eichengreen rivalry and crisis

CHAPTER 5
CRISIS
Stability is the sine qua non of a currency that is widely used in international transactions. Whether they rely on it as a means of payment, unit of account or store of value, a currency’s stability is the first thing to which exporters, importers, and investors all look. Nothing, it follows, can more seriously damage the regard in which a currency is held than a full-blown financial crisis—that of course being precisely what the United States experienced in 2008–2009.
It is entirely natural, therefore, that in the wake of the crisis questions should have been asked about the dollar’s international role. The assertion that the United States had a comparative advantage as an originator of high-quality financial assets came to be dismissed as a joke. The belief that the complex financial instruments retailed by U.S. institutions were as reliable as treasury bonds was shown to be false. The immensely large budgetary costs of digging the economy out of the hole created by the crisis raised suspicions that the Fed might seek to inflate away the debt, especially that part held by foreigners. All this pointed to the possibility that the crisis was a turning point. It could prompt a large-scale migration away from the dollar on the part of importers, exporters, and foreign investors alike. And the existence now of alternatives such as the euro seemed to suggest that the possibility was more than hypothetical.
But there were also those who suggested that all this dollar doom and gloom was overdone. Each time the crisis reached new heights—first in July 2007 with Bear Stearns’s liquidation of two in-house hedge funds, then with the collapse of Bear in March 2008, and finally with the bankruptcy of the investment firm Lehman Brothers the following September—the dollar strengthened against the euro and other currencies. The dollar was still the ultimate safe haven for frightened investors around the world. And if it could survive these events with its international role intact, it could survive anything. Or so its defenders insisted.
Making sense of their arguments and determining whether they are right require understanding not just the impact of the crisis but also the role of the dollar’s exorbitant privilege in bringing it about.
ROOTS OF THE CRISIS

At the root of the crisis lay financial irregularities unchecked by adequate regulation. Banks outsourced the origination of mortgage contracts to specialized brokers. Often these individuals had little professional training, there being no meaningful federal or in some cases even state licensing requirements. In many states, brokers had no fiduciary responsibility to their so-called clients, the homeowners they sometimes all but dragooned into signing contracts. This meant that the broker bore no legal responsibility if a homeowner somehow, just somehow, misunderstood the terms and ended up unable to pay. And the broker suffered no financial consequences, having been paid his fee and passed the mortgage along to the bank that was in effect his real employer.
Banks then packaged batches of these contracts into residential-mortgage-backed securities or passed them on to other banks and special-purpose vehicles able to do so. The resulting securities were then sold not just to other banks but to pension funds, mutual funds, and other investors. There being a limited market for bonds backed by some of these securities, they were then repackaged as collateralized debt obligations (CDOs). The holders of these even more complex instruments had different claims on the income streams associated with the underlying mortgage-backed securities. The typical CDO was separated into a senior tranche, holders of which got paid first out of the income on the underlying mortgage-backed securities; a mezzanine tranche, holders of which got paid second; and the disarmingly named “equity” tranche, holders of which got paid last if at all. The next step was then CDOs squared, CDOs made out of other CDOs. Next were CDOs cubed.
This, argued the investment banks that earned generous fees for slicing, dicing, and repackaging the mortgage-backed securities in question, was all a way of more efficiently separating out risks and shifting them to those with the right risk-bearing ability. The reality, we now know, was different. This complex process was in fact a way of disguising risks rather than simply assembling them into more efficient bundles. It ended up shifting risks from those with more risk tolerance to others with less risk tolerance—to, say, the Missouri State Employees’ Retirement System and the Teachers Retirement System of Texas—who often had little idea of the attributes of the assets they were buying.1 But all this is wisdom after the fact. At the time, the riskiness of these instruments was inadequately appreciated by the wider investment community. It was not something that the originating investment banks were inclined to advertise. Nor were they obliged to do so since they did their work in the absence of meaningful regulatory oversight.
It was also a business model that would not have been viable without help from confederates, starting with the rating agencies. Pension funds and mutual funds, whose covenants limited the amount of risk they were permitted to take on, needed investment-grade ratings from Standard & Poor’s and Moody’s to purchase CDOs.2 The rating agencies rated the CDOs, but they also advised their originators how to structure them so that the senior tranche would win an AAA rating. For this latter activity they earned handsome fees. It is easy to imagine how their employees would have felt pressure to confer the expected rating. The rating agencies deny that they were subject to conflicts of interest. One wonders.
But the originator still might find few willing buyers of the speculative equity tranche. Not infrequently the originating bank ended up having to hold it on its own balance sheet. This was a constraint on expanding the business. A solution was then found in the form of insurance to further “enhance” the securities. For a fee, the risk of default could be transferred to another entity. Suitably insured, some portion of the equity tranche could then be sold off to other investors. The mechanism for obtaining this insurance, once obscure, was the now notorious contract known as a credit default swap. And the leading underwriter of these insurance contracts was the thinly capitalized American International Group (AIG). Exactly what those responsible for decision making at AIG were thinking will forever be a mystery. Be this as it may, they, too, took their decisions in the absence of meaningful regulatory oversight.3
Savvy investors, starting with the investment banks themselves, understood that holding the equity tranche was risky. There would be losses if the housing market turned down. The credit default swap providing the insurance might not be worth the paper it was written on if the provider, or counterparty, got into trouble. Investment banks presumably did not anticipate the size of the hole that CDOs could blow in their balance sheets. But they probably did anticipate earning low returns on this part of their portfolios. Their response was to attempt to boost the return on capital by expanding their balance sheets. In other words, they used more borrowed money—in some cases, much more borrowed money—to maintain the now customary profit margin. From this flowed the explosive growth of leverage—the ratio of borrowed funds to own capital—of bank and nonbank financial institutions.
This was not a phenomenon limited to the investment banks at the center of the crisis. Other financial institutions not so deeply invested in residential-mortgage-backed securities, CDOs, and other “sophisticated” financial instruments also levered up. Behind this was the growth of a wholesale money market on which financial institutions could borrow large sums for periods as short as overnight. There was also the willingness of regulators to look the other way. In particular, investment banks, which once gambled only their partners’ money but now gambled the money of others, and their conduits and special-purpose vehicles remained largely outside the regulatory net.
ASLEEP AT THE SWITCH

But even commercial banks were permitted to take on more leverage. Standard regulatory practice dictated requiring a commercial bank to hold core capital—its shareholders’ own funds—equal to 8 percent of the bank’s investments as a cushion against losses. Now the regulators, in their wisdom, allowed them to substitute less liquid instruments for shareholders’ common equity. Commercial banks were permitted to substitute so-called hybrid instruments and junior debt, the holders of which get paid just before equity holders but after other claimants. These not so liquid instruments were known as “Tier 2 capital” to distinguish them from the funds of bank shareholders, so-called Tier 1 capital. Under the Basel Accord, the agreement on capital adequacy negotiated by the Basel Committee on Banking Supervision (the committee of national regulators that meets at the Bank for International Settlements, the bankers’ bank in Basel), banks were permitted to hold as little as 2 percent common equity as a share of risk-weighted assets.
As a result, commercial banks had less capital to cushion themselves against losses. They had fewer reserves out of which to pay their debts if things went wrong. Again, the regulators averted their eyes. Even worse, they bought into the idea that the banks were now capable of more efficiently managing their risks, justifying the substitution of cheaper and less liquid forms of capital. Internal models of the riskiness of banks’ activities and, where they were too small or “backward” to possess them, commercial credit ratings were used to place assets into different categories, inelegantly called “buckets,” according to their risk.
The distinction between Tier 1 and Tier 2 capital was incorporated into the Basel Accord on capital adequacy signed onto by the so-called financially advanced countries in 1988. This is a first hint, then, that the trends in question had been under way for some time and were not limited to the United States. In fact, many of those trends, from excessive leverage to the tendency for regulators to buy into the arguments of the regulated, infected the banks and financial systems of other countries, from Germany to the United Kingdom. For European banks, the American model of minimizing capital and using high levels of leverage was something to be emulated, not scorned. Ultimately, common global tendencies produced a common global crisis.
This securitization machine, itself almost as complex as the securities it spit out, had a voracious appetite for fuel. With leverage rising, portfolios expanding, and investors stretching for yield, it required extensive inputs of high-yielding securities. This need in turn encouraged the creation of more CDOs and residential-mortgage-backed securities, which in turn encouraged the origination of more mortgages. Banks loosened their credit and documentation standards. Mortgage brokers moved down the credit-quality spectrum in search of borrowers. It was an elaborate dance, although not one in which all participants were fully in touch with their partners.
To be sure, the rapid growth of subprime mortgages involved more than just this financial legerdemain. But the originate-and-distribute model, and the perverse incentives it created, was an important factor in what started out as the now quaintly sounding “subprime crisis” and developed into the most serious global credit crisis in 80 years. And as for what enabled all involved to act on those incentives, the answer is simple: inadequate regulation.
Once the dominos were lined up, just one had to be toppled to bring them all tumbling down. The first domino was the residential property market, which peaked in 2006. The second, starting in 2007, was losses for specialized investment funds invested in complex securities backed by subprime mortgages.4 As CDO prices fell, investors received collateral calls and were forced to sell other securities. Declines in the prices of those securities then forced still other investors to sell. Before long, a full-fledged fire-sale was under way.
Suddenly aware of the risks, banks drew in their horns. In precisely those parts of the financial system where leverage was greatest, deleveraging now occurred with a vengeance. And precisely those financial institutions like Bear Stearns that had funded themselves most aggressively on the wholesale market now found themselves with collateral calls they could not meet.
Faced with extraordinary uncertainty, spenders stopped spending, plunging the economy into a tailspin and causing banks to be hit by problems on previously sound loans and investments. Regulators, finally roused from their slumbers, responded unpredictably when deciding whom to save (AIG) and whom to sacrifice (Lehman Brothers). By the final months of 2008, the situation had degenerated into a full-blown financial crisis and set the stage for the deepest recession since World War II.
DIGGING DEEPER

So far, so good. But this account begs as many questions as it answers. It begs the question of who permitted the development of an immensely large and dangerous market in complex securities. It begs the question of why banks were allowed to use their own models to gauge the riskiness of their investments and determine, essentially for their own convenience, the size of the capital cushion to be held against them. And it begs the question of how it was that the regulators remained asleep at the wheel.
The key players in the run-up to the crisis firmly positioned themselves as believers in letting markets work. In particular, they were believers in letting derivatives markets work on the grounds that they were efficient mechanisms for redistributing risk. Larry Summers, the Harvard professor who served as undersecretary, deputy secretary, and secretary of the treasury in the Clinton years, was convinced, in his incarnation as an official, that derivatives “serve an important purpose in allocating risk by letting each person take as much of whatever kind of risk he wants,” as his views were described by Robert Rubin, his political mentor and predecessor as treasury secretary.5 Summers himself put it more technically but no less unequivocally. “By helping participants manage their risk exposures better and lower their financing costs, derivatives facilitate domestic and international commerce and support a more efficient allocation of capital across the economy. They can also improve the functioning of financial markets themselves by potentially raising liquidity and narrowing the bid-asked spreads in the underlying cash markets. Thus, OTC [over-the-counter] derivatives directly and indirectly support higher investment and growth in living standards in the United States and around the world.”6 Summers the academic had been more skeptical about the efficiency of markets, but that kind of iconoclasm did not transfer easily to the policy domain.7
Rubin himself was more circumspect, having managed the fixed-income division at Goldman Sachs, which traded mortgage-backed securities, fixed-income futures, options, and other derivatives. The fixed-income division under Rubin had experienced problems in 1986 as a result of traders not anticipating “unlikely market conditions.”8 Rubin’s traders had placed big bets on the assumption that the prevailing level of interest rates would remain broadly unchanged. When rates dropped unexpectedly, the division took losses of $100 million, a large amount of money at the time.
With his awareness of the tendency for traders to be incapable of imagining the worst, one would have expected Rubin to have been an even stronger proponent of regulating derivatives markets. His autobiography suggests that, with benefit of hindsight, he agrees.9 But hindsight is 20/20. At the time—specifically, in 1998—Rubin, together with Summers and Federal Reserve chairman Alan Greenspan, opposed measures to regulate derivatives trading.
Specifically, they opposed such measures when they were proposed by Brooksley Born, head of the Commodity Futures Trading Commission (CFTC). Born was a formidable opponent. As an undergraduate at Stanford University in the early 1960s, her desire to become a doctor rather than a nurse had been frustrated by a guidance counselor who insisted that this was no occupation for a woman. Born went to Stanford Law School instead, where she was one of only four women in her graduating class, and from there to a leading Washington, D.C., law firm, where she developed its derivatives practice before being appointed to the CFTC in 1995. All this suggests that Born had considerable strength of will. She was not strong enough, however, to prevail over “the Committee to Save the World,” as Rubin, Greenspan, and Summers were dubbed by Time Magazine for their actions following the collapse of the mega hedge fund Long-Term Capital Management in August 1998.
What Born actually proposed was relatively modest: a “concept paper” identifying the risks posed by the growth of unregulated financial derivatives and sketching a framework for regulating them. This was still enough to provoke a furious reaction from Greenspan, Rubin, and Summers, whose saw it as the camel’s nose under the tent. The three opposed CFTC regulation that would have forced derivatives traders to engage in greater disclosure and hold a larger capital cushion against losses, as suggested in Born’s concept paper when it finally appeared. They supported only the creation of a clearinghouse to net transactions in derivatives, something that would have helped to mitigate the problems that arose in 2008 as a result of AIG’s immensely complex party-to-party transactions.10 Even then they supported just a voluntary clearinghouse, not a mandatory one. And a clearinghouse was not something for which dealers like Goldman Sachs that made a profit on each and every derivatives trade were particularly interested in volunteering their support.11
Subsequent treasury secretaries Paul O’Neill and John Snow, coming from business rather than finance, were in no position to rock the boat. Nor were they inclined to do so as long as things were going well. Henry Paulson, also coming as he did from Goldman Sachs, was a bird of a different stripe, but that hardly made him an advocate of hardheaded regulation.
Alan Greenspan, chairman of an institution with considerable responsibility for supervising and regulating financial institutions, was fundamentally a believer that markets knew best. For one normally inclined toward oracular statements, Greenspan put it with uncharacteristic bluntness in testimony to the House Committee on Banking and Financial Services: “Regulation of derivatives transactions that are privately negotiated by professionals is unnecessary.”12 Greenspan opposed regulation of the derivatives market when it was considered by the Congress in 1994. He opposed stricter oversight of derivatives by the CFTC in 1998. In 2003 he told the Senate Banking Committee that it would be a mistake to more extensively regulate derivatives markets. “What we have found over the years in the marketplace,” he asserted, “is that derivatives have been an extraordinarily useful vehicle to transfer risk from those who shouldn’t be taking it to those who are willing to and are capable of doing so.” By October 2008, of course, he was singing a different tune, warning the U.S. House Oversight and Government Reform Committee against relying on “the self-interest of lending institutions” and acknowledging the need for regulation of derivatives markets.13
Greenspan’s successor, Ben Bernanke, as a student of the Great Depression, might have been expected to hold different views. But it was hard, intellectually and politically, to abandon the score handed down by the maestro, especially when the investment community was enjoying such healthy returns. Bernanke also may have had a tendency to overlook the vulnerability of the “shadow banking system,” that is, the hedge funds, conduits, and special-purpose vehicles where so many CDOs and so much housing-related risk were held, since no shadow banking system had existed in the 1930s. Other regulatory agencies, meanwhile, saw their budgets and human resources cut by a President George W. Bush who may not have been a financial sophisticate but knew one thing: regulation was the problem, not the solution.
HERD BEHAVIOR

Yet this blame game, which became understandably popular in the wake of the crisis, assigns too large a role to individuals, even individuals in positions of power. What informed decisions in the run-up to the crisis was not the personal ideology of a few powerful individuals but a powerful collective psychology. A central tenet was the belief that it had become possible, using modern mathematical tools, to more effectively price and manage risk. Elegant mathematical formulae like the Black-Scholes model could be used to determine prices for options and other derivatives. Subject to the simplifying assumptions needed to render the model tractable, it was possible to give a numerical estimate of the maximum loss that would be incurred on an investment in the course of the next day with, say, 99 percent probability. Subject to yet more assumptions about, say, the correlations of returns on different investments, it was possible to characterize the distribution of returns on a bank’s investment portfolio. The maximum loss that might be incurred on that portfolio was soon given its own name, Value at Risk (VaR).
These techniques were not without value. Black-Scholes could be used to detect instances where the price of a complex derivative diverged significantly from the value of its components. It offered profits for those prepared to engage in arbitrage, for example buying the underpriced components while selling, or shorting, the overpriced composite. It provided a business model for entities like the hedge fund Long-Term Capital Management (LTCM), founded by the serial financial entrepreneur John Meriwether and his merry band of Nobel laureates in 1994. Similarly, the practice inaugurated by J.P. Morgan chief executive officer Dennis Weatherstone after the 1987 Wall Street crash, that he should have a “4:15 Report” summarizing the risk of the bank’s investment portfolio on his desk each day within 15 minutes of the market close, was a fundamentally sensible request for a CEO. This was the practice that evolved into VaR, with J.P. Morgan being the institution that developed the methodology and published it, also in 1994.14
The problem was the tendency to push these processes too far. LTCM initially made handsome profits from its arbitrage transactions. Typically they involved buying and selling similar securities, for example U.S. treasury bonds of the same maturity issued on slightly different dates, as a way of exploiting small differences in the prices at which they traded.15 But the more money LTCM took in from investors and the more traders adopted its techniques, the smaller those arbitrage opportunities became. The very success of the model and the drive to maintain profitability encouraged Meriwether and his LTCM partners to use more and more borrowed money to exploit smaller and smaller opportunities. The pseudoscientific nature of the undertaking created excessive confidence about the outcome. When the unlikely happened—Russia defaulted on its foreign bonds in August 1998, and asset prices moved unexpectedly—LTCM was pushed to the verge of bankruptcy, nearly toppling the U.S. financial system.
This should have been a warning shot across the bow of those using mathematical methods in the elusive quest to master risk. But the failure of LTCM was dismissed as an anomaly; big countries like Russia, it was said, do not default every day. If LTCM had overreached, this reflected the overweening ambition and supreme self-confidence of the principals, who included the Nobel laureate Myron Scholes of Black-Scholes fame, and not any intrinsic limitations of their techniques. Complex mathematical formulae thus came to be used even more widely to price even more complex securities. The fact that every quant now had a powerful microcomputer on his desk encouraged the building of ever more complex models, which of course fed the salaries of the quants, the only ones capable of manipulating the models. Banks built more elaborate models of Value at Risk, supported by specialized practitioners—J.P. Morgan having spun off the methodology and its practitioners into an independent company with the confidence-inspiring name RiskMetrics Group in 1998—whose incomes depended on how widely the practice was adopted. The methodology was taught in business schools, giving it scholarly legitimization, and prescribed by newly minted risk-management consultants.
The fact that the models were based on simplifying assumptions, necessarily in order to render them tractable, meant that in the hands of careful practitioners they were never used as more than a starting point for thinking about risks. Careful practitioners similarly understood that the model was fitted to a relatively short series of observations of the prices of certain assets. Information on the prices of complex mortgage-related securities spanned only the period when home prices had gone up, for example, and consequently contained little information about what might happen if prices came down. The problem was that there were few incentives to be a careful practitioner.
In this way the starting point became the end point. What started as a daily dose of self-discipline for a single CEO was applied more widely and mechanically. Banks, confident that they had reduced their risks to a single number, became confident of their ability to shoulder more. Regulators, convinced by the regulated of the reliability of the methodology, allowed VaR to be used as an input into setting capital requirements. The Securities and Exchange Commission, when requiring financial institutions to provide their shareholders more information about the risks they were taking, accepted VaR as a logical summary measure. And banks, seeing how VaR affected the amount of costly capital they had to hold, had an incentive to tweak their portfolios—and the methodology—to produce more favorable estimates of Value at Risk. The sense that risk, having been mathematicized, had been mastered grew pervasive. False confidence encouraged institutions to take on more risk. It encouraged the regulators let them. So long as things turned out as expected, more risk meant more profits. Many people were generously compensated for going along. No one was generously compensated for exercising “undue” caution.
The timing is important here. When disaster struck in 2008, VaR was still less than 15 years old. The microcomputer revolution was still recent; LTCM had been famous for spending lavishly on state-of-the-art workstations, which were thought to give it a leg up on its competitors by allowing it to solve more complicated equations faster. Periods of rapid financial innovation are seedbeds for crisis. They are periods when there has not been sufficient time for innovations to be fully road tested. And with financial practice changing rapidly, it is especially hard for the regulators to keep up. In particular, it is hard for them to keep up with claims by the regulated that they have become more adept at managing risks.
NO MORE COZY LIVING

A further ingredient in this toxic brew was the intensification of competition. The Glass-Steagall Act separating commercial and investment banking and limiting the investment activities of deposit-taking banks was revoked by the Gramm-Leach-Bliley Act of 1999. Commercial banks, now able to more efficiently manage risk, could be entrusted to take on a wider range of investment activities, or so the authors of the bill believed. Defenders of Gramm-Leach-Bliley object that commercial banks were not at the center of the subprime crisis. It was not they but investment banks that originated CDOs and ended up stuck with the risky equity tranche. It was not they but investment banks and broker-dealers like Bear Stearns that were so dangerously leveraged.
But to argue this point is to miss the big picture. As commercial banks branched into new activities, they disturbed the investment bankers’ cozy lives. Seeing their rents competed away, investment banks responded by moving into riskier activities and using more borrowed money in the scramble to survive. It is no coincidence that Bear Stearns, which once upon a time had earned a comfy living charging fixed commissions for stock trades but now saw this business eroded by deregulation and technological innovation (the emergence of discount brokers like Charles Schwab, for example) took the most dangerous gambles involving the highest levels of leverage.
Competitive pressure was further ratcheted up by financial globalization. The changes in technology and practice fostering the belief that banks could manage more risk and use more leverage while requiring less regulation were not limited to the United States. The same logic implying that financial institutions could master more lines of business and invest in a wider class of assets suggested that they could do business in more places. The European Union, as part of its Single Market program, removed all restrictions on the ability of banks to do business in other European countries. More generally, the period saw a sharp intensification of cross-border competition.
And, again, institutions feeling the chill winds of competition took on more risk in the effort to maintain customary profit margins. The British building society Northern Rock levered up its bets by supplementing the deposits of retail customers with money borrowed from other banks. Icelandic banks offered suspiciously high-interest online savings accounts to British, Dutch, and German households to finance risky bets. Sleepy German savings banks took on some of the worst performing U.S.-originated-and-distributed CDOs. Leverage was even higher among European financial institutions than in the United States. Either false confidence was higher, or the intensification of competitive pressure was greater, or both. In Europe, too, regulators averted their eyes.
Gambling to survive—doubling up one’s bets—is not the only conceivable response to an existential threat. The alternative is to hunker down. Hunkering down in this case would have meant shrinking the enterprise. But compensation schemes provided no incentive to respond in this way. Successful bets meant big paydays. If those bets put the firm in an untenable position tomorrow, well, that was someone else’s problem. The need for corporate boards and, failing that, government agencies to regulate compensation to better tie it to the long-term performance of the enterprise is now widely understood. But this was not something on which regulators insisted, or even mentioned, before it was too late.
LIQUID ACCELERANT

If flawed regulation was the spark, then central bank policy was the accelerant. Financial excesses would not have spread so quickly to such destructive effect had the Fed not poured fuel on the fire.
Ironically, the very success of the Fed in stabilizing the economy and, thereby, financial markets may have been a factor in the buildup of risk. The period after Paul Volcker’s conquest of inflation saw a reduction in economic volatility that came to be known as “the Great Moderation.” Whether good policy or good luck was mainly responsible is contested, but it is hard to imagine that the improvement in policy since the G. William Miller years played absolutely no role.16 The same naive belief that the Fed had tamed the business cycle had underlain talk of a “New Era” of stability in the 1920s and fueled an earlier Wall Street boom. A less volatile economy, it had been argued then and was argued again now, meant less volatile financial markets. Institutional investors, indeed investors of all kinds, felt more confident about taking on risk. As Donald Kohn, vice chairman of the Federal Reserve Board, put it in 2008, not long after things went south, “In a broader sense, perhaps the underlying cause of the current crisis was complacency. With the onset of the ‘Great Moderation’ back in the mid-1980s, households and firms in the United States and elsewhere have enjoyed a long period of reduced output volatility and low and stable inflation. These calm conditions may have led many private agents to become less prudent and to underestimate the risks associated with their actions.”17 The very success of the Fed at becoming more transparent about its intentions and reducing uncertainty about the future may have been a prime factor in the development of this complacency.
Even if one is skeptical that policy was responsible for the decline in economic volatility and, through this channel, for complacency on the part of investors, there was always the belief that the Fed under Greenspan would intervene to put a floor under asset prices in order to prevent a destabilizing crash. Virtually the first act of Chairman Greenspan had been to cut rates following the 1987 Wall Street crash. The Fed cut again following the LTCM debacle in 1998. By this time “Greenspan put”—the idea that the Greenspan Fed was ready to effectively guarantee a minimum level of asset prices—had become part of the financial lexicon.
But the idea crystallized when the Fed cut rates to 1 percent following the collapse of the tech bubble and Greenspan spoke in 2002 on the subject of asset prices and monetary policy.18 The chairman argued that asset bubbles are hard to identify while they are developing, making it impractical to direct monetary policy against them. He went on to suggest that there is no such difficulty of detecting bubbles after the fact and that the role for monetary policy is to limit the destabilizing consequences. Investors came to believe, rightly or wrongly, that because the Fed would intervene it was no longer necessary to worry about the slim possibility, the so-called tail risk, of asset prices collapsing. The danger of large losses being less, the temptation to take on risk was more. And there was nothing to restrain the risk takers, what with the Federal Open Market Committee and the regulators, including the Fed itself, in denial about their ability to detect bubbles, much less to limit their growth.
TAYLOR RULES

The final grounds on which to implicate the Fed is that monetary policy was significantly looser in 2002–2005 than it would have been had the Fed followed the same script as in the previous 15 years. From the end of the Volcker deflation in the mid-1980s through the bursting of the tech bubble in 2000, Fed policy was well captured by the “Taylor Rule.” Named after the Stanford economist John Taylor, who had identified it in 1992, the Taylor Rule was a stylized relationship linking the central bank’s main policy lever, the interest rate at which banks lend to one another, known as the federal funds rate, to its principal policy objectives: inflation and the level of idle resources.19 The Taylor Rule had done a good job of capturing the Fed’s reaction to changes in inflationary pressures and business cycles from the mid-1980s through the end of the 1990s. It continued to track policy after the turn of the century. When the tech bubble burst in 2000, causing unemployment to rise and inflation to subside, the Taylor Rule pointed to the need to cut rates. Over the course of 2001, the federal funds rate was cut from 6.25 percent to 1.75 percent, in line with its predictions.
But once the economy bottomed out late in 2002, causing the gap between actual and potential output to stop widening, the Taylor Rule suggested raising rates. Instead the Fed reduced the funds rate still further until it reached a low of 1 percent in mid-2003. There the policy rate remained for a year, at the end of which it was fully 3 percentage points below the levels suggested by the rule.
This was when Greenspan and his colleagues grew concerned that the economy was slipping into a Japanese-style deflation from which it might be difficult, even impossible, to extricate it. (Actually, the word “deflation” was too alarming to use in public. Greenspan referred instead in congressional testimony to “an unwelcome further fall inflation.”) Given the Fed’s awareness that monetary policy affects the price level and economy with a lag, it sought to make policy with the future in mind. In other words, it was expected deflation rather than actual deflation, something that was not in fact visible, that dictated its actions.20
With benefit of hindsight we can say that the Fed overestimated the risk of deflation from early 2002 through mid-2004. It extrapolated too mechanically from the deflation that followed the bursting of Japan’s financial bubble in the late 1980s.21 Chairman Greenspan’s self-named “risk management approach” to monetary policy dictated erring on the side of averting the risk of a Japanese-style deflationary crisis.22 In the event, this policy erred in the direction of fueling an even greater boom and bust down the road.
This story of lax monetary policy fueling the mother of all credit booms has its critics, notably Federal Reserve officials past and present. They object that the Fed controls only short-term interest rates, not the long-term rates that matter for investment. But while long-term rates may be what matter for firms contemplating investments in factories, this is not equally the case of individuals deciding whether to buy a home. Insofar as rates on fixed-rate mortgages did not fall enough, households desperate to join the homeownership society could opt for adjustable mortgages, the initial rates on which, linked to one-year interest rates, were temporarily low.23 Lenders seeking to attract additional business by offering teaser rates on which interest payments were below market rates for the first few payment periods were better able to finance the tease. And with a boom in the demand for housing, there was then a boom in housing starts.24
STRETCHING FOR YIELD

But while the run-up in asset prices in 2003–2006 centered on residential real estate, it was a broader phenomenon. For monetary policy to have been important, it would have to had to fuel leverage and risk taking not just by home buyers and builders but by investors in other assets. Monetary policy operating on short-term rates could have done so in a number of ways. First, lower nominal interest rates encouraged institutions to take on more risk in order to match previous returns. Some investors use previous returns as a gauge of managers’ performance. If returns go down, they blame the managers. To retain clients, the manager is forced to make riskier investments and use more leverage.
Second, some financial firms, such as pension funds and insurance companies, are required to pay out fixed nominal amounts to their investors.25 A pension fund operating a defined benefit plan, for example, is obliged to make a specified monthly payment to its contributors. If market interest rates go down by more than the company expected when signing the pension contract, the yield on safe securities may not be enough for it to meet its obligations. A bank that has offered certificates of deposit and whose other liabilities bear fixed interest rates may likewise find itself squeezed. In both cases portfolio managers, to meet the institution’s obligations, will have to move into riskier investments or take on more leverage.
Third, lower interest rates cheapen wholesale funding—they reduce the cost of borrowing chunks of money from other institutional investors. Lower money market rates thus encourage financial intermediaries to borrow more and expand their balance sheets. This behavior will be particularly visible among broker-dealers who rely on the wholesale money market for their funding.26 Predictably, broker-dealers like Bear Stearns were among the most highly leveraged of all financial institutions in the run-up to the crisis. They then fell the hardest.
Finally, if lower interest rates and more ample liquidity boost stock prices, including the stocks of financial institutions themselves, banks will want to increase their lending. Higher share prices mean that banks have more capital—the funds that their owners have subscribed to fund the operations of the institution are worth more. If the bank doesn’t expand its lending, some of this capital will be sitting idle. If the firm is not fully utilizing its lending capacity, it will be leaving money on the table. This forgone opportunity is something it will seek to correct. Low interest rates that translate into higher equity prices will thus trigger a lending boom.
Under these circumstances, equities and land will become more valuable, encouraging financial institutions to lend against them. Banks will give more loans to borrowers with lower credit scores, since if bad behavior recurs they can always seize and liquidate their collateral. If something causes the value of that collateral to fall, all bets are off. But that is a problem for the future. It was not something about which lenders, caught up in the moment, especially worried.
Low interest rates thus encouraged investors to assume more risk and use more leverage. Although the level of interest rates was not the only factor at work, it fanned the flames. Similarly, Federal Reserve policy was not the only thing keeping interest rates low. Not just short- but also long-term rates, which are not directly under the Fed’s control, were unusually low around the middle of the decade. Even when the Fed allowed the funds rate to rise in 2004, rates on bonds and fixed-rate mortgages remained anomalously low. This was the bond market “conundrum” on which Chairman Greenspan commented in a much-noted February 2005 speech. These low long-term rates may have been yet another manifestation of the Fed’s success at becoming more transparent and limiting perceived uncertainty about the future. The reduction in uncertainty gave investors the confidence to bid up bond prices.27
ENTER THE DOLLAR

But another, potentially more important culprit, fingered by Fed chairmen Greenspan and Bernanke once the low level of long-term rates came to be seen as less boon than problem, was foreign central bank purchases of U.S. government bonds and the securities of the quasi-governmental agencies Freddie Mac and Fannie Mae (so-called agency securities). And behind those foreign purchases lay the status of the dollar as the world’s reserve currency.
Some of the facts are incontrovertible. Central banks made extensive purchases of U.S. treasury and agency securities. In 2008–2009 they became the dominant foreign purchasers as private investors, increasingly concerned about the stability of the dollar, drew back. Central banks were motivated by the lessons they drew from the Asian financial crisis of 1997–1998, namely that capital flows are volatile and the only guaranteed protection against an abrupt reversal is to stockpile dollars so that short-term foreign liabilities, not just of the government but of the private sector as well, can be paid off. Korea, one of the countries most traumatized by the crisis, boosted its reserves, mainly of dollars, from 5 to 25 percent of GDP. Others followed, accumulating U.S. treasury and agency securities hand over fist. It must have been true that the prices of these dollar assets were higher, and the interest rates they bore were lower, than in the absence of this additional demand. And as the return on relatively safe assets fell, other investors “stretched for yield” by shifting into riskier securities.
But beyond these points, agreement does not extend. Bernanke referred to a “global savings glut” when explaining why foreign capital had been flowing into U.S. debt securities.28 This terminology was unfortunate in that global savings and the savings of emerging Asian economies, China in particular, which were among the principal buyers of U.S. securities, had been trending downward for 7 years. Emerging Asian savings, after having averaged nearly 33 percent of GDP in the last 4 years of the 1990s, fell to 31 percent in the first half of the 2000s, the so-called savings-glut years.29 But investment rates in emerging Asia fell even more, first because of the crisis-induced recession in 1998 and then as Asian governments abandoned the practice of running their economies under high pressure of demand. With Asia’s saving having risen relative to its investment, its excess funds had nowhere to flow but abroad. There being no shortage of U.S. debt securities, the U.S. treasury and agency market became the logical destination.
China was its own story, of course. China is not accurately characterized as having engaged in lower levels of investment. But its savings rates soared even higher as Chinese households socked away funds to provision for education, health expenses, and other contingencies. Enterprise managers, under no pressure to pay out dividends, retained earnings for capacity expansion at home and acquisitions abroad. All that additional savings had to go somewhere; in practice it could only go overseas. Capital outflows from China ballooned from one-tenth of 1 percent of global GDP in 2002 to nearly 1 percent of global GDP in 2007. On this basis it is argued that China’s financial underdevelopment contributed to the buildup of systemic risks.30
TANGO LESSONS

But it takes two to tango. High levels of Chinese saving were matched by low levels of U.S. saving. Household savings rates in the United States fell from 7 percent in the early 1990s to near zero in 2005–2007. At the time it was argued that this drop reflected the robust health of the economy and a Great Moderation that justified higher asset prices. Higher asset prices that included higher real estate prices made U.S. households, whose most important investments were their homes, feel wealthier. And on this basis they saved less.31 We now know that this vision of a permanent increase in wealth was an illusion, albeit one on which too many households that extracted equity from their homes based their decisions.
In the absence of this behavior—that is, with more U.S. savings—flows of capital toward the country would have been less. Still, there is little question that developments in the rest of the world, whether they are characterized as a savings glut or an investment strike and whether they are located primarily in emerging Asia or more broadly, contributed to the flow of capital into U.S. debt markets. They also had a self-reinforcing character. Capital inflows contributed to the rise in U.S. asset prices, which made the country appear more creditworthy, encouraging foreign investors to lend it more, much as a bank lends more against more valuable collateral.
There is also the question of whether all this dollar accumulation by central banks really reflected the demand for insurance. Some central banks accumulated reserves not as insurance against a sudden reversal in the direction of capital flows but as an inadvertent by-product of policies of export-led growth. Emerging-market central banks, most notoriously the People’s Bank of China, bought foreign currencies to prevent their exchange rates from rising. This encouraged the exports of manufactures that were the engine of economic development and the vehicle for transferring workers to the modern industrial sector.


Figure 5.1. S&P/Case-Shiller U.S. National Home Price Index.
Source: Standard & Poor’s.

Views of the relative importance of these motives tended to shift over time. Early observers emphasized worries about financial volatility on the part of Asian policymakers as the main factor motivating the accumulation of reserves. Subsequently some of the very same commentators emphasized the reluctance of governments to allow their exchange rates to rise and risk disrupting the process of export-led growth.32
Whatever the motivation, the result was an enormous accumulation of reserves. But why dollar reserves? Europe is as important as the United States in providing a market for Asian exports. In principle, Asian central banks and governments could have intervened to prevent their currencies from rising against the euro. They could have acquired euros—and British pounds and Swiss francs—instead of dollars.
PRIVILEGE ONCE MORE

The fundamental explanation for the decision to target the dollar was its status as the leading international currency. With other countries still shadowing the dollar, doing likewise produced stable exchange rates not just vis-à-vis the United States but more generally. For countries engaged in intra-Asian trade in parts and components, this indirect way of stabilizing exchange rates was of considerable benefit. With the largest share of world trade invoiced and settled in dollars, stabilizing local currencies vis-à-vis the dollar was particularly convenient for exporters. And pegging to the dollar encouraged the practice of accumulating dollar reserves.
Then there is the fact that the market in U.S. debt securities was so liquid. The costs of buying and selling them were low. Central banks and governments could make purchases and, when necessary, sales without moving prices. And those markets were so liquid precisely because of the participation of so many foreign central banks and governments. This was the dollar’s exorbitant privilege as the once and still reserve and international currency.
The issue is how much impact all this foreign finance had on U.S. interest rates. Were foreign purchases mainly responsible for Greenspan’s bond market conundrum? While foreign capital inflows were large, U.S. debt markets were larger. As late as the end of 2006, a majority (55 percent) of U.S. government bonds and an even larger fraction (85 percent) of the securities issued by the quasi-governmental agencies Freddie Mac and Fannie Mae were held by domestic investors. That said, foreign and especially Asian central bank purchases became increasingly important as the period progressed. One study finds that yields on 10-year bonds were at least half a percentage point (50 basis points) lower in 2005 than if there had there been no additional foreign purchases since the beginning of 2004.33 Another suggests that that 10-year bond yields were 70 basis points lower as a result of foreign capital inflows.34 Still another suggests that the increase in U.S. treasuries held by foreigners depressed yields by 90 basis points.35
Given the notorious inability of economists to agree, this is a remarkable degree of consensus. It suggests that foreign purchases of U.S. debt securities were largely, even wholly, responsible for Greenspan’s conundrum. This was the dollar’s exorbitant privilege in yet another guise.
But would long-term interest rates a half or even full percentage point higher have made all that much difference for the course of the crisis? The perverse financial practices and lax regulation that were at its root still would have been there. Mortgage brokers still would have had no fiduciary responsibility to the households with which they did business. Financial institutions still would have repackaged mortgage-backed securities into complex derivatives. Standard & Poor’s and Moody’s still would have advised originators on how to structure CDOs before proceeding to rate them. The belief that the economy and financial markets had become less volatile still would have encouraged risk taking.
But not to the same extent. With mortgage finance more expensive, the housing market would not have overheated so dramatically. There would not have been as much lending on the security of overvalued collateral. Borrowed money would have been more expensive. Leverage would have been less. When the process unwound, it would have unwound less violently.
APRÈS LE DELUGE

In the event, the violence with which it unwound was unprecedented. Investors took deep losses on the financial derivatives that had been the signature of the boom. Foreign central banks halted their acquisition of U.S. agency securities when the troubles of Freddie and Fannie became apparent. The happy belief that capital, whether private or public, flowed toward the United States because of the country’s singular capacity to originate and distribute high-quality financial assets dissolved in the face of these events. The slogan “They sell us high-quality merchandise, we sell them high-quality financial assets” was replaced by “They sell us toxic toys, we sell them toxic securities.”
There was no sign, however, of foreign investors and, specifically, foreign central banks withdrawing from the U.S. treasury market. Foreign purchases continued unabated, although there was some tendency now for central banks to buy shorter term treasuries. But with the deep recession caused by the crisis and the massive budget deficits that followed, questions were increasingly asked about the sustainability of the Treasury’s debt. With a majority of U.S. government debt now held by foreigners, the temptation to inflate it away was greater. Foreign investors could see the writing on the wall. Questions were increasingly asked about whether foreigners would retain their healthy appetite for U.S. treasury securities—and, if not, what their growing distaste might imply for the dollar.
There was growing dissatisfaction as well with an international monetary system that gave the United States access to cheap foreign finance that it deployed in such counterproductive ways. It was entirely reasonable that central banks should want to accumulate reserves as a buffer against volatile capital flows. But it was unreasonable that the only way of doing so was by shoveling financial capital into the United States and fueling the excesses that drove the global financial system to the brink. America, it followed, was no longer to be entrusted with its exorbitant privilege. A privilege abused was not a privilege that others would willingly extend. On how a new international monetary system should be structured there was little agreement. The one point on which the critics agreed was that it should entail a more limited role for the dollar.
On all these grounds, pessimism reined about the dollar’s future as a store of value, means of payment, and unit of account and, by implication, about its international role. The dollar was doomed. The dollar was in terminal decline. Without official foreign demand to prop it up, the dollar exchange rate would collapse, eroding America’s living standards and geopolitical leverage.
The only question being: were these pessimistic forecasts right?
CHAPTER 6
MONOPOLY NO MORE
In the wake of the crisis, doubts are pervasive about whether the dollar will retain its international role. Recent events have not exactly enhanced the reputation of the United States as a supplier of high-quality financial assets. It would not be surprising if demonstrations of the dysfunctionality of American financial markets soured investors on U.S. debt securities. Meanwhile, a budget-deficit-prone U.S. government will be pumping out debt as far as the eye can see. It will be tempted to resort to inflation to work down the burden. That temptation will be even greater now that a majority of its marketable debt is held by foreigners.
Foreign investors, the Cassandras darkly warn, will not sit still for this. They will seek to protect themselves by curtailing their dollar holdings. The end result, the worriers caution, could be a mass migration to other currencies.
If developments in the United States raise doubts about the dollar’s international role, developments abroad deepen them. The post–World War II recovery of Western Europe and Japan and now the emergence of China, India, and Brazil have reduced the economic dominance of the United States. It is not obvious why the dollar, the currency of an economy that no longer accounts for a majority of the world’s industrial production, should be used to invoice and settle a majority of the world’s international transactions. Nor is it clear why the dollar should still constitute a majority of the reserves of central banks and governments. As the world economy becomes more multipolar, its monetary system, logic suggests, should similarly become more multipolar. This reasoning implies at a minimum that the dollar will have to share its international role.


Figure 6.1. Foreign Holdings of U.S. Securities as a Percentage of Total U.S. Amount Outstanding.
Source: Federal Reserve Board Flow of Funds Accounts of the United States.

Moreover, what is true of the economic logic for a dollar-based international monetary and financial system is true also of its political logic. When after World War II the United States stationed large numbers of troops in Europe and Asia, our allies there saw supporting the greenback as an appropriate quid pro quo. Today, in contrast, China, our largest foreign creditor, is not a close ally. In many parts of the world, the American security umbrella is neither as essential as it once was nor, indeed, as welcome. It is not obvious that the best way for foreign countries to ensure their security is by propping up the dollar. All this makes them increasingly critical of America’s exorbitant privilege.
AN INCONVENIENT TRUTH

There is only one problem with these arguments. It is that there has been little actual diminution of the dollar’s role in international transactions. There has been no discernible movement away from the dollar as a currency in which to invoice trade and settle transactions. One recent study for Canada, a country with especially detailed data, shows that nearly 75 percent of all imports from countries other than the United States continue to be invoiced and settled in U.S. dollars.1 The dollar similarly remains the dominant currency in the foreign exchange market. The most recent Bank for International Settlements survey showed that the dollar was used in 85 percent of foreign exchange transactions worldwide, down only marginally from 88 percent in 2004.2 Some 45 percent of international debt securities are denominated in dollars.3 OPEC continues to price its petroleum in dollars.
While U.S. nemeses like Iran and Venezuela regularly offer proposals for pricing oil in another currency, there is no agreement about what constitutes an attractive alternative. The famous instance was in November 2007 when, at a closed-door session in Riyadh, a camera was inadvertently left on, broadcasting into a nearby press room a quarrel between the Iranian and Saudi foreign ministers over whether OPEC should move away from dollar pricing. In October 2009 a sensational if undocumented press report had the Gulf States conspiring with China, Russia, Japan, and France—now there’s an odd coalition—to shift the pricing of oil away from dollars.4 But so far all this has been a tempest in a teapot.
Data on the currency composition of central banks’ foreign reserves are incomplete, since not all countries report. China, importantly, is among the nonreporters. But data from the IMF, the best source on the subject, show the share of dollars in total identified official foreign exchange holdings as of the first quarter of 2010 as 61 percent, down only marginally from 66 percent in 2002–2003.5 If one goes back further, to the first half of 1990s, the dollar’s share in total identified official holdings of foreign exchange was actually lower than recently. Plus ç a change.
Although the IMF’s statistics are not perfect, other sources point in the same direction. For example, surveys of financial institutions conducted by the U.S. Treasury suggest that foreign central banks continued to accumulate treasury bonds following the outbreak of the crisis—if anything at an accelerating pace.6 There was a sharp fall in foreign central bank accumulation of “agency securities”—the securities of the quasi-governmental mortgage agencies Freddie Mac and Fannie Mae—but not of U.S. treasuries.


Figure 6.2. 52-Week Change in Custodial Holdings on Behalf of Foreigners.
Source: Federal Reserve Board.

STILL THE ONE

What explains the gap between rhetoric and reality? Above all the simple fact that, jeremiads about American declinism notwithstanding, the United States remains the largest economy in the world. It has the world’s largest financial markets. This may not be true forever, but remains true now.
Moreover, the dollar has the advantage of incumbency. Consider an exporter deciding in what currency to quote the prices of his exports. Exporters want to limit fluctuations in their prices relative to those of competing goods in order to avoid confusing their customers. If other exporters are invoicing and settling their transactions in dollars, each individual exporter has an incentive to do likewise.7 And to continue doing so.
And what is true of trade is true of other international transactions. That so many exports are priced and settled in dollars makes the dollar the dominant currency in foreign exchange markets, since exporters from other countries, when they want to pay their suppliers, workers, and shareholders, must first convert the proceeds back to their home currency. It makes the dollar the dominant unit in currency forward and futures markets, since exporters will want to use those markets to ensure against unexpected exchange rate movements while the transaction is still under way. Since it pays for exporters of financial services, like exporters of merchandise, to avoid confusing their customers, they, too, will price their products in the same currency as their competitors. Thus, the fact that international bonds were denominated in dollars in the past creates a tendency for them to be denominated in dollars in the present.
For many central banks, it similarly makes sense to stabilize their exchange rates against the dollar even though the United States no longer accounts for a majority of their foreign trade and financial transactions—if for no other reason than that other countries stabilize their exchange rates against the dollar. Because other countries peg to the dollar, doing likewise stabilizes a country’s exchange rate not just against the United States but more broadly. Beyond that, there is a reluctance to shift away from dollar pegs, since they are the established basis for monetary policy, and a change may sow uncertainty.
Central banks will want to hold reserves in the same currency in which the country denominates its foreign debt and invoices its foreign trade, since they use those reserves to smooth debt and trade flows. They will want to hold reserves in the currency of the country to which they peg, since they use them to intervene in foreign exchange markets.
Although central banks naturally welcome returns on their investments, they also seek to limit the riskiness of their reserve portfolios. Importantly, the currency to which they peg will be the most stable in terms of its domestic purchasing power (that is, in its command over domestic goods and services). And the identity of the predominant anchor currency is no mystery. As of mid-2009, fifty-four countries pegged to the U.S. dollar, compared to just twenty-seven to the euro, the runner-up.8
Calculations of what combination of dollars and other currencies are attractive to central bank reserve managers assume for convenience equally liquid markets in bonds and deposits denominated in different currencies.9 This assumption may be unrealistic, but relaxing it only works in the dollar’s favor. Central banks value liquidity in their reserve instruments so that they can use them in market intervention. If a financial instrument is not readily convertible into cash, then it is not readily used in market operations.
It therefore matters greatly that the market in U.S. treasury bonds and bills has unrivaled liquidity whether measured by turnover or transactions costs. The U.S. treasury market is, quite simply, the most liquid financial market in the world. This reflects the scale of the U.S. economy and its financial development. But the status quo is self-reinforcing. Because the U.S. market is so liquid, foreign investors undertake transactions and concentrate their holdings there. The fact that they undertake their transactions and concentrate their holdings there in turn lends it additional liquidity.
Thus, in the same way that incumbency is an advantage in the competition to be an international financial center, it is an advantage in the competition for reserve-currency status. Incumbency is not everything. And its advantages may be weakening; the costs of comparing prices in different currencies and switching between them are declining with the development of modern information technologies. But as any politician will tell you, the advantages of incumbency are not to be dismissed. They are one reason that, questions about its reelection prospects notwithstanding, the dollar is unlikely to be voted out of office just yet.
SMALL POTATOES

Yet another factor favoring a continuing role for the dollar is that all the other candidates for international currency status have serious shortcomings of their own. The UK and Switzerland are simply too small for the pound sterling and Swiss franc to be more than subsidiary reserve and international currencies. Both lack the size to provide debt instruments on the scale required by the global financial system. The UK is barely a sixth the economic size of the United States. Switzerland is barely a thirtieth. Given the importance of market size for liquidity, the share of their currencies in global reserves is even less. Sterling accounts for less than 4 percent of identified global reserves, the Swiss franc for less than 1 percent.
The same is even truer of still smaller economies. When Russia’s central bank announced in 2009 that it was diversifying its reserves to include Canadian dollars (nicknamed “loonies”), those anticipating the death of the U.S. dollar gleefully took note. But the announcement caused nary a ripple in the U.S. dollar exchange rate. Canadian government bond markets are simply too small to make a dent in global reserve portfolios or for Russia’s decision to buy loonies to have a discernible impact on the greenback.
Japan is a larger economy, but its government long discouraged international use of the yen on the grounds that this would undermine Japan’s ability to maintain a competitive exchange rate and would otherwise complicate its conduct of industrial policy.10 This reluctance to internationalize the yen may now be a thing of the past. Japanese officials are anxious to see their currency play a larger role, especially in Asia. But past policy continues to shape market liquidity. And a decade of no growth and zero interest rates have made holding reserves in yen unattractive. The yen accounts for barely 3 percent of total identified official holdings of foreign exchange.11 Going forward, Japan’s aging population and antipathy to immigration do not favor a rapidly expanding global role for its economy or its currency.
WHOM TO CALL

This leaves the euro, notwithstanding its recent difficulties, as the most serious rival to the dollar for now.12 The euro area possesses the requisite scale. Its exports are nearly double those of the United States.13 Germany is a major exporter of capital goods to emerging Asia. Euro-area companies operate branch plants in countries to their east, countries that are increasingly linked into Western Europe’s production networks and supply chains. Euro-area banks own and operate many of Eastern Europe’s banks. This makes the euro a logical currency in which to invoice and settle transactions for importers and exporters in economies adjoining the euro area as well as other parts of the world.
If a first-class international currency needs a first-class central bank, then this, too, is something the euro possesses. The ECB has shown itself to be extraordinarily serious about the maintenance of price stability. Although it was roundly criticized for its exceptional purchases of government bonds in the spring of 2010, it continues to take its price-stability mandate seriously. It shows absolutely no inclination to embark on reckless inflationary polices.
At the same time the ECB understands its responsibility as an emergency lender. In 2008, at the height of the crisis, it extended emergency loans to countries whose banks and firms had borrowed in euros. It provided other central banks with euros in exchange for their currencies. Such swaps are the signature of a central bank that recognizes its currency’s international role. Henry Kissinger’s quip, “You don’t know who to call when you want to telephone Europe,” no longer applies to monetary policy. You call the ECB.
And appropriately for an aspiring supplier of reserve assets, the euro area also possesses an ample stock of government debt securities. Its bond markets are accessible to foreign investors, controls on capital flows being a thing of the past.
It is important to recall these positive attributes, especially at times when the euro becomes currency traders’ punching bag. Recall, for example, that the euro area served as a safe harbor in the 2008 crisis. That episode demonstrated that the European Central Bank, as the issuer of a recognized international currency, has more capacity to provide emergency liquidity than, say, the National Bank of Denmark or the Swedish Riksbank, the central banks of countries still outside the euro area. The ECB provides emergency assistance in euros, a unit that is widely used in cross-border transactions and to which risk-averse investors flee in a crisis. And in times of crisis, these attributes make the euro area a safe place to be.
The ECB’s intervention was most critical in the period following the failure of Lehman Brothers, when it cut interest rates and flooded financial markets with liquidity. In contrast, the National Bank of Denmark, still the steward of a national currency, had to raise interest rates in response to deleveraging by foreign investors that led to a sharp fall in the value of the krone. Even had the Danish central bank disregarded the implications for the exchange rate and flooded financial markets with krone, this would have been no help to Danish firms and banks with obligations in euros. At the height of the crisis the National Bank had to negotiate emergency swap lines with the ECB, which provided it with the euros needed to relieve the pressure on those firms and banks.
But those arrangements were ad hoc. Whether they will be repeated is uncertain. The only guarantee of access to the ECB’s liquidity facilities is by adopting the euro. The governor of the Danish central bank, Nils Bernstein, acknowledged as much in an interview in the Irish Independent: “The crisis has shown that we can manage economically outside the euro, but it has also demonstrated that there are big advantages during a crisis to be inside and much more protected against turmoil and to have access to the euro system’s facilities.”14 The euro area’s difficulties in 2010 will undoubtedly cause candidates for membership to think twice before coming in. But as they cast their minds back to 2008, they will recall that it is not particularly attractive to stay out either. Estonia, by choosing to adopt the euro at the beginning of 2011, has already voted with its feet.
The obvious exception is Britain, where the crisis tarnished the reputation of a pro-EU Labour Government and led to its replacement by a euro-skeptical opposition. Given London’s position as an international financial center and sterling’s history as a reserve currency, Britain’s joining the euro area would make the biggest difference to the euro’s status as an international currency. But this is not something that is going to happen anytime soon, given the sour aftertaste from Britain’s earlier flirtation with the Exchange Rate Mechanism.15
This leaves growing the members’ own economies as the best way of creating a larger platform for the euro.16 But Southern European countries, grappling with a difficult process of fiscal consolidation, now face an extended period of slow growth. More generally, the capacity to grow is not one of Europe’s strengths. The continent has inflexible product and labor markets. Firms are reluctant to hire because they are unable to fire. Start-ups are slow to ramp up because of obstacles to ramping down. They hesitate to scale up until they are confident that they can continue to employ those they take on.
Some will say that these problems are overstated. But even if these other criticisms of Europe’s economic prowess are contestable, there is no question that the continent is challenged demographically. Its population is aging. Its net reproduction rate, the number of daughters per woman who survive to average reproduction age, is only 0.75, well below the value of unity required for a stable population. Immigration from Turkey and North Africa could make up the difference, but few in Europe relish this prospect. Most Europeans oppose admitting Turkey to the EU, which would be the obvious way of solving Europe’s demographic problem. What is true for Japan is true for Europe: a stagnant population will mean a stagnant economy. This stasis will not translate into a larger platform for the euro or make it more attractive as an international currency.
A CURRENCY WITHOUT A STATE

The euro, as the currency of an economic zone that exports more than the United States, has well-developed financial markets, and is supported by a world-class central bank, is in many respects the obvious alternative to the dollar. While currently it is fashionable to couch all discussions of the euro in doom and gloom, the fact is that the euro accounts for 37 percent of all foreign exchange market turnover. It accounts for 31 percent of all international bond issues. It represents 28 percent of the foreign exchange reserves whose currency composition is divulged by central banks.17 It is second only to the dollar on all these dimensions of “international currenciness,” although it remains a considerable way behind the leader.
So why hasn’t its progress been faster? There is the novelty of a unit that came into existence barely a decade ago. There is the fact that Europe’s bond markets are not larger and more liquid.18
But most fundamentally, the problem is that the euro is a currency without a state. It is the first major currency not backed by a major government, there being no euro-area government, only the national governments of the participating countries.19 The European Commission is the proto-executive not of the euro area but of the European Union, which encompasses also the UK, Sweden, Denmark, and other EU countries that have not adopted the euro. The Commission has limited power to help governments with financial problems or to force them to take steps that they do not deem in their national interest. The manifest inability of the Commission to enforce the Stability and Growth Pact, which is intended to limit budget deficits, is a prime case in point.
This absence of a euro-area government is the main factor preventing the euro from matching the dollar in international importance. When Europe develops economic and financial problems, managing them requires cooperation among its national governments, which is far from assured. When a government develops budgetary problems so serious that it is impossible to resolve them without international assistance, providing it is something on which European countries as a group must agree. Their leaders have to negotiate a burden-sharing agreement, and then a host of national parliaments have to ratify it. The possibility that they won’t do so quickly, or at all, raises fears of unpredictable financial fallout. This in turn creates reluctance on the part of central banks in other parts of the world to put their eggs in the euro basket.
The Greek crisis illustrates the point. How Athens developed a “big fat Greek deficit” in excess of 13 percent of national income is a story in itself, one that unfolded over a period of years. Be that as it may, in early 2010 investors awoke to the fact that Greece’s financial position was untenable. They demanded that the Greek government reduce its budget deficit by fully a tenth of national income in 3 years, a Herculean task, or else they would go on strike.
Greek prime minister George Papandreou made a valiant attempt to narrow the deficit. But public-sector workers, unconvinced that they should bear the burden of austerity, resisted accepting big pay cuts. Households and companies similarly resisted heavy increases in taxes.20 While there were good reasons why the government and Greek society had to go back to living within their means, the adjustments required met with considerable political opposition. But absent those adjustments the markets refused to advance the Greek government the funds it needed to service its debt and finance its other operations. The country was stuck between a rock and a hard place.
The only solution, barring a restructuring of the public debt that, for better or worse, no one was prepared to contemplate, was to stretch out the adjustment. But an extended adjustment might be possible only if other European countries advanced Greece a loan, conditional on Athens laying out a credible plan for balancing its budget. Here the absence of a powerful executive at the level of the euro area came to the fore. There was no European government with the capacity to lend, only the German, French, and other national governments.21 German officials, with regional elections coming, catered not to Europe’s interests but to German voters. And German voters were angry as heck about having to bail out their profligate Greek neighbors.
All this encouraged political grandstanding. German chancellor Angela Merkel made aggressive remarks about refusing to participate in a bailout for consumption by her domestic constituents while at the same time providing private reassurances to Papandreou. Papandreou, seeking a better deal, threatened to shun his European partners for the IMF, causing embarrassment in Paris and other European capitals. Investors began to worry that Greece would be forced to default on its debt, damaging French and German banks stuffed full of Greek bonds. The absence of a convincing resolution of the Greek problem caused investors to develop similar doubts about Portugal, Spain, and other Southern European countries. This led to panicked sales of their bonds, raising questions about the solvency of the European banks that had invested in those bonds themselves. Catastrophe was averted only when national leaders, meeting in Brussels, agreed to a $1 trillion fund to guarantee national debts and the ECB agreed to emergency purchases of government bonds. Obviously, this brinkmanship hardly reassured foreign central banks and others contemplating whether to use euros in their international transactions.
Everyone understands what is needed to address the problem. Europe needs stronger oversight of national budgets to prevent governments from getting into this pickle in the first place. It needs closer coordination of fiscal and other policies to prevent competitive positions from getting out of whack. And it needs no more emergency agreements at two o’clock in the morning. Instead it needs a proper emergency financing mechanism—a euro area crisis management institution run by a committee of technocrats answerable to the European Parliament and the member states. They would have a pool of resources that could be loaned to countries with strong policies experiencing financial problems through no fault of their own.22 They would have the power to extend emergency financial assistance, either unilaterally or in conjunction with the IMF, and conditioned on the recipient’s implementation of adjustment measures.23 They could purchase euro-zone government bonds in the event of disruptions to sovereign bond markets—and prevent the ECB from being pushed into doing so. When such assistance was not enough, they would step in to restructure the debts of the insolvent country. They would impose a “haircut” on the bondholders (writing down the value of their claims to a fraction of their face value) and giving them a menu of new bonds to choose from.24 A permanent mechanism of this sort would help to deal with Europe’s immediate difficulties but, more important, it would be in place to address future problems.
All this would involve significant additional delegation of national prerogatives to a European institution, which is why there was more talk than actual movement in this direction. Yet, three months into the Greek crisis, German finance minister Wolfgang Schäuble had come around to view that Europe needed this kind of mechanism. The European commissioner for economic and financial affairs, Olli Rehn, proposed requiring governments to clear their budgets with the Commission before submitting them to their national parliaments. He proposed extending Commission oversight to sensitive national arrangements that had previously been regarded as off-limits, such as the parliamentary procedures used to negotiate the budget, which might pose an obstacle to sensible outcomes, and the structure of national wage-bargaining arrangements, where these interfered with the maintenance of competitiveness. The European Central Bank issued a document endorsing both strengthened surveillance of national policies and the creation of a euro-area crisis management institution.25 Even more significantly, in late June the European Council, made up of EU heads of state or government, agreed.
But it remained unclear whether European governments were prepared to move ahead with such changes. Would German voters agree to have their taxes go to fund the operations of a euro-area crisis management institution? Would the EU get serious about budgetary rules with teeth? Would the members of the monetary union allow the European Commission to interfere with delicate national prerogatives such as the procedures and conventions used to negotiate the budget and bargain over wages?
The answer, in each case, is unclear. And so long as Europe lacks the political will to create an emergency financing mechanism and, more generally, to put in place the other policies needed to complete its monetary union, the euro’s economic attractions as an alternative to the dollar will remain limited.
POWER OUTAGE

Historically, the leading international currency has always been issued by the leading international power. Nothing threatens that country’s existence. One reason the dollar dominated after World War II was that Fortress America was secure. As the English political economist Susan Strange put it in Cold War days, “It is just possible to imagine a future scenario in which West Germany is overrun by an exuberant Red Army while Fortress America remains inviolate across the Atlantic, but it is impossible to imagine the converse: a West German state surviving while the United States is overrun or the North American continent laid waste by nuclear attack. As long as this basic political asymmetry persists, there is no chance whatever of the Deutsche mark being the pivot of the international monetary system.”26 Foreign central banks and other investors want to know that their money is safe. And safety has more dimensions than just the financial.
The leading power also has the strategic and military capacity to shape international relations and institutions to support its currency. After World War II the United States could insist that its allies support the dollar, and countries like Germany had no choice but to comply. The unmatched power of the United States permitted it to shape international institutions, at the Bretton Woods Conference and after, to support the dollar’s exorbitant privilege. Europe, in contrast, lacks a common foreign policy. It doesn’t even have a position on how to reform the International Monetary Fund and the international financial system to reduce their dependence on the dollar.
With time, Europe will move to a common position on reform of the international monetary system in ways that enhance the position of the euro. It will develop a common position on IMF reform. But these innovations won’t spring full-blown from Europe’s brow tomorrow or the day after. The continent will move gradually, if in spurts, toward deeper integration, as it always has. And because institutional reform will be slow, the euro’s rise as an international currency will be slow.
The one place where the euro is likely to gain market share rapidly is on the euro area’s own fringes. The euro is already the dominant currency for trade settlements and invoicing in non-euro-area EU countries. The EU is also seeking to develop stronger ties to the non-EU countries to its south and east. It has put in place a “Union for the Mediterranean” to deepen its links with non-EU countries bordering that sea. It relies on Russia for its energy supplies, which means that Russia in turn relies on it for revenues.
As countries in its neighborhood deepen their links with the EU, they will rely more on the euro. Thus, in 2009 Russia announced that it was raising the weight of the euro in the basket of currencies used to guide its exchange rate policy, reflecting the growing importance of the euro area in its foreign trade and payments. And as the euro becomes more important to Russia as a guide for policy, its central bank will want to hold a larger share of euro-denominated reserves.
Still, this is a recipe for a regional reserve currency, not a dominant global unit. For the euro to rival the dollar as a global currency, one of two things would have to happen. Attitudes toward sovereignty would have to change. Europe would have to move toward deeper political integration; it would have to issue euro-area bonds and create government bond markets with the liquidity of the U.S. treasury market. Or the United States would have to badly bungle its economic policies, sowing distrust of its currency.
PRISONERS OF THEIR OWN DEVICE

In its annual report for 2008, the Central Bank of the Russian Federation revealed that it had reduced the share of dollars in its reserves from 47 to 41.5 percent between the end of 2007 and the end of 2008, while raising the share of the euro from 42.4 to 47.5 percent. Then in mid-2009, in a poke at the United States, the central bank’s first deputy chairman, Alexei Ulyukayev, announced that Russia intended to reduce the share of dollar-denominated assets in its portfolio still further.
But where Russia can do pretty much as it pleases without causing too much trouble, China is different by virtue of the sheer size of its holdings. Its official dollar assets are roughly eight times Russia’s. China is estimated to control nearly half of all U.S. treasuries in the hands of official foreign owners.27 Some 65 percent of China’s $2.5 trillion of reserves are in dollar-denominated assets.28 China selling U.S. treasury securities in quantities sufficient to significantly alter the composition of its reserve portfolio would cause their prices to tank. To the extent that dollars still comprised a significant portion of its reserves, the People’s Bank would suffer additional accounting losses. If it moved significant amounts of money into other currencies, the dollar would depreciate, making for further losses on those residual holdings.29
Since dollar depreciation would make U.S. imports more expensive, Chinese exporters would suffer. This is not a minor matter for a China that depends on exports for employment growth. It is a major consideration for a country that experiences some 70,000 civil disturbances a year.
Moreover, disruptions to the U.S. treasury market that sharply raise U.S. interest rates would not endear China to its American interlocutors. Transactions that cause the dollar to depreciate abruptly, leaving investors wrong-footed and roiling international markets, would not please other countries. One is reminded of Keynes’s line, “When you owe your bank manager a thousand pounds, you are at his mercy. When you owe him a million pounds, he is at your mercy.”


Figure 6.3. China’s Holdings of U.S. Securities.
Source: Following Bertaut and Tryon (2007).

The sensible strategy under such circumstances is to adjust one’s portfolio gradually and inconspicuously. This is, in fact, what China has been doing. It is yet another reason that the declining dominance of the dollar in reserve portfolios, to the extent that it occurs, will be gradual rather than sudden.
To be sure, Chinese officials feel pressure to do something. That the issue has become a flashpoint domestically is not surprising when one observes that China’s foreign reserves amount to $2,000 per resident. They are the equivalent of a third of Chinese per capita income. In 2009 China’s Global Times newspaper ran an online poll in which 87 percent of respondents called China’s dollar investments unsafe. During his 2009 visit to China, U.S. treasury secretary Timothy Geithner felt compelled to address the issue before an audience of students at Beijing University, attempting to reassure them that Chinese holdings of U.S. treasury bonds were secure. His answer drew hoots of laughter.30
China would like the United States to compensate it for any losses on its dollar-denominated securities, like the guarantee the British government extended to the members of the sterling area after the pound’s 1967 devaluation. But it is hard to imagine any circumstances under which the U.S. Congress would agree.
FUNNY MONEY

Recognizing that selling dollars is risky and that, even if the transactions can be safely executed, it is not clear what to replace them with, the Chinese have begun exploring other options. In March 2009 Zhou Xiaochuan, the cerebral governor of the Chinese central bank, drew attention by arguing that the IMF’s Special Drawing Rights should eventually replace the dollar as the world’s reserve currency.31 SDRs, recall, are the bookkeeping claims on the IMF first created in the late 1960s to supplement dollars in official international transactions. Zhao in his speech even made explicit reference to the Triffin Dilemma that provided the analytical underpinning for SDRs in the first place.32
SDRs quickly became something of an intellectual fad. China, Russia, and Brazil announced their willingness to buy $70 billion of SDR-denominated bonds as their contribution to topping up the IMF’s resources.33 A United Nations commission chaired by the Nobel laureate Joseph Stiglitz advocated an expanded role for an international unit resembling the SDR, although the members of the commission indicated a preference for it to be issued not by the IMF, of whose policies they disapprove, but by a new “Global Reserve Bank.” How this would work, exactly, is unclear. As the commission dryly observed in its report, “in setting up such a system, a number of details need to be worked out.”34
It is not hard to understand the appeal of this idea in the abstract. Empowering the IMF or some similar entity to provide bookkeeping claims in the quantities required by the expansion of global trade and finance would address the need for balance-of-payments insurance. Rather than having to accumulate dollars with which foreign loans could be paid off and foreign goods could be purchased in an emergency, governments could use their SDRs, since other governments would be obliged to accept them. Having SDRs satisfy this need would eliminate the exorbitant privilege enjoyed by the United States and make the world a safer financial place. By creating an alternative to existing national currencies, it would solve the dilemma of large reserve holders like China.
MINOR OBSTACLES

At the moment, however, the SDR is only a bit player. Even after the April 2009 decision to proceed with the distribution of an additional $250 billion of SDRs to IMF members, SDRs still accounted for less than 5 percent of global reserves.
Even more fundamental than this question of scale is the question of utility. Reserve assets are attractive only if they can be used, and the usefulness of SDRs is limited. SDRs can be used to settle debts to governments and the IMF itself, but not for other purposes. They cannot be used to intervene in private markets because there are no private markets where SDRs are traded. They cannot be used to invoice and settle trade because no trade is invoiced and settled in SDRs.35 Central banks will find it attractive to hold SDRs only when a significant fraction of trade is invoiced and settled in SDRs. They will find it attractive to do so if and when private lending and borrowing take place in that unit. Until then, central banks will have to convert their SDRs into dollars or euros when they want to use them, incurring additional cost and inconvenience. Under current arrangements this process takes a minimum of five days, which is an eternity in a crisis.
Making the SDR attractive would require building deep and liquid markets on which SDR claims can be bought and sold. It would be necessary to build markets on which governments and corporations could issue SDR bonds at competitive cost. Banks would have to accept SDR-denominated deposits and extend SDR-denominated loans. It would be necessary to restructure foreign exchange markets so that traders seeking to buy, say, Korean won for Thai baht first sold baht for SDRs rather than first selling baht for dollars.
Anyone serious about going down this road should familiarize himself with the earlier failed attempt to create a private market in SDRs. In 1981 the IMF sought to jump-start the market by reducing the number of currencies making up the SDR from sixteen to five. The sixteen had included the currencies of all countries accounting for at least one percent of world trade. However, such a large number made the SDR hard to understand, and it included currencies like the Saudi riyal, the South African rand, and the Iranian rial that were not freely traded or for which forward markets did not exist. Banks refused to accept SDR-denominated deposits since they couldn’t hedge the risk on forward markets; they couldn’t protect themselves against losses due to exchange rate changes.
By simplifying the SDR basket to include only dollars, German marks, Japanese yen, French francs, and British pounds, the IMF thought it could solve these problems. Commercial banks seeking to test the market took out ads offering certificates of deposit in SDRs. Investment banks offered to underwrite SDR bonds on behalf of governments and corporations. There were a few modest indications of interest among investors in the first quarter of 1981. But with Paul Volcker at the Fed still at work wringing inflation out of the economy, this period was also one of high U.S. interest rates and a strengthening U.S. currency. The SDR depreciated by 7 percent against the dollar in the first quarter of 1981, and all interest on the part of savers and lenders dried up.
One might think that borrowers would have wanted to do business in a unit that became less valuable over time. Sweden in fact obtained a syndicated credit in SDRs in early 1981, but the only other governments that followed its example, the likes of Ireland and the Ivory Coast, were small potatoes. As the Federal Reserve Bank of New York put it in an understated assessment at the end of 1981, “nonbank investor and borrower interest has been modest to date.”36 It became even more modest subsequently.
It is easy to see why there was so little progress. The first private entity issuing an SDR bond or deposit incurred extra costs as a result of the instrument’s illiquidity. The first private SDR, by definition, was not traded in a broad and deep market. Purchasers required additional compensation to hold it. And since liquid markets in claims denominated in national currencies already existed, private SDRs traded at a disadvantage.
Moreover, anyone who preferred borrowing or lending in a basket of currencies was apt to favor a tailor-made basket that would suit his or her financial needs more closely and whose components trade in more liquid markets. There was no particular reason that the weights attached to the five currencies making up the SDR basket would be the same as the proportions in which an investor would want to hold bonds denominated in those five currencies. If the diversification benefits of holding different countries’ bonds appealed to an investor, he could roll his own portfolio. And since the costs of buying and selling private SDRs were high, SDRs had no cost advantage to offset the attractions of a bespoke portfolio.
GETTING SERIOUS

Building private markets in SDR-denominated securities will require sustained investments by the relevant stakeholders, in this case, governments. If China is serious about giving the SDR reserve-currency status, in other words, it should take steps to create a liquid market in SDR claims. It could issue its own SDR-denominated bonds in Hong Kong. This would be a more meaningful step than buying SDR bonds from the IMF. Those bonds will not be traded, so they will do nothing to enhance market liquidity. A Chinese government bond denominated in SDRs would be another matter. Like U.S. treasury bonds, that instrument would be actively traded by investors. And where China led, Brazil and Russia could follow.
The first governments issuing SDR bonds will pay a price, since investors will demand an interest-rate premium to hold them. Bondholders will demand additional compensation for the novelty of the instrument and its lack of liquidity. But nothing is free. That price will be an investment in a more stable international system. Time will tell whether countries like China are willing to pay it.
Then there is the question of exactly who will find it attractive to buy the securities that governments sell. Many government bonds are held by pension funds and insurance companies, since the maturity of those bonds matches the maturity of their obligations to their clients. Domestic government bonds have the advantage that they are in the same currency as the pension or insurance payments that the company makes to its customers. This relieves it of having to worry about changes in exchange rates.
SDR bonds, on the other hand, would not be in the same currency as those pension fund and insurance company obligations. If the dollar depreciates against the euro, a European insurance company with SDR-denominated investments but euro-denominated liabilities would quickly find itself in the soup.37 One day, far in the future, policy holders and pensioners may be prepared to accept payouts in baskets of currencies. But putting the point this way is a reminder that the day when there is a deep and liquid market in SDRs, with adequate demand and supply sides, remains very far away.
A decision to create large numbers of SDRs on a regular basis to meet the global demand for reserves would also have to confront the delicate question of who gets them. When IMF members agree to increase the number of SDRs, they are allocated according to an agreed formula. But what formula should be used in the future? Would SDRs be allocated in proportion to currently existing reserves? Or mainly to the poorest countries with the most need? In the absence of a consensus about who gets the goodies, there is unlikely to be a commitment to ongoing SDR issuance on the scale needed to replace existing reserve currencies.
Finally, in a world where the SDR was the dominant international currency, the IMF would have to be able to move quickly to issue additional SDRs at times of crisis, much as the Fed and ECB provided dollar and euro swaps to ensure adequate dollar liquidity in 2008. Under current rules, countries holding 85 percent of IMF voting power must agree before SDRs can be issued.38 This is not a recipe for quick action. IMF management would have to be empowered to decide on SDR issuance, just as the Federal Reserve can decide to offer additional dollar swaps. For the SDR to become more like a global currency, in other words, the IMF would have to become more like a global central bank and provider of emergency liquidity. Again, this is not something that is going to happen overnight.
The case for a global currency issued and managed by a global central bank is compelling in the abstract. A series of ambitious IMF managing directors, seeking to expand the ambit of the institution, have suggested moving in this direction. But as a practical matter, so long as there is no global government to hold it accountable for its actions, there will be no global central bank. No global government, which means no global central bank, means no global currency. Full stop.
At most, one can imagine a limited role for the SDR in supplementing existing reserve holdings. Because the SDR is defined as a basket of currencies, accumulating SDR claims will be another way for central banks to modestly adjust their reserve portfolios in the direction of fewer dollars. Issuing SDRs has the attraction that doing so is cheap. Since the SDR is simply a bookkeeping claim, it costs no real resources to produce. Countries seeking additional reserves do not have to forgo consumption and run export surpluses in order to acquire them. This is also a way of limiting the exorbitant privilege of future reserve currency countries.
But limiting is not the same as eliminating. Central banks will hold only a fraction of their reserves in this form, since SDRs are not liquid or readily used in market transactions. The SDR will not replace national currencies in central bank reserves because it will not replace national currencies in other functions.
SYMBOLIC GESTURES

Governor Zhou is aware of these difficulties. He has been around for a long time, having worked his way up through a series of Chinese banks and helped to run the country’s State Administration of Foreign Exchange (SAFE). He attends the Jackson Hole retreat of the Federal Reserve Bank of Kansas City, where issues like the viability of the SDR as an international currency are discussed around the campfire.
One might ask, in light of this, what motivated Zhou to make his case for the SDR. One answer is that his SDR proposal was intended as a stalking horse for a Substitution Account through which the international community would take China’s dollars off its hands. The idea of an account at the IMF through which SDRs would be substituted for dollars on the books of central banks was first raised in the late 1970s, an earlier period of angst over the prospects for the greenback.39 It foundered then over the question of who would bear the losses on the dollars absorbed by the account. Since the U.S. government was not prepared to do so, the risk would have remained in the hands of IMF members as a group. But because it was those same IMF members who were anxious to get dollars off their books, this rendered the operation purposeless. It amounted to little more than shifting those dollars from one pants pocket to the other.
Recently there has been another flurry of interest in a 1970s-style Substitution Account to exchange SDRs for the dollars that central banks are anxious to sell.40 But again the proposal is certain to founder over the question of who will absorb the losses to the account if the dollar depreciates against other currencies. If the members of the IMF, which operates the account, take the losses, then it achieves nothing. Almost 85 percent of shares in the IMF are owned by countries other than the United States—the same countries that are anxious for a Substitution Account to take some of their dollars. In contrast, if the United States agrees to compensate the IMF for losses to the account, it would open itself up to a very large financial liability, which it is not willing to do. Governor Zhou is savvy enough to understand this.
A more compelling explanation for Zhou’s initiative is that he was engaged in symbolic politics. He wanted to signal China’s unhappiness with prevailing arrangements. By delivering his speech on the eve of a G20 economic summit, he reminded other countries that China’s views are to be reckoned with. By suggesting an enhanced role for the SDR, he was positioning China as an advocate of a rules-based multilateral system.
In addition Zhou was playing to his domestic audience. He was seeking to deflect criticism that the Chinese authorities, by failing to more actively seek out alternatives to the dollar, had not been careful stewards of their country’s international reserves.
WHAT CHINA IS AFTER

But perhaps the most fundamental reason that the SDR proposal will go nowhere is that China has a preferred alternative, namely establishing the renminbi as an international currency. Were the renminbi used widely in international transactions, China would be freed of having to hold foreign currencies to smooth its balance of payments or aid domestic firms with cross-border obligations. It could just print more or less of its own currency as called for, like the United States. It would enjoy all the advantages of a reserve-currency country.
It is clear that Chinese officials are thinking along these lines. To cite one example, Zhang Guangping, vice-head of the Shanghai branch of the China Banking Regulatory Commission, suggested to reporters in 2008 that the renminbi could become an international currency by 2020.41
But 2020 is a long way off. It is a long way off in that the renminbi will remain inconvertible for the foreseeable future. Inconvertibility means that foreigners can only use it to purchase goods from China itself, with a few exceptions. China permits the currency to be used in cross-border trade only with its immediate neighbors, countries like Mongolia, Vietnam, Cambodia, Nepal, and North Korea and the special administrative zones of Hong Kong and Macau. Even there only “select” trustworthy companies are permitted to settle their transactions in renminbi.42
These limitations are designed to prevent the value of merchandise imports and exports from being misstated as a way of circumventing China’s capital controls. If a Hong Kong resident wanted to smuggle money into China in order to invest in apartments, he could overstate the value of the goods he was importing from Guangzhou, inflate his payment, and on his next trip to Guangzhou recover the funds from the company he was in cahoots with. Or if a businessman in Guangzhou wanted to ferry money out of the country, he could overstate his payments to an exporter in Hong Kong.43 This is why only trustworthy importers and exporters are allowed to use the renminbi in cross-border trade. These restrictions insulate the Chinese economy from capital flow volatility. They allow Chinese officialdom to manipulate financial markets as they choose. But they also limit the renminbi’s international use.
Brazil and China made a splash in 2009 by announcing that they were exploring ways of using their own currencies in bilateral trade.44 But such explorations are mainly useful for advertizing the fact of that trade. What use would the typical Brazilian firm have for renminbi given that the Chinese currency cannot be converted into reais? A Brazilian firm will take renminbi for its exports only insofar as it imports from or seeks to invest in China—not your typical case. Brazil and Argentina reached a similar agreement to settle their bilateral trade in their own currencies in September 2008 but, revealingly, still use dollars in practice.
China is not Argentina. Its trade will continue to grow, and Chinese firms will encourage their customers to invoice and settle their transactions in renminbi, since doing so will protect them from currency fluctuations. Something analogous happened with Japanese trade in the 1980s. As Japanese firms acquired more bargaining power, they insisted that more of their exports be invoiced and settled in yen. Still, the share of Japanese exports invoiced and settled in yen never rose above 40 percent, the yen lacking the other attributes of an international currency. Given that the renminbi will likewise lack these attributes for the foreseeable future, it is not clear that it will do better.
Similarly, China’s currency swap agreements with Argentina, Belarus, Hong Kong, Indonesia, South Korea, and Malaysia are not so much practical measures as a way for it to signal its ambitions. Other central banks can’t use the renminbi to intervene in foreign exchange markets. They can’t use it to import merchandise from third countries or to pay foreign banks and bondholders. Contrast the $30 billion swap that the Bank of Korea received from the Federal Reserve in November 2008, which the Bank used to intervene in the foreign exchange market. China could become more consequential as a supplier of emergency credits if it offered other countries swap lines in dollars. But so much, then, for swaps as a device for enhancing the renminbi’s international role.
With time China can strengthen the international role of the renminbi by developing liquid securities markets and liberalizing access to them. With time it can make its currency freely usable for financial as well as merchandise transactions. The question is: how much time? China has been feeling its way in this direction for more than a decade yet even now has moved only part of the way down the path. With good reason: reconciling financial stability with full freedom to buy and sell domestic and foreign assets has formidable prerequisites. Markets must first become more transparent. Banks must be commercialized. Supervision and regulation must be strengthened. Monetary and fiscal policies must be sound and stable, and the exchange rate must be made more flexible to accommodate a larger volume of capital flows.
China, in other words, must first move away from a growth model of which bank lending and a pegged exchange rate have been central pillars. This is easier said than done. Witness that the Chinese authorities’ reaction to the 2008 crisis was to move in the opposite direction, ordering the banks to boost their lending and hardening the renminbi’s peg to the dollar to sustain exports. All the evidence suggests, then, that China’s move to more open financial markets will remain gradual.
SLOW BUT STEADY

Policy toward bond markets is a case in point. Until recently, renminbi-denominated bonds were sold only by Chinese banks and by multilateral banks like the World Bank and the Asian Development Bank, and only in China. The authorities were reluctant to allow foreign corporations to issue bonds, since doing so would have interfered with the government’s ability to channel savings to Chinese industry (shades of Japan in the 1970s). If foreign companies offered Chinese investors more attractive terms, their savings would not automatically flow to Chinese banks, to be lent out to enterprises that the government deemed worthy.
In the summer of 2009 HSBC Holdings became the first foreign bank to sell renminbi-denominated bonds in Hong Kong. The following September the Chinese government then issued 6.3 billion of renminbi-denominated sovereign bonds there. The equivalent being less than $1 billion, this was a drop in the bucket, but it was an indication of what is to come. Then in July 2010 Hopewell Highway Infrastructure, a Hong Kong-based highway construction firm, became the first company other than a bank to receive authorization to issue renminbi-denominated bonds offshore.
A market in renminbi-denominated financial instruments in Hong Kong is one thing. So long as financial markets in Hong Kong and the mainland are separated by administrative controls, the actions of foreign investors would not compromise the ability of the government to channel funds to Chinese industries of their choosing. But a market in renminbi-denominated bonds in Shanghai fully open to foreign issuers would be another matter. It would destroy the ability of the Chinese authorities to channel savings to domestic industry. Chinese savers would regard these bonds, with their returns guaranteed in domestic currency, as an attractive alternative to the captive bank deposits that are funneled into industrial development. The very foundations of the Chinese development model would be threatened.
That said, Chinese policymakers are serious about transforming Shanghai into an international financial center by 2020. Doing so will require deeper and more liquid markets. It will require liberalizing the access of foreign investors to those markets, which will in turn imply other changes in the country’s tried-and-true growth model. Liberalizing the access of foreign investors to China’s financial markets will in turn require a more flexible exchange rate to accommodate a larger volume of capital inflows and outflows.45 While these are not changes that can occur overnight, it is worth recalling how the United States moved in less than 10 years from a position where the dollar played no international role to one where it was the leading international currency. There is precedent, in other words, for the schedule that the Chinese authorities aspire to meet.
The creation of the Federal Reserve System and a market in dollar acceptances in New York were major institutional changes, but the changes to the Chinese economy required to make the renminbi convertible would be even more far-reaching. Whether China can complete them in as few as 10 years is an open question.
The other reason that 2020 seems a bit ambitious for elevating the renminbi to reserve-currency status is that even if China grows at a 7 percent annual rate for the next decade (slower than in the past, reflecting less favorable demographics, but still exceptional by historical standards), its GDP in 2020 will still be only half that of the United States at market exchange rates—market rates being what matter for international transactions. The renminbi will still have a smaller platform than the dollar from which to launch its international career. The liquidity of markets in renminbi will still not be comparable to markets in dollars.
This means that the share of reserves in renminbi will be limited. Renminbi reserves will be most attractive to countries trading heavily with China and doing their financial business there. They will be most attractive to countries for which fluctuations of the renminbi on the foreign exchange market matter most. This suggests that the practice of holding renminbi reserves will be disproportionately concentrated in Asia, much as the practice of holding euro reserves will be disproportionately concentrated around Europe.
Someday, perhaps, the renminbi will rival the dollar. For the foreseeable future, however, it is hard to see how it could match the currency of what will remain a larger economy, the United States. Regional reserve currency? Yes. Subsidiary reserve currency? Yes. But dominant reserve currency? This is harder to imagine.46
GOING BUGGY

Finally, there are some minor alternatives to be dismissed. Gold has its bugs. They argue that if there is a loss of confidence in the dollar—or even if there isn’t—gold is an obvious asset for international investors, including central banks, to scramble into. In practice, of course, central banks have been scrambling in precisely the opposite direction for the better part of a century. Where gold accounted for nearly 70 percent of central banks’ international reserves in 1913, its share today is barely 10 percent. In every year since 1988, central banks have been selling, not buying, gold.
Why they have done so is clear. Financial instruments are more convenient for emergency financial transactions. When a currency is under pressure and the central bank is forced to support it, it is simpler just to buy that currency for dollars than to first sell gold in order to obtain the requisite dollars. Gold is not a convenient instrument with which to purchase imports—how many foreign companies will be as inclined to accept it as they are a dollar-denominated check? It is not a convenient instrument with which to pay interest on foreign debts or to engage in other financial transactions.
The exceptions prove the rule. Late in 2009 the Reserve Bank of India bought 200 tons of gold from the International Monetary Fund.47 This raised the share of gold in the Reserve Bank’s reserve portfolio from 4 to 6 percent. Notice, however, how this was a sale of gold by an institution, the IMF, with reason to expect that it might actually have to use its financial resources. Starting in 2008 the IMF had to leap in with emergency financial packages for Hungary, Iceland, Latvia, and Ukraine. In response to the financial crisis, it needed to provide them with cash—actual dollars and euros. Shipping them gold bars wouldn’t have been convenient.48
India, on the other hand, has a flexible exchange rate, restrictions on capital inflows and outflows, and a relatively sound banking system. At the moment its central bank has little need to use its foreign reserves in market transactions. Given worries about whether the dollar will hold its value, it therefore made sense for the Reserve Bank to modestly increase the share of its portfolio in gold.49
Will other central banks follow? In the wake of India’s high-profile transaction, there were small purchases of gold on local markets by Venezuela, Mexico, and the Philippines and from the IMF by the central banks of Sri Lanka and Mauritius. Russia’s central bank has reportedly purchased limited amounts of gold. But in early 2010 the IMF was forced to acknowledge that it had been unable to find buyers for the remaining 200 tons of gold it wished to sell. So much, then, for the mass migration of central banks into gold.
Again the big player, potentially, is China. Since 2008 China has added modest amounts of gold to its portfolio. Were it to move further in this direction, it could have important implications for the role of gold in central bank portfolios. But once more China faces the dilemma that its reserves, the majority of which are in dollars, are so large. Selling significant quantities of dollars for gold would push down the value of the greenback, creating losses on the dollars that the central bank has not yet sold and pushing up the cost of Chinese exports. Doing so would also push up the price of gold and hence the cost of obtaining more.
Gold bugs are forever. But it is not obvious that one can say the same for the monetary role of gold.
TIMBER

The other oft-mooted possibility is real assets. Several countries, China among them, have transferred a portion of their reserves to sovereign funds that invest in timber acreage, oil reserves and refineries, and other real assets. It might be possible for the United States to inflate away the value of its treasury securities, the argument goes, but not the value of timberland in New Zealand or petroleum in West Africa.
Again, this is a fine strategy for countries with more foreign reserves than they will ever use. It is unlikely that China will ever have to use more than a fraction of its $2.5 trillion of reserves to intervene in foreign exchange markets or to recapitalize banks with dollar liabilities. It makes perfect sense to lock up a share of those reserves in real assets. That those investments are illiquid is no big deal since there is little prospect that their Chinese owners will find themselves having to sell them in an emergency. It is no coincidence that Norway, which has accumulated more foreign assets through sales of natural gas than it will ever need to intervene in financial markets, was among the first to pursue this strategy, creating a sovereign wealth fund to invest in real assets.
But the situation is different for countries that foresee circumstances in which they might actually have to use their reserves. Selling timberland for dollars in a financial emergency is even more difficult than selling gold. Just ask Harvard University, whose endowment fund bought up large tracts of timber in New Zealand in the years leading up to the crisis. At 2006 valuations, timber accounted for 12 percent of the Harvard endowment’s real estate portfolio. This created huge difficulties in 2008 and 2009 when the fund needed cash and couldn’t easily sell its timber acreage. This is not a strategy for investors who value liquidity. It is not an investment strategy for central banks.
WEALTH OF ALTERNATIVES

So where does this leave us? It leaves us with the prospect of multiple international currencies. A world of multiple international currencies is coming because the world economy is growing more multipolar, eroding the traditional basis for the dollar’s monopoly. Once upon a time, after World War II, when the United States dominated the international economy, it made sense that the dollar should dominate international monetary and financial affairs. The United States dominated world trade. Only it had deep and liquid financial markets. As an international currency, the dollar had no rivals.
Today more countries are consequential traders. More countries have liquid financial markets. The shift away from a dollar-dominated international system toward this more multipolar successor was accelerated, no doubt, by the 2008 crisis, which highlighted the financial fragility of the United States while underscoring the strength of emerging markets. But even before the crisis, it was clear that the tension between a multipolar economic world and a dollar-dominated international monetary system would have to be resolved. And even then, there was little question about the form this resolution would take.
The speed with which this world of multiple international currencies arrives will depend on the advantages of incumbency. Recall the argument that the competition for international currency status is subject to status quo bias. It pays individual exporters and bond issuers to use the same currency as other exporters and bond issuers. This works to maintain the status quo. And the status quo choice is the dollar.
But this mechanism is unlikely to carry the same weight in the future as the past. Once upon a time it made sense for importers, exporters, and bond underwriters to use the same currency as other importers, exporters, and bond underwriters, since doing otherwise could create confusion for their customers. The difficulty of obtaining up-to-date information on the value of different currencies and the high costs of switching between them meant that it paid to use the same unit as everyone else. The currency in which everyone transacted had the most liquid markets as a result of the fact that everyone transacted in it. The dominance of the leading currency was self-reinforcing.50 International currency status was a natural monopoly, like municipal water supply or electricity.51
The twenty-first century is different. Everyone now carries in his pocket a device capable of providing the real-time information needed to compare prices in different currencies. Comparing the prices of bonds or the cost of trade credit in dollars and euros is no longer a problem. Changes in technology have allowed for freer competition in other industries long assumed to be natural monopolies, like electricity and telephony. Why should international finance be any different?
In addition, the sheer size of the twenty-first-century world economy means that there now is room for more than one market with the liquidity that makes for low transactions costs. Again, the natural-monopoly argument for why there should be only one consequential international currency has become less compelling as a result of economic and financial development.
Forecasts are risky, especially when they involve the future. But there is little uncertainty about the identities of the leading players in this new, more multipolar system. The dollar, the currency of the single largest economy with the most liquid markets, will remain first among equals. The euro, the currency of a monetary zone whose economic size approaches that of the United States, will become more attractive, particularly on the periphery of Europe, if not now, then once Europe has sorted out its problems. China, for its part, is already encouraging select nonresidents to use its currency to invoice and settle merchandise transactions. In the not-too-distant future, perhaps in as few as 10 years, the renminbi will be an attractive unit, especially in Asia, for use by international investors and central banks.
While the dollar, the euro, and the renminbi will be the leading international currencies, they may not have the field to themselves. The same arguments suggesting that there is room for three international currencies suggest that there is room for more than three. The additional candidates are not likely to be the currencies of demographically challenged countries like Japan and Russia. Size matters for the depth and liquidity of financial markets. And assuming the continued convergence of living standards, population will be a key determinant of economic size.
This points to India’s rupee and Brazil’s real as additional runners. Like China, both India and Brazil have work to do before their currencies are used internationally. Like China, they continue to restrict foreign participation in their financial markets, thereby limiting the attractions of their currencies for international use. Like China, their financial systems are bank based: they have a way to go in building liquid markets in the bonds and bills that central banks and other international investors find attractive. Still, their favorable demographics suggest that their currencies, like China’s, may acquire growing roles. That their economies are smaller than China’s and that they engage in less foreign trade and investment suggest more limited international use of their currencies. But they are coming.
None of this means that the dollar will lose its international currency status, only that it will have rivals. It will have to compete for business because exporters and investors will have a growing range of alternatives. That said, there is no reason that it shouldn’t succeed at that competition—barring a homegrown economic disaster of the first order.