The IMF has also recently introduced the Rapid Financing Instrument (“RFI”) for member countries facing urgent balance of payment problems; see The IMF’s Rapid Financing Instrument (RFI), IMF (Oct. 13, 2017), http://www.imf.org/en/About/Fact- sheets/Sheets/2016/08/02/19/55/Rapid-Financing-Instrument [perma.cc/Y6ZX-AZP6]. However, RFI loans are capped at 37.5% of a mem- ber country’s quota on an annual basis and 75% on a cumulative basis. RFI loans are also subject to same terms as the FCL and PLL. Id. 251 See The IMF’s Flexible Credit Line (FCL), IMF (Oct. 5, 2017), http://www.imf.org/en/About/Factsheets/Sheets/2016/08/01/20/40/Flexible- Credit-Line [perma.cc/89MW-6U8P]. 252 Id. At the same time, the IMF has signaled that member countries do not need to show strong performance against all these criteria. See IMF, THE FLEXIBLE CREDIT LINE—GUIDANCE ON OPERATIONAL ISSUES (2009), AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 1011 to draw on the FCL at any time during the term of the facility without the need to satisfy further conditions. The amount that member countries can borrow is theoretically unlimited, but ultimately subject to the general constraints on the IMF’s financial resources. Member countries that qualify for access to the FCL must pay a commitment fee and, when drawn upon, an interest rate that varies with the amount borrowed and the time that borrowed funds remain outstanding.253 To date, the only countries that have applied for access to the FCL are Mexico, Poland, and Colombia—with none of these countries having thus far drawn down under this new facility. Introduced in November 2011, the PLL is designed to pro- vide emergency liquidity assistance to member countries that do not meet the ex ante qualifications for access to the FCL.254 In addition to meeting a less stringent set of qualifications, member countries borrowing under the PLL are subject to ex post conditions reflecting the IMF’s assessment of their poten- tial vulnerabilities.255 Compliance with these conditions is then monitored and assessed by the IMF Executive Board on a semi-annual basis. Once qualified, member countries can borrow up to 250% of their quota per year, depending on the duration of the loan and the source and severity of the eco- nomic shock precipitating the need for emergency liquidity as- sistance.256 Qualified countries must pay a commitment fee, plus interest and service charges on any drawn amounts.257 http://www.imf.org/external/np/pp/eng/2009/110209.pdf [perma.cc/35CV- KXX3]. 253 Drawn amounts are also subject to a service charge of fifty basis points. For a more detailed description of the methodology for calculating the commitment fee, interest rate, and service charge, see The IMF’s Flexi- ble Credit Line (FCL), supra note 251. 254 The PLL replaced the Precautionary Credit Line introduced in Au- gust 2010. See The IMF’s Precautionary and Liquidity Line (PLL), IMF (Oct. 10, 2017), http://www.imf.org/en/About/Factsheets/Sheets/2016/08/ 01/20/45/Precautionary-and-Liquidity-Line [perma.cc/K9QK-5FP2]. 255 Id. 256 Id. 257 Id. AWREY – FINAL 1012 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 To date, the only countries to have applied for access to the PLL are Macedonia and Morocco.258 While still essentially untested, the FCL and PLL repre- sent an important step in the development of a credible mul- tilateral mechanism for the provision of FCLA. Perhaps most importantly, the design of these facilities reflects a discerna- ble shift in the IMF’s focus from financial crisis firefighting toward meaningful fire prevention. This shift is most clearly reflected in the FCL’s utilization of ex ante qualifications as opposed to ex post conditionality. The use of ex ante qualifica- tions injects a degree of certainty from the perspective of po- tential creditor countries that they will have immediate and unconditional access to the facility in the event of an emer- gency. The absence of ex post conditionality also eliminates an important source of potential procedural obstacles to the rapid extension of FCLA. While the introduction of these facilities is undoubtedly a positive development, the FCL and PLL are still a far cry from representing a truly credible framework for the provision of FCLA. First, only five countries—together representing less than 3% of global GDP—have thus far applied for access to these facilities.259 Second, unlike conventional ELA facilities, any loans ultimately extended under the FCL and PLL will not be collateralized against the assets of either the borrowing countries or the financial institutions that ultimately receive FCLA. This is likely to exacerbate potential moral hazard problems and leaves the IMF member countries that provide financing under these facilities exposed to potentially signifi- cant losses in the event of default. Third, and perhaps most importantly, the financial re- sources currently available under the FCL and PLL may simply be too small to mount an effective full-scale response to a systemic foreign currency liquidity crisis. As of March 2017, IMF quotas totaled approximately $USD645 billion.260 258 Id. 259 See The IMF’s Flexible Credit Line (FCL), supra note 251; The IMF’s Precautionary and Liquidity Line (PLL), supra note 254. 260 See The IMF at a Glance, supra note 242. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 1013 Subject to the approval of member countries representing at least 85% of the fund’s resources, the IMF can also borrow up to $USD250 billion from member countries under an existing multilateral facility known as the ‘New Arrangements to Bor- row’ (the “NAB”). However, with 17.47% of the IMF’s total quota (and 16.54% of the voting rights of the Board of Gover- nors), this threshold notably gives the United States a de facto veto over the use of the NAB.261 Furthermore, it is not clear whether any countries other than the United States would be in a position to lend the IMF U.S. dollars in the middle of a global liquidity crisis. After subtracting current commitments of approximately $USD159 billion, a conservative estimate thus leaves the IMF with just over $USD500 billion of dry powder theoretically available in response to an emerging global crisis. Yet only a small fraction of this $USD500 bil- lion—approximately $USD102.6 billion as of January 31, 2017—is comprised of the U.S dollars that are likely to be most in demand in the thick of any such crisis.262 By way of comparison, it worth remembering that U.S. dollar lending under the swap lines between the Federal Reserve and other major central banks peaked at $USD586 billion in December 2008. By almost any measure, the IMF is thus not in a position to comply with the imperative of Bagehot’s dictum to lend freely in response to widespread panic and financial instabil- ity. 261 When the requisite 85% supermajority does not approve a proposal to activate the NAB, the IMF can still borrow under the ‘General Agree- ments to Borrow’ (“GAB”). As of April 2017, the IMF was entitled to borrow up to approximately $USD23 billion under the GAB. IMF Standing Borrow- ing Arrangements, IMF (Apr. 2017), http://www.imf.org/en/About/Fact- sheets/Sheets/2016/08/05/17/55/IMF-Standing-Borrowing-Arrangements [perma.cc/YY4F-QLRW]. 262 Figure calculated on the basis of the U.S. total of the IMF’s currency holdings (75,480.9 million SDR) multiplied by the notional SDR-USD ex- change rate as of January 31, 2017 (1.358830). See IMF, FINANCIAL STATE- MENTS FOR THE QUARTERS ENDED JANUARY 31, 2017, AND 2016 (2017), http://www.imf.org/External/Pubs/FT/quart/2017fy/013117.pdf [perma.cc/PEM4-SRMX]; Currency Units per SDR for January 2017, IMF, https://www.imf.org/external/np/fin/data/rms_mth.aspx?SelectDate=2017- 01-31&reportType=CVSDR [perma.cc/ETY9-JHQ2]. AWREY – FINAL 1014 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 Given these resource constraints, Emmanuel Farhi, Pierre-Olivier Gourinchas, and Hélène Rey propose granting the IMF the right to borrow from national treasuries and cen- tral banks and to raise capital on international bond mar- kets.263 In order to address the moral hazard problems asso- ciated with the FCL’s use of ex ante qualifications (and the absence of ex post conditionality), they also propose subjecting borrowers to ongoing monitoring of private sector foreign ex- change imbalances over the term of any loan. Both of these proposals seem prudent. Ultimately, however, the effective- ness of these proposals hinges on the IMF using its new pow- ers proactively to raise capital under normal market condi- tions rather than waiting for a global liquidity crisis to appear on the horizon.264 Moreover, as Farhi, Gourinchas, and Rey concede, these reforms would likely need to be considered as a part of a far broader discussion regarding the governance of the IMF.265 At least for the moment, this discussion does not appear to be high on the IMF’s agenda. In the absence of an external catalyst, the FCL and PLL are thus likely to remain incomplete and imperfect arrangements for the provision of FCLA. Are the proposed reforms to either the central bank swap lines or the IMF’s new emergency lending facilities feasible in the current political climate? Perhaps not. Indeed, engender- ing widespread political support for these reforms may prove especially challenging precisely because they envision the widening and deepening of the international financial safety net. Nevertheless, the vagaries of domestic and international politics do not detract from the reality that there are no free lunches in global finance, and that the drive for deeper eco- nomic and financial integration will eventually need to be matched by more credible and effective commitments in the realm of international financial policy, regulation, and crisis 263 FARHI ET AL., supra note 195, at 41. 264 By definition, raising capital in these circumstances would be more expensive or, potentially, impossible. See supra Section III.B. 265 FARHI ET AL., supra note 195, at 43. AWREY – FINAL No. 3:934] FOREIGN CURRENCY LIQUIDITY ASSISTANCE 1015 management. Until they are prepared to make these commit- ments, policymakers will continue to face a choice between au- tarky and the instability generated by foreign currency liquid- ity crises. VI. CONCLUSION There are a great many lessons from the global financial crisis. This Article has identified some of the key lessons from an extremely important but relatively unexamined episode from the crisis: the international U.S. dollar liquidity shortage that began in December 2007 and reached its apex in Septem- ber and October 2008. Two lessons stand out. The first is that the widespread issuance of short-term foreign currency liabil- ities by banks and other financial institutions should be viewed as a form of private money creation that poses a grave threat to financial stability. The second is that, within a glob- ally integrated financial and economic system, there are sig- nificant practical constraints on the ability of sovereign states to effectively respond to this threat. It is perhaps tempting to view the network of central bank swap lines established during the financial crisis, along with the IMF’s new emergency lending facilities, as evidence that society has learned the important lessons from this episode. However, while there is little doubt that the swap lines played a critical role in stabilizing the financial system at the height of the crisis, their current design gives rise to significant dis- tortions, thereby undermining their likely effectiveness as a comprehensive and durable solution to foreign currency li- quidity problems. The new IMF emergency lending facilities, meanwhile, though undoubtedly a constructive step toward a viable multilateral facility for the provision of FCLA, are also far from perfect and remain essentially untested. This Article has sketched out a blueprint for how one might improve the effectiveness of this emerging international ar- chitecture for the provision of FCLA. This blueprint hinges on strengthening the credibility of the commitments underpin- ning FCLA, eliminating the distortions associated with the underpriced provision of liquidity insurance under the current facilities, and enhancing cross-border data sharing and AWREY – FINAL 1016 COLUMBIA BUSINESS LAW REVIEW [Vol. 2017 macroprudential surveillance arrangements. While largely in- cremental, these reforms could serve to dramatically reduce the prospect of widespread financial instability stemming from foreign currency liquidity problems. Lastly, this Article has raised a host of important and—at least for the moment—unanswered questions. Paramount amongst these questions is how well the current patchwork architecture for the provision of FCLA will perform during the next global financial crisis. While there is reason for cautious optimism, there is also enormous scope for political interfer- ence, coordination problems, and policy failure. From an em- pirical perspective, other important questions include whether and how the current architecture might serve to dis- tort international capital flows, along with how these distor- tions might contribute to foreign exchange imbalances or fo- ment potential financial instability. From a policy perspective, important questions then include whether these distortions are best addressed through the development of bilateral ar- rangements such as central bank swap lines or multilateral arrangements such as the IMF’s emergency lending facilities. The answers to these questions remain to be explored in sub- sequent work and, ultimately, will play an important role in building a more sustainable international financial system. gorton