DEBT, EQUITY, AND EINANCIAL OPENNESS'

Will the Sovereign Debt Market Survive?

By ANDREI SHLEIFER*

The market for sovereign debt of developingcountries is alleged to be gravely ill. In the lastfive years, several East Asian countries, Russia,Argentina, and other borrowers defaulted ontheir sovereign bonds. Financial crises in bor-rower countries, as well as prolonged, messyand litigious debt renegotiations accompaniedthese defaults. Some see the so-called "roguecreditors." who sue to get their money back anddelay settlement, as the culprit of the market'sproblems (Michelle White, 2002), To con-front them, reformers propose to change debt-restructuring procedures by reducing creditorrights. The IMF in particular advocates usingsome features of U.S. Chapter U corporatebankruptcy that are favorable to borrowers insovereign debt restructuring (Anne Krueger,2002).

This paper uses the experiences with bank-ruptcy in other debt markets to reexamine sov-ereign debt restructuring. I argue that theailment of sovereign-debt markets has beenmisdiagnosed. Experience shows that debt mar-kets work best when the rights of creditors areprotected most effectively. From this perspec-tive, the troubles afflicting sovereign-debt mar-kets result from creditor rights being too weak,not too strong.

There is a more evocative way to put thispoint. When he was Secretary of the Treasury,Lawrence Summers asked: Why can U.S. mu-nicipalities borrow at such low rates, and have

* Discu.'i.'ianl.s: Kenneth Froot. Harvard University: Ken-iieih RogotT, International Monetary Fund; Robert Shiller,Yale University.

* Deparlinent of Economics, Harvard LIniversity, Cam-bridge, MA 02138, Thanks to Olivier Blanchard, JeremyBulow, Mihir Desai, Ken Froot. Rubin Greenwood, OliverHan, Robert Inman, Anne Krueger, Rafael La Porta, Fk)-rencio Lopez-de-Siianes, Randall Picker, Michael Rashes.Lawrence Summers, Michelle White, and Nancy Zimmer-man for helpful comments.

such low rates of default, compared to sover-eign borrowers? How can the sovereign debtmarket become more like U.S. municipal bondmarket, with few defaults and low interestrates? The answer, I believe, is that creditorrights vis-a-vis defaulting U.S. municipalitiesare much greater than those of lenders vis-a-vistheir defaulting sovereign borrowers. As a con-sequence, defaulting municipalities repay theircreditors in full or almost in full, and municipalbond markets function well.

The objective of the paper is to consider howsovereign-debt markets can be improved. Someacademics believe that the problems of sover-eignty doom these markets. Except for verygood borrowers who care about their reputation,the markets should not really exist (JeremyBulow, 2002). According to this view, existingdebt should be forgiven and replaced with for-eign aid. This view may well be correct. Still, inthis paper, I focus on the possibility of improv-ing the sovereign-debt market rather than writ-ing it off.

I. The Structure of Debt Contracts

Debt is a contract, in which the borroweraccepts some money and agrees to pay it back.If the borrower fails to repay, the creditor ac-quires certain rights and powers vis-a-vis hisassets. Except for some possible reputationalconcerns, those rights are the main reason thatborrowers ever repay loans (Oliver Hart, 1995).

A typical debt contract is collateralized. Thelender's most basic power is to seize the cot-lateral of a defaulting borrower. This powersupports the mortgage market in the UnitedStates, much of personal lending, and standardliquidation-foe used bankruptcy procedures inmost countries.

For several good economic reasons, grabbingand liquidating collateral can be inefficient. Asa consequence, softer bankruptcy procedures

85

86 AEA PAPERS AND PROCEEDINGS MAY 2003

have emerged in many debt markets. In thepersonal-debt market, some assets of the bor-rower are exempt from being seized by thecreditors. Likewise, many legal jurisdictions re-place liquidation with reorganization, in whicha plan for the borrower to pay back the lender isdeveloped under the supervision of a court,avetting the seizure of assets.

In the United States, this alternative bank-ruptcy procedure for corporations is known asChapter 11. In Chapter 11, the management of adefaulting company has a few months to presentthe coutt with a reorganization and repaymentplan. During this period, management retainscontrol, creditors cannot grab assets, and thecompany can borrow fresh capital to continueoperations. If the creditors do not accept themanagement's plan, they can propose an alter-native, which might seek to replace the man-agement or to liquidate the company. Crucially,under the law, the objective of the court is tochoose a plan "in the best interest of thecreditors."

In many U.S, states, municipalities are unlikecorporations in that they cannot seek bank-ruptcy protection from creditors at all. In a fewinstances, municipalities can seek (and havesought) such protection, known as Chapter 9.This municipal-bankiTiptcy procedure followssome of the principles of Chapter 11, but furtherweakens creditor rights. Municipal governmentis not replaced, and the creditors are not allowedto propose competing plans. At the same time,the law continues to specify clearly that anyrepayment plan must be "in the best interests ofthe creditors" (Michael McConnell and RandalPicker, 1993). In addition, state governmentsin the United States appear to take responsi-bility for municipal debts, and to interveneaggressively in the fiscal affairs of defaultingmunicipalities. The work of the MunicipalAssistance Corporation in New York City andthe assistance to Orange County by the Cali-fornia state government both show that de-faulting local governments lose much of theirfiscal sovereignty.

The U.S. Chapter 11 bankruptcy (and relat-edly municipal Chapter 9 bankruptcy) is one ofthe softest (nicest to debtors) bankruptcy proce-dures in the world (see e.g., Rafael La Porta etal., 1998). Even the possibility of reorganiza-tion, as opposed to liquidation, does not exist

in some advanced-market economies. In theUnited States, management can unilaterallyseek Chapter 11 protection from creditors,whereas in many other countries creditor con-sent is required to initiate bankruptcy. In theUnited States, secured creditors cannot grab as-sets (there is an automatic stay on assets), whichthey can do in about half of the developedcountries. Perhaps most unusually, in theUnited States, management stays after filing forbankruptcy. Consistent with this picture, moststudies of the U.S. Chapter 11 bankniptcy see itas too slow, too soft on the existing manage-ment, and often ineffective in setting the com-pany on the path to economic recovery (EdithHotchkiss, 1995; Matthias Kahl, 2002a, b).

Yet even in Chapters 11 and 9, creditors havelegal rights counterbalancing the softness of theprocedure. In Chapter 11, the managementmakes its proposal under the threat of alterna-tive proposals from creditors that might involveits removal or liquidation of the company. Fur-thermore, in evaluating proposals in both Chap-ters 11 and 9, courts have a clear obligation tochoose one in the best interest of the creditors.

These powers of the creditors are the mainreason that borrowers pay back and debt mar-kets exist and prosper. The more the creditorscan recover from a defaulting borrower, and themore cheaply they can do so, the more they willlend to begin with, and on more attractive terms.

The sticks in the form of creditor rights arenot the only reason why borrowers pay back.There are also carrots that encourage borrowersto settle with creditors. The main carrot in thecases of persona! and municipal bankruptcy isthe ability to borrow again. For corporations,the benefits of settling with creditors and emerg-ing from bankruptcy include avoidance of liq-uidation, the ability of managers to keep theirjobs and get paid, and the resumption of divi-dends and unrestricted borrowing. The combi-nation of sticks and carrots allows the variousdebt markets to function.

II. Creditor Rights and Debt Markets

In the data, the stronger are creditor rights inthe event of a default, the better developed arethe debt markets. In the case of personal bank-ruptcy, creditor rights differ across the U.S.states. Some states have a much larger personal

VOL. 93 NO. 2 DEBT. EQUITY. AND FINANCIAL OPENNESS

exemption (the assets beyond the reach of cred-itors in the event of personal bankruptcy) thando others. The available evidence shows that,looking across states, higher exemptions reducethe amount of credit available to low-assethouseholds (Reint Gropp et al., 1997), raise thecosts and reduce the availability of mortgages(Emily Lin and White, 2001), and reduce smallfirms' access to credit (Jeremy Berkowitz andWhite, 2002).

La Porta et al, (1997, 1998) examine therelationship between creditor rights and the sizeof debt markets in 49 countries. They constructan index of creditor rights based on legal rulesgoverning the automatic stay on assets, the re-spect for seniority of creditors, the restrictionson going into reorganization, and the replace-ment of management in bankruptcy. The au-thors present evidence suggesting that countrieswith stronger creditor rights have larger debtmarkets (see also Rene Stulz and Rohan Wil-liamson, 2002), The evidence also shows thatgreater legal shareholder rights are associatedwith broader and more valuable equity markets.

The few existing municipal bankruptcies il-lustrate a similar pattern. Courts expect credi-tors to be repaid, and generally speaking, theyare repaid in full or nearly in full, and with onlya short delay. This is done through municipaltax increases, expenditure cuts, and assistancefrom state governments in the form of cash orguarantees of long-term bonds. Because theyknow they have to repay, municipalities bor-row moderately, and lenders provide funds atlow rates. There is even insurance againstmunicipal default, pointing to the insignifi-cance of moral-hazard problems. In light ofthe evidence, the answer to Summers's ques-tion is straightforward: The municipal bondmarket in the U.S. functions so well, with lowspreads and few defaults, because creditors'rights are sufficient to get paid in full after adefault.

III. Sovereign Debt Markets

Sovereign debt markets could not be moredifferent. Some countries borrow massively, of-ten to the point where both they and the lendersknow they cannot pay back, and at interest ratesthat refiect the high likelihood of default. Theydefault with some regularity, and when they do.

often pay back a fraction of what they haveborrowed, even when they have the ability topay, as was the case in Russia in 1998, Debtrenegotiation takes years and is accompaniedby massive litigation. In short, in this market,unlike in the municipal bond market, overbor-rowing, default, and limited repayment arecompletely normal and expected by both bor-rowers and lenders.

As Bulow and Kenneth Rogoff (1989a, b)showed in now-classic papers, this is exactlywhat one would expect in a market where lend-ers have no power. Because the borrowers insovereign debt markets are, well, sovereign,creditors have virtually no rights. Creditors can-not grab assets in the country. The most theycan do is to seize a few airplanes or barges ofoil, which does not get them far except as astrategy of harassment. Management (i,e., gov-ernment) is not replaced. Countries default uni-laterally, without the consent of creditors. Lastbut not least, there are no courts with authorityover sovereign states whose mandate is to pro-tect the interest of creditors. Bulow and Rogoffconclude by suggesting that there is no goodreason why the sovereign debt market exists.

The situation is not quite this dire, since atleast some sticks and carrots are available. Onepower of the creditors, recognized by Bulowand Rogoff, is to disrupt and litigate, sincesovereign-debt contracts are signed under inter-national law. As importantly, sovereign borrow-ers settle with their creditors because they wantto borrow again, and they run the risk that thefresh loans are seized by the old creditors unlessthere is a settlement. Under the current system,sovereign borrowers in default continue someborrowing from the IMF and in local marketsunder domestic law. To resume borrowing ininternational markets, countries settle with cred-itors and pay them more than zero. Indeed, localfinancial markets interpret settlement with cred-itors as very good news (Serkan Arslanalp andPeter Henry, 2002).

The development community is quite un-happy with the status quo, and in particular withthe litigation and disruption surrounding sover-eign default. Such litigation delays settlement,possibly prolonging recessions and raising thecost of IMF programs. There is also a concernthat the payment to creditors is in part subsi-dized by the IMF loans, which puts political and

AEA PAPERS AND PROCEEDINGS MAY 2003

economic pressure on the Fund to demand thatburdens be shared.

In making reform proposals, the IMF has twoadmirable objectives. The first is for sovereignstates to be able to borrow at low rates in privateinternational markets to finance their develop-ment. The second is for sovereign states not topay their creditors back, especially during thetimes of economic distress. The tension be-tween these objectives explains the challengesof reform.

At least superficially, the current proposalsfocus on making sovereign bankruptcy morelike the U,S. Chapter 11. There are three keyelements of the current IMF proposal. The firstinvolves collective-action clauses in debt rene-gotiations, the purpose of which is to make itmore difficult for "rogue" or "maverick" credi-tors to opt out of proposed settlement agree-ments and litigate. The IMF indeed argues thatthe goal of its proposals is to limit the rights ofthese minority creditors, rather than of all cred-itors. The second feature is the creation of asovereign bankruptcy court, which at the timeof the initial proposal was intended to be theIMF itself, but is now envisaged as an indepen-dent agency. Third, in parallel to U.S. Chapter11, the sovereign state in default will be allowedto borrow in arrears so that its economy is notdisrupted, and the new debt will be senior to theexisting debt.

In the discussions of the IMF proposals,collective-action clauses have received by farthe most attention. Nouriel Roubini (2002) ex-plains persuasively why the problem of "roguecreditors" is vastly overstated. Lenders are al-ready under tremendous pressure to settle withsovereign borrowers, since any lender who doesnot agree to a proposed debt exchange or othersettlement risks receiving nothing. Collective-action clauses increase the pressure to settle onthe borrower's terms. Litigation is one of thefew (relatively weak) powers the lenders have ifthey do not like these terms. Indeed, many ofthe recent settlements, unlike those for munici-pal debt in the United States, take place onterms highly favorable to the borrowers. This iswonderful ex post for countries in financial cri-sis, but unlikely to be conducive to the long-term health of the sovereign debt market.

The creation of a bankruptcy court can be avery beneficial development for the sovereign

debt market if, like the courts in U.S. municipalbankruptcy, the objective of settlement is to bestserve the interests of creditors. Of course, theobjective of the IMF is exactly the opposite—again admirable ex post, but incompatible withlong-term health of the market. It wants thecourt to protect the country from creditors ratherthan to serve their interests.

But perhaps the crucial, and most neglected,feature of the IMF proposal is a radical increaseof lending into arrears. The development objec-tive here is clear and laudable: such borrowingcan finance exports and deficit spending andthereby relieve economic hardship. But basiclogic, as well as the evidence discussed in theprevious section, suggests that this refonn islikely to damage the market by eliminating thecrucial carrot that allows recovery by creditors.If the country can borrow fresh funds withoutrepaying old debt, it has no incentive to paycreditors anything and, in fact, has every reasonto remain in "sovereign bankruptcy" forever.The crucial features of U.S. Chapter 11 whichcounterbalance borrowing in arrears (the threatsof liquidation or of acceptance of the creditors'reorganization plan by the judge) do not existfor sovereign debt. Without these threats, andwith the possibility of continued borrowing se-nior to the old loans, there is no reason to repay.And with no reason to repay, there is no sover-eign debt market in the long run.

In summary, all the proposals by the IMFshare a common element: they reduce creditorrights. These ideas serve the admirable goal ofreducing the burden of default on debtor coun-tries. Yet, by cherry-picking tbe features ofChapter 11 most favorable to creditors, theseproposals ignore the delicate balance betweendebtor protections and creditor rights that existsin both Chapter 11 and Chapter 9. As all theevidence indicates, enhancing borrower rightswhile emasculating creditor rights is unlikely topromote the health of emerging debt markets.

IV. Alternatives

Sovereign debt markets are likely to shrink ifthe IMF's proposals are adopted, at least fordeveloping countries. Indeed, governments ofseveral significant emerging-market borrowersoppose these proposals, precisely because they

VOL 93 NO. 2 DEBT, EQUITY, AND FINANCIAL OPENNESS

recognize the consequences of weakening cred-itor rights for their ability to borrow in thefuture. Still, there may well exist good strategiesfor reform. Such strategies must look at suc-cessful institutions in their entirety, includingboth debtor protections and creditor rights.

Traditional collateralized lending, supportedby the creditors' right to seize collateral, canprobably be expanded even in emerging mar-kets. Countries can pledge their natural-resourceexport revenues, for example, especially whenthere is no practical way to stop production andexport. Mexico successfully pledged a portionof its oil revenues to the United States duringthe 1995 rescue, and there is no reason why thisexperience could not become more common.

Relatedly, the IMF could encourage sover-eign issuance of securities whose returns arelinked to commodity prices or economic per-formance. Such equity-like securities wouldrequire lower payments during economic hard-ship, and higher payments in times of prosper-ity. They would avoid many of the problems ofself-fulfilling crises that the IMF is concernedwith.

Suppose, however, that the reformers wish topursue bankruptcy procedures similar to Chap-ter 9 or 11, the concern about the softness ofthese procedures notwithstanding. These re-forms could include collective-action clausesand lending into arrears. If they do, they couldonly work if counterbalanced by other featuresof Chapters 9 and 11, The most crucial of theseis the presence of a court with power to enforceits decisions and a duty to find solutions thatbest serve the interests of creditors. To be ef-fective, such a court must stay away from pur-suing development objectives. It should alsohave jurisdiction over both sovereign borrowersand their lenders, including the IMF.

If the analogy to Chapter 11 is taken seri-ously, creditors should also be able to present tothe court restructuring plans that compete withthe proposals of the borrowers (perhaps goingsecond). As long as these proposals are feasible,they should be considered by the court, whichshould select and enforce the plan "in the bestinterest of the creditors." Likewise, if the devel-opment community wishes to imitate the successof the U.S. municipal bond market by pursu-ing a procedure like Chapter 9, mechanisms oftakeover of national finances by an international

authority must be considered. Ricardo Cabal-lero and Rudiger Dombusch (2002) make pre-cisely such a proposal for Argentina.

In the end, a supra-national bankruptcy courtmay or may not be feasible. But such a court isessential if sovereign default and bankruptcy areto proceed as smoothly as the IMF wants. Thesuccess of the municipal bond market in theUnited States suggests that a bankruptcy proce-dure for a government can work well. But thefoundation of such a procedure is the existenceof a credible arbiter whose mandate is to protectcreditor rights. Without exception, the powersof such an arbiter in every kind of bankruptcyentail some loss of sovereignty by the borrower.Without such an arbiter, however, the looseanalogies to Chapter 11 lose credibility. Itwould be better to leave the market alone. De-spite all the turmoil, it has delivered vastamounts of private capital to the developingworld.

REFERENCES

Arslanalp, Serkan and Henry, Peter. "Debt Re-lief: What Do the Markets Think?" Mimeo,Stanford Business School, 2002.

Berkowitz, Jeremy and White, Michelle. "Bank-ruptcy and Small Firms' Access to Credit."National Bureau of Economic Research(Cambridge, MA) Working Paper No. 9010,2002.

Bulow, Jeremy. "First World Governments andThird World Debt." Brookings Papers onEconomic Activity, 2002, (1), pp. 229-55.

Bulow, Jeremy and RogoiT, Kenneth. "A Con-stant Recontracting Model of SovereignDebt." Journal of Political Economy, Febru-ary 1989a, 97(1), pp. 155-78,

"Sovereign Debt: Is to Forgive to For-get?" American Economic Review, March1989b, 79(1), pp. 43-^50.

Caballero, Ricardo and Dombusch, Rudiger."Argentina: A Rescue Plan that Works."Mimeo, Massachusetts Institute of Technol-ogy, 2002.

Gropp, Reint; Scholtz, John Karl and White,Michelle. "Personal Bankruptcy and CreditSupply and Demand." Quarterly Journal ofEconomics, February 1997, 112(1), pp. 217-52.

AEA PAPERS AND PROCEEDINGS MAY 2003

Hart, Oliver. Firms, contacts, and financialstructure. Oxford, U.K.: Oxford UniversityPress, 1995.

Hotchkiss, Edith. "Postbankruptcy Performanceand Management Turnover." Journal of Fi-nance, March 1995, 50(1). pp. 3-21.

Kahi, Matthias. "Economic Distress, FinancialDistress, and Dynamic Liquidation." Journalof Finance, February 2002a, 57(1), pp. 135-68.

"Financial Distress as a SelectionMechanism: Fvidence from the UnitedStates." Mimeo, University of California-Los Angeles, 2002b.

Krueger, Anne. "A New Approach to SovereignDebt Restructuring." Unpublished manu-script. International Monetary Fund, Wash-ington, DC, 2002.

La Forta, Rafael; Lopez-de-Silanes, Florencio;Shleifer, Andrei and Vishny, Rohert. "LegalDeterminants of Fxternal Finance." Jour-nal of Finance, July 1997. 52(3), pp. 1131-50.

"Law and Finance." Journal of Polit-ical Economy, December 1998, 70(5(6), pp.1113-55.

Lin, Emily and White, Michelle. "Bankruptcy andthe Market for Mortgage and Home Improve-ment Loans." Journal of Urban Economics,July 2001,50(1), pp. 138-62.

McConnell, Michael and Picker, Randal. "WhenCities Go Broke: A Conceptual Introductionto Municipal Bankruptcy." University of Chi-cago Law Review, Spring 1993, 60(2), pp.425-495.

Roubini, Nouriel. "Do We Need a New Bank-ruptcy Regime?" Brookings Papers on Eco-nomic Activity, 2002, (1), pp. 321-33.

Stulz, Rene and Williamson, Rohan. "Culture,Openness, and Finance," Dice Center Work-ing Paper No. 2001-2, Ohio State University,2002.

White, Michelle. "Sovereigns in Distress: DoThey Need Bankruptcy?" Brookings Paperson Economic Activity, 2002, (1), pp. 287-319.