Documents

Speculative Attacks and Risk Management

Categories
Published
of 38
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Description
Speculative Attacks and Risk Management
Transcript
  Speculative Attacks and Risk Management ∗ Parag A. Pathak † and Jean Tirole ‡ August 7, 2006 Abstract The paper builds a simple, micro-founded model of exchange rate management, spec-ulative attacks, and exchange rate determination. The country may defend a peg in anattempt to signal a strong currency and thereby boost the government’s future re-electionprospects or attract foreign capital. The paper relates the size of the speculative attackand government’s defense strategy to the market’s prior beliefs about the strength of thecurrency, the ability of foreign speculators to short sell the currency, domestic politics, andinitial debt composition. Speculative activities can exhibit strategic complementarity or sub-stitutability. Finally, features of srcinal sin covary with the maintenance of pegs, as lettingresidents hedge the currency or incentivizing them to lengthen their debt maturity structureis an admission that the currency is overvalued and undoes the signal sent by defending thecurrency. Keywords:  Speculative attacks, hedging, financial crises. JEL class:  D82, F33, F34. ∗ We are grateful to Alberto Alesina, Olivier Jeanne, Henri Pag`es, Ken Rogoff, and Jaume Ventura and toparticipants at the Lehman Brothers FMG-LSE Liquidity conference, the 4th Annual IMF Research Conference, aFondation Banque de France conference, and seminars at Toulouse and Harvard Universities for helpful comments.For financial support, Pathak thanks the Bourse Chateaubriand, the National Science Foundation, and the Divisionof Research at Harvard Business School and Tirole thanks the Fondation Banque de France. † Harvard University, e-mail: ppathak@fas.harvard.edu, www: http://www.fas.harvard.edu/ ∼ ppathak ‡ IDEI and GREMAQ (UMR 5603 CNRS), Toulouse and MIT, e-mail: tirole@cict.fr, www:http://idei.fr/member.php?i=3.  “The president who devalues is devalued.”–Mexican President Jos´e L´ opez Portillo, 1984 1 1 Introduction While no two financial crises are identical, most recent ones 2 share (at least) three keyfeatures: i) currency peg: The exchange rate is fixed at an ambitious and ultimately unsus-tainable rate; 3 ii) poor risk management: The corporate sector (banks and firms) is overlyexposed to a depreciation of the local currency, and thus suffers from “srcinal sin”; 4 (iii)sudden stop: Capital inflows, initially large, rapidly come to a sudden halt. 5 These familiar observations raise a host of nagging questions, such as: (i) Why do somany countries defend pegs (and floaters manage their exchange rate 6 ) if capital mobilitymakes exchange rate management hazardous? (ii) Why is the corporate sector exposedto exchange rate risk and the concomitant threat of facing liquidity shortages? Or, evenmore basically, why do domestic residents not enter into insurance contracts with foreignersin which the latter would deliver dollars in bad times and receive dollars in good times?(iii) Why do international financial institutions not take more advantage of the post-crisisdepreciated exchange rate to invest in the country?This paper suggests a common signaling hypothesis as a potential answer to these ques-tions. It builds a simple, micro-founded model of exchange rate and corporate risk manage-ment and speculative attacks on the following premises:a) The domestic government is privately informed about variables, including its own politicalintentions, that affect the future exchange rate: the level of reserves (broadly defined toaccount for the State’s off-balance sheet liabilities or the quality of reserves tied in contractsor in commodities); the political support necessary to sustain the currency’s value; thegovernment’s willingness to implement structural reforms, deal with corruption or protectproperty rights; or the level of fiscal needs.b) The government has preferences over the market’s exchange rate expectations. This paperanalyzes the case in which the government benefits from the market’s perception of a strongcurrency, because domestic voters infer that the country is well-managed (which we will usefor illustration) or this facilitates its firms’ access to the international capital market (seesection 7). As we discuss in the conclusion, the government’s desire for the perception of  1 Quoted in Kessler (2000). 2 E.g., Mexico (1994), Southeast Asia (1997), Russia (1998), Brazil (1999) and Argentina (2001). 3 See, for example, Fischer (2001) and Summers (2000). 4 See Eichengreen and Hausmann (1999). 5 See Calvo and Reinhart (2000). 6 See Calvo and Reinhart (2002). 2  a strong currency, which we endogenize, is of course not the only case of interest, but itis certainly the relevant one for countries that underwent financial crises and motivate thispaper.c) Maintaining the peg is costly to a country with an overvalued currency as it must sell theforeign currency at an unfavorable rate, which will later on reduce the country’s standardof living. Alternatively, delaying a devaluation will result in even higher bankruptcy costsor more interrupted investment projects in the future.d) Exchange rate management is only one component of a  cluster of policy signals  . Thegovernment’s equilibrium policy choices ought to be coherent in that it is irrational for thegovernment to expend resources on policy A in an attempt to signal a strong currency and tosimultaneously undo the signal through policy B. In this spirit, we endogenize the domesticresidents’ ability to hedge foreign exchange risk and, thereby, the source of exchange ratevolatility (models of currency risk usually implicitly assume that hedging contracts are forsome reason limited). We ask whether the government optimally facilitates or hinders suchhedging. Similarly, we ask whether the government would like to use policy to alter theprivate sector’s liability maturity structure.The paper’s main economic insights are:(1) We identify two wealth effects which are factors of strategic complementarity (SCmeans that a larger attack by other speculators increases one’s incentive to attackthe currency) or substitutability (SS). Speculation has an  immiserizing   effect on acountry with weak fundamentals and a  windfall gain   effect on a country with strongfundamentals. The former is conducive to SC and the latter to SS.(2) Hedging is endogenously incomplete. To be credible in its attempt at convincing themarket of the currency’s strength, the government cannot encourage, and actually mustdiscourage hedging by residents. Letting the residents hedge is a clear admission thatthe currency is overvalued by the market and makes any complementary attempt atexchange rate management futile.(3) A peg may make domestic borrowers eager to issue short-term liabilities so as to provideforeign investors with an advantageous exit option. Furthermore, the government doesnot incentivize firms to lengthen the maturity structure even when it wants to, becausedoing so would again be an open admission of a future depreciation.(4) Performing comparative statics with respect to the government’s objective function,the model predicts that pegs are more likely to be maintained before an election, andless likely to be maintained, the larger the tradable goods sector.Our predictions are largely supported by available empirical evidence. While successfulattacks often capture the headlines, episodes of speculative pressure can also lead to exchangerate appreciation, consistent with the wealth effects we identify. Indeed, if all attacks were 3  guaranteed to be successful, then a strategy of always attacking a peg would be an arbitrageopportunity. Eichengreen, Rose and Wyplosz (1995) identify episodes of speculative pressurein OECD countries from 1959-1993 as when a weighted index of quarterly macroeconomicindicators deviates significantly from their mean level and find that 21% of their episodes of speculative pressure result in failed attacks. Moreno (1995), using a similar methodology forEast Asian countries from 1980-1994, shows that speculative pressure leading to appreciationaccount for 40% of the episodes in his dataset. 7 The lack of hedging in recent financial crises has been documented by a number of authorsincluding Eichengreen, Hausmann and Panizza (2002) and Alesina and Wagner (2005). Bothpapers document that features of srcinal sin are positively correlated with managed floatingand fixing. There is also evidence that governments may signal by exposing their countries todevaluation risk. For instance, before the Mexican crisis, a substantial portion of the publicdebt was restructured from  cetes   (peso-denominated) to  tesobonos   (dollar-denominated). Byensuring that a devaluation would increase the public debt, the creation and rapid growthof this financial instrument signaled to foreign investors that the government was not goingto allow the currency to fall. Also, the Bangkok International Banking Facility, initiallyestablished to intermediate foreign investment, functioned as a conduit for short-term foreignlending. Thailand also famously offered tax breaks which encouraged offshore borrowing.The model explicitly analyzes the incentives for hedging and suggests why governmentsunder duress, as in these two cases, may not want to encourage it.The idea that the choice of exchange rate regime and its management arbitrates between(primarily) domestic interest groups has been frequently discussed in case studies and, morerecently, in cross-sectional empirical analysis. See Eichengreen (1995) and Frieden (1997)for examples of the former and Alesina and Wagner (2005), Eichengreen, Hausmann, andPanizza (2002), and Levy-Yeyati, Sturzenegger and Reggio (2003) for examples of the latter.Frieden, in his overview of the Mexican crisis in 1997, discusses how depreciation would haveaffected the purchasing power of swing voters in the urban middle and working classes.Numerous cross-sectional studies have documented that devaluations are less likely shortlybefore elections, but often occur right after an election when a new government can blamethe previous government for mismanagement. Cooper (1971), for instance, was the first toempirically document the political costs of devaluation. His study, which has been updatedand extended by Frankel (2005) for the period 1971-2003, showed that in the aftermath of devaluations, nearly 30 percent of governments collapsed within 12 months, as compared to14 percent in a control group. These facts suggest that the electorate revises their beliefsabout the government in the wake of a devaluation and that governments understand this,viewing exchange rate management as an important signal. Indeed, governments often viewthe exchange rate as a foundation of macroeconomic policy and their source of economic 7 Both authors, of course, recognize that they may miss episodes of unsuccessful attacks by using quarterly datawhen repelled attacks may only last days, and that detecting the magnitude of government intervention in foreignexchange markets is notoriously difficult. 4
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks