The British Pound in the ERM

Home , Black Wednesday, Deutsche Mark, Italian lira, Speculative attack

Mathias Zurlinden

Mathias Zurlinden, an economist with the Swiss National Bank, was a visiting scholar at the Federal Reserve Bank of St Louis when this paper was started. Thomas A. P01/mann provided research assistance.

The Vulnerability of Pegged Exchange Rates: The British Pound in the ERM

ETWEEN SEPTEMBER 1992 and August conditions of easy access to foreign exchange 1993, the European Monetary System (EMS) reserves and free capital mobility. went through the most serious crisis since its This article concentrates on the British epi- start in 1979. Member countries cross-pegging sode in the EMS crisis. Since the United King- their exchange rates in the framework of the dom’s participation in the ERM was suspended Exchange Rate Mechanism (ERM) were confront- in September 1992, only the early phase of the ed with a string of speculative currency attacks. Associated with these attacks were five realign- crisis is covered. First, I describe a brief history of the pound in the EMS. Next, I have selected ments and the suspension of two major macroeconomic indicators on the eve of the cri- currencies—the Italian lira and the British sis to provide a picture of the economic situa- pound—from the ERM. The situation eased off tion and the credibility of the exchange rate only when the fluctuation margins were widened band as perceived by the markets. Then, I dis- considerably in August 1993. cuss the main features of the speculative attack There are two reasons to review these events. on the pound against the background of the First, there had been no genuine realignment in basic model of balance-of.payments crisis. To the EMS for more than five years. The EMS had this end, I introduce the model originated by widely come to be seen as a model for a viable Krugman (1979), along with extensions motivated pegged exchange rate system. Second, most of by the British situation. the recent cases of speculative currency attacks occurred in developing countries, where access to foreign exchange reserves is rather limited BRITAIN’S PARTICIPATION 2 and capital controls usually play an important in the ERM role in maintaining pegged exchange rates? Hence, the near-collapse of the ERM provides When the EMS started operating on March 8, a useful example of a speculative attack under 1979, Britain did not participate in the central 1 See Edwards (1989) for a detailed analysis of devaluations in developing countries. 2 There is a vast literature on the EMS. Ungerer, et al. (1983, 1986, 1990) provide accessible reviews of EMS developments. Also see Fratianni and von Hagen (1992), Giavazzi and Giovannini (1989), and Gros and Thygesen (1992) for more advanced discussions.

~cpTrL1ncRJnrTnnrn iooq 42 piece of the new system, the ERM? In the view growth targets for monetary aggregates. Capital of British monetary authorities, the loss of room controls were removed rapidly and fiscal policy for maneuvering under a system of pegged was oriented toward balancing the budget.8 This exchange rates outweighed probable gains. Many strategy resulted in a large reduction of inflation observers did not give the ERM much credit (from 18 percent in 1980 to less than 5 percent either. Some predicted an inflationary bias, in 1983), albeit at the price of substantial output while others expected the system to he drawn losses. A complicating factor was the increasingly apart soon by the widely differing inflation rates unstable relationship between the targeted among participating countries. In the event, the monetary aggregate (sterling M3) and nominal ERM performed surprisingly well. Inflation rates income.’ This made sterling M3 a questionable decreased substantially (albeit not more than in indicator, which risked a reduction in the credi- non-ERM countries), and the variability of nomi- bility of monetary policy. In response, monetary nal and real effective exchange rates fell a great authorities tried several alternatives. First, sever- deal. al aggregates were targeted simultaneously. Certainly, many realignments were required Then, the emphasis shifted to narrow monetary for the ERM to survive during its early years. aggregates. Finally, in 1987-88, the free float was The 17 realignments witnessed in the period replaced by a managed exchange rate shadowing 1979-93 are summarized in Table 1.~‘fivo fea- the deutsche mark. tures stand out. First, the deutsche mark never was devalued against other ERM currencies. In retrospect, the mark exchange rate target- Second, realignments became less frequent until ing of 1987-88 was an ill-fated attempt at finding 1992, reflecting the decline in intra-ERM infla- a stable nominal anchor. Initially, monetary poli- tion rate differentials.~ cy loosened due to the determination of the Capital controls also played a role in the government to stick with the unofficial target survival of the ERM. They had intensified in the exchange rate of 3 marks per pound. As a result, the economy overheated and forced final years of the Bretton Woods System and monetary authorities to tighten the policy stance many European countries continued to use them and to let the pound appreciate.8 after that. It was not until 1988 that an EC directive stipulated the complete liberalization of Despite this troubled experience with an capital movements. For’ most member countries, exchange-rate oriented policy, ERM membership this was accomplished by mid-1990 (extensions remained an option favored by the chancellor of were granted for Greece, Ireland, Portugal and the exchequer, Nigel Lawson, and supported by Spahi). leading husinesspeople. Tn June 1989, at the EU Britain chose a different way. Rather than summit in Madrid, the government set the participate in the ERM when it began in 1979, terms for Britain’s entry to the ERM. These Britain decided to pursue a deliberately tight terms were: British inflation close to the EU monetary policy based on a free float and average; i-cal progress towards completion of

3 The EMS includes all members of the European Communi- ~Achronological account of British economic policy is ty (EC). The ERM originally included only Belgium- provided by the annual surveys on the United Kingdom Luxembourg, Denmark, Germany, France, Ireland, Italy and published by the Organization for Economic Cooperation the Netherlands. Portugal, Spain and the United Kingdom and Development (OECD). joined in April 1992, January 1990, and October 1990, ‘Sterling M3 is MS less residents’ deposits abroad. respectively. Italy and the United Kingdom suspended their 8 participation in the ERM in September 1992. Greece has See Belongia and Chrystal (1990) for a critical discussion never participated in the ERM. of this episode. ~Itmay be argued that there were only 16 genuine realignments. On January 8, 1990, when the Italian lira switched from ±6 percent to ±2.25 percent fluctuation margins, the central rate was devalued relative to the current market rate. The new lower intervention margins were not below the old margins, however, except for the Spanish peseta exchange rate of the lira. 5 Giavazzi and Giovannmni, among many others, argue that the EMS became a greater deutsche mark area by 1983; Germany is the center country, and countries such as Italy and France peg their currencies to the mark. See Fratianni and von Hagen for qualifications of this view.

FEDERAL RESERVE BANK OF St LOUIS 43

N ‘N

‘~ ~flbIs1 N “ ~MSneanw,meêts~~ercent~à~angáin Bilateral centS fiatest

\NDSt~~\ ~‘:N e~ \PJ~R ~rnK flA~~~P~‘~lRL’N~’NuTN:N~l*L> ~lsc~ Nu$L

\/\ \\/\\ <$sptember 4197t ~ \~N N~ *Q~ \N /$ \ \“NN NN is \\\\\N /N ‘N’N~ • \ -~ -SD M ‘k~$abttry ~!. \* 3ft, ~NN~ N

~4W*14l98?N N N & ‘N 278’N 42 N M*Is 983 6 fl N~t5 -4s -25 35 *92198$ tO 2,0 b 60 2fl NA )‘19$6 TN Nj $~ NN N tON N 2 NNJatw* 19W 2.0 // NN to

JUt>409 N N~ N N NN \NN N 9 $set*ettaaaN N / ~NN N 7Ø,N N/S~~St~ 19S NNNN ~ ‘-~W N NNN* N N’ N~~8 Nna~ ‘N \\N\ ~ N/// ,/,\ \N~, N ~‘G~9”t N’”’ N ,/,~//N//\/ / N 4* N~ N’N N NN /N~ ~ N / /N N NNNNN ~ N NN t NNNNN:4 N:. NN N / “‘N N ~NN / !t!~$~~~ ~ N N N ~. ~ t~ssf tpSTtSJMatlwkS~’ ~ N’ //<~)4~a~ue “at øeSsa. py4~ / ~ ~ “~ N’~, ~ kjs

N N ~“ N ~ ~N N’ / N

N C~/// N ~ N ;N ‘N/j 4” 44 N,’

the single European market; financial-market to keeping the exchange rate within these mar- liberalization; end of exchange controls; and gins. Essentially, two instruments were available strengthened competition policy in the EC.9 to this end: interest rate policies and direct in- When Britain actually joined the ERM 15 months terventions on the foreign exchange market?2 later, all conditions except inflation convergence were virtually met?” Consider a case where the pound approaches the lower margin of its deutsche mark band. Britain Enters the ERM The Bank of England can sell foreign currency or it may raise short-term interest rates to pre- Britain entered the ERM on October 8, 1990, vent the pound from depreciating further. lb with fluctuation margins of ±6 percent around finance the intervention, it may either draw on bilateral central rates, instead of the usual ± its own reserves or borrow from other sources 2.25 percent.” As with Italy and Spain before, and (international capital markets, central banks). In Portugal later, the ERM allowed wider margins the ERM, access to foreign exchange reserves is to provide the newcomer some flexibility to ad- facilitated by the Very Short-lerm Financing just. By joining the ERM, Britain committed itself Facility (VSTF). Under the VSTF, the Bank of

9 See OECD(1989/1990), p. 40. “The effective margins were approximately —2.22 percent ‘°Theremay have been short-term reasons to speed up and +2.28 percent for the narrow (4.5 percent) bands, and —5.82 percent and +6.18 percent for the wide (12 percent) Britain’s entry in the second half of 1990. With the inter- governmental conference on monetary union scheduled for bands. Margins of about ±2.25percent and ±6percent would violate the condition that central bank A’s compulso- December 1990, Britain knew that it had to be part of the ry buying rate for currency B must be the same as central ERM if it wished to influence the path of monetary integration in the EC. Furthermore, a general election was bank B’s compulsory selling rate for currency A. approaching, and the government could expect ERM mem- “In Britain, changes in the policy stance are typically sig- bership to provide some honeymoon” effect by allowing naled by changes in short-term interest rates. See Batten pound interest rates to fall without jeopardizing the et al. (1990) for a discussion of the operating procedures. exchange rate.

SEPTEMBER/OCTOBER 1993 44

England is allowed to borrow marks from the notable first in Italy, where the discount rate Bundesbank virtually without limits. The was raised over the summer of 1992 in several Bundesbank is obliged to grant such credits 3 steps, from 12 percent to 15 percent. The pound upon request? felt some pressure too, and British monetary After Britain joined the ERM, the pound authorities began to step up interventions on moved most of the time comfortably in the ±6 the foreign exchange market in late August. percent band around its central rate. After a On September 3, 1992, Britain announced a temporary appreciation, the pound stayed for program to borrow ECU 10 billion, about $14.3 more than one year in an implicit narrower billion (U.S.) at the time, in the international band in the neighborhood of ±2.25 percent market to increase foreign exchange reserves. around central parity. Pressure on the pound Ultimately, the rise in domestic interest rates usually was short-lived and quickly reversed by did not save the Italian lira. On September 13, either a slight increase in domestic interest rates the lira’s central rate was devalued by 7 percent. or modest interventions in the foreign exchange TWo days later, Germany slightly eased monetary market. Interest rates were raised temporarily policy. The discount rate was lowered by 50 when Margaret Thatcher stepped down as basis points to 8.25 percent and the Lombard prime minister in November 1990 and in the rate was lowered by 25 basis points to 9.5 per- weeks before the general election of April 1992, cent. These adjustments—the first German in- when opinion polls pointed to a victory for the terest rate cuts in nearly five years—were opposition Labour party. On the whole, however, perceived as unexpectedly small by the markets, interest rates (as well as inflation) decreased and comments attributed to the Bundesbank substantially during the period Britain partici- president, who appeared to question the ade- pated in the ERM. quacy of the pound’s central rate (subsequently denied), raised tensions further?~ Tensions in the ERM The tensions in the foreign exchange markets Britain Withdraws from the ERM that finally led to the near-collapse of the ERM in September 1992 were triggered by doubts On September 16—Black Wednesday—the Bank about the progress toward monetary union?~ of England intervened massively on the foreign In June 1992, the Danes had voted no in a exchange market to prevent the pound from referendum on the Maastricht Tt1’eaty, which falling below the lower margin of its Deutsche included a chapter on European Monetary mark band. Furthermore, it raised the base Union (EMU). Moreover, the outcome of the lending rate from 10 to 12 percent and an- French referendum on the treaty in September nounced later in the day a further rise to 15 was expected to be close. Since the prospect of percent, to be effective the following morning. monetary union had provided an anchor for These measures did not succeed in relieving the expectations, the outlook for the current parities pressure on the pound. In the evening, British in the ERM looked rather bleak if France reject- monetary authorities announced the temporary ed the treaty also. For reasons discussed below, suspension of the pound from the ERM. It pressure on the exchange rate became most seemed hardly feasible to fix a new parity less

3 4 ‘ Even if access to the VSTF is said to be unlimited, this is ‘ For a detailed account of the events prior to Britain’s with- not literally true. The Bundesbank has two reasons to drawal from the ERM, see Bank of England Quarterly Bulle- make sure that the VSTF is not overburdened. First, it tin (November 1992). bears an exchange rate risk since the credits are denomi- 5 ‘ See Financial Times, September 15, 1992, and September nated in European Currency Unit (ECU), a basket currency 16, 1992, respectively. As is now known, the Bundesbank defined by fixed quantities of member currencies. So, if the actually did suggest a trade—with the size of German in- pound depreciates relative to the mark, the Bundesbank terest cuts depending on the size of the realignment—to will be repaid the value of depreciated ECU. Second, the the chairman of the EC Monetary Committee. Since the selling of the mark during interventions raises Germany’s Italian lira finally was the only currency devalued on Sep- monetary base and would jeopardize its inflation objective tember 13, German interest rate cuts consequently were unless the Bundesbank is able to sterilize the intervention. Therefore, the Bundesbank insisted right from the start of small; see Financial Times, December 11, 1992, for a detailed account. the ERM on an opt-out clause. According to Central Bank- ing (1992; Robert Pringle, ed), this clause was confirmed in a letter to the German government by the then-president of the bank, Otmar Emminger.

FEDERAL RESERVE BANK OF ST. LOUIS 45 than a week before the French referendum on ties. This, of course, turns out to be difficult. the Maastricht ‘Theaty The next day, the base The market knows that pegging the exchange lending rate was moved back to 10 percent. At rate ultimately is a conditional commitment. Ex- the same time, Italy announced the lira’s tem- pectations are formed that monetary authorities porary withdrawal from the ERM, while Spain will realign if the perceived cost of defending devalued its currency by 5 percent and imposed the exchange rate is larger than the perceived temporary capital controls. cost of a realignment. This section takes a look at selected macroeconomic indicators and asks The floating of the pound marks the end of whether the exchange rate target bands were Britain’s ERM episode. In the following weeks, perceived as credible by the markets. when the markets did not calm down after the French had narrowly approved the Maastricht General Economic Conditions ‘meaty, it became clear thai Britain and Italy would not rejoin quickly ‘The ERM struggled on Table 2 shows macroeconomic indicators on for another 10 months. Then, on August 2, the eve of the crisis for all countries participat- 1993, EC member states decided to raise the ing in the ERM (except Portugal, because of data margins of the exchange rate bands to ±15per- limitations). The countries are ordered in terms cent around the central parities, an action com- of their relative size (measured by GDP). Indica- ing close to a suspension of the system. The old tors refer to August 1992 for monthly data and margins continued to be valid only for the the second quarter of 1992 for quarterly data. deutsche mark/Dutch guilder exchange rate. Annual figures are OECD forecasts for 1992, published in June of the same year. The indica- In retrospect, the distinctive feature of tors can roughly be divided into three groups: Britain’s defense of the pound was the almost monetary indicators (money supply growth, complete lack of interest rate policies. Domestic short-term interest rate, long-term interest rate); interest rates were raised only on the final day fiscal indicators (primary deficit/GDP ratio, of the crisis and the size of this rise was rather debt/GDP ratio); and indicators describing final small. Most other weak-currency countries in goals (output growth, inflation)?~ the 1992-93 crisis of the EMS raised their short- term interest rates more aggressively?6 As indi- These indicators shed some light on the diver- cated above, however, this did not provide them gent economic forces in the ERM. The most im- durable relief from speculative attacks. portant single event hitting the EC during recent years has been German unification. Basic economic theory suggests that the German unifica- ECONOMIC INDICATORS PRIOR tion would lead to an increase in that country’s TO BRITAIN’S WITHDRAWAL aggregate demand and a real appreciation of the FROM THE ERM mark (where tile real appreciation is equal to the nominal appreciation adjusted by the infla- When Britain participated in the ERM, British tion rate differential against foreign countries). monetary authorities regularly emphasized their These effects were heightened by the decision commitment to the established exchange rate to finance higher public spending by borrowing 8 parities. It is not enough to make such a pledge, rather than by raising taxes? Under pegged ex- however. The market has to be convinced that change rates, the real appreciation of the mark monetary authorities have no incentive to drop is brought about by a positive inflation differen- their commitment and that they will take tial between Germany and the other ERM whatever action is necessary to defend the pan- member countries. Since the Bundesbank was

‘°SeeGoldstein et al. (1993), Annex VI and the statistical of GDP; and (4) the public debt/GDP ratio must not exceed appendix in particular. 60 percent. Note, however, that table 2 uses the primary “Four out of the seven indicators closely correspond to the deficit (i.e. the budget deficit net of interest payments) to macroeconomic convergence criteria for EMU mentioned in indicate whether the evolution of the fiscal indicators goes the Maastricht Treaty. These criteria are: (1) the inflation in the direction required by the EMU criteria. 8 rate of the country under review must not exceed the aver- ‘ These are short-run effects. In the long run, an expansion- age of the three EC countries with the lowest inflation rate ary fiscal policy that permanently raises the debt/GDP ratio by more than 1.5 percent; (2) the interest rate on long-term will cause the real exchange rate to fall, since the country government securities must not be more than 2 percentage must export more to offset the effects of the decline in its points higher than the average rate of the same three net external asset position. countries; (3) the budget deficit must not exceed 3 percent

etnttliAatD,nrTnnco Inn.3 46

‘N, NNNNN’ N N NNN N N” ,NN NNNN,, NNNNNN ,N N NN N \N N N

NNN NN NNNNN N’ t~’~ N’ NNN N N N N tablea N NNN N

Macroeconomic indicators on the, E~eof’ tt!e Septembé 1992 Crr~s N N (~npe~pent) NNNNNNNNN\ :,~ N\~ N N ,NN N\ :.N NN\

NNNNNNNNNNN

N N NNNNN N NN N ::NN N N N / gOVetflfl NN NGS N NN NNNNNNNNNNSh tetmN N~4fl94~tflt N::. Pr!mMt N gosern~nt Ctnsitrflet NNN N NNNN N ~ NMOfl hIteSt NN N sn~test N N static.! ‘N debt/ N Npace NOOP N CaUñtnes. shares N grGwtht ~NN N N GDfl N itlsonc ~towtM N

Gefatafty. 2S1TNNNI.t N~ NNNNs N12r 4$,57N NSSN it 7 4 Ft~fteeNNNNNN195 ~16NNt0*NN~N93~ NNfl N -485 N20 N~~22N:. NN NIBSN N Q9 Nfl4 N j5NNN NN’~flN N NNtD7BN 50 0 StedNKirtg4offi NN4~ NNNfl NNNN S N Nfl N OAN N NSNN SS ~i~: NtG NN NN N NNN~4RNN NN 4$,Q NNN~ 58 N *5 :NN K~~NNN N47 $ØN 9~tNNNNN NN,~ NNNNNNN~ NN7fl S,S NNEeIjJfaN\ NNNa A NN*S ,UNNN4S NNNN$ZPN NN N[TtNNN N95NN t0~ t4~ Nfl, N 1 816 22 ______Nay -se iS N4 N 7

SQ tIN N NN 0’S SIS U N4 ~

‘‘N’ ,, , NNN, N N NNNNNNN,,NN ,,NN NN N, N N N

iast ‘Intetnanal N11a’ncla( Stattattes VP ‘WeStilt NN N N N 4s’Mfl scentaoe eM flflna $ S çser~ N aAugust tQ9~$5 hne6Ob N NN 4 Aug~ist1992 $5 tWa St (exc.~tIP& ltne Sm ot3W~N N, NNN Pmfaehqns, S 1St DECO Eceh*$ic ONOc$jMteN199~ N, !Mgust’Nta,* pSreS*9e 119* tv ba’ssØdØ year SWa JEt ‘West N N N N N’> > ~~N,N N NNNNN

%&vel~WS9btS4~by GOP shates NNN N N> N NNNN>,N:NNNNN> N,NNN NN NN~:. N:: ~::~NN:> N NN N~ N NNNNNNN N N \NN NN NN4NNN NN. N

detertnined to keep inflation low and to let to reduce its large debt/GDP ratio in view of EMU. short-term interest rates rise, the required real Britain was in an even better position with appreciation of the mark could come only by a regard to inflation. Italy and Spain clearly had substantial reduction in inflation in the other an inflation problem reflected in both inflation countries. rates and long-term interest rates. Despite some success in bringing inflation rates down from So, Germany’s economic policy on the eve of double-digit levels, there was still a substantial Britain’s crisis was characterized by a slightly gap compared with the other ERM countries. loose fiscal policy (reflected in the primary deficit) and a tight monetary policy (at least as Britain, on the other hand, had inflation and reflected in short-term interest rates of nearly long-term interest rates below the average of the 10 percent, substantially higher than the infla- ERM countries. Britain’s main problem was rela- tion rates). Despite slow output growth in tively slow growth. The (JECD’s forecasts for Germany, the chances of German interest rate GDP growth in 1992 had Britain at the bottom cuts seemed bleak. The growth of the German of the EC countries. Starting in 1990, the British money supply still was above the target range recession had been particularly stubborn, and and the inflation rate—running at 3.5 percent hopes for an economic recovery had been disap- and accelerating—was considered too high. pointed repeatedly. As a result, it was difficult not only to accept the current interest rate level What were the specific problems of Britain, imposed by Germany, hut also to convince the Italy and Spain, the three large EC countries market that domestic interest rates would be forced to adjust their currencies in September raised even further should the pound come 1,992? Table 2 indicates that a deficit in the under pressure. primary balance was expected for Britain and Italy. This was much more trouhlesotne for Italy, No attempt is made in table 2 to weigh the since Britain’s debt/GOP ratio was low, while various indicators to calculate an overall Italy actually had to realize substantial surpluses weighted-average indicator for each country.

rrflr~ftI QrQrn’tr QAMW flr QT flfl~C 47

A review of EMS realignments since 1979, how- 1987 for both the French franc and the Italian evet~indicates that the chief cause for a devalu- lira) or the entry to the ERM (January 1990 for ation was probably a persistent inflation differen- Spain).20 The figures exhibit substantial real ap- tial, leading to an overvaluation of the currency. preciation for both the Italian lira and (to a less- Such an overvaluation often was built up over er extent) the Spanish peseta. The franc, while an extended period of tine. Therefore, a better appreciating in real terms during the first few indicator than an annual performance measure years, largely retraced its rise by 1990 thanks to is the cumulated rate of change of the real ex- low inflation. change rate over the period starting with the date when the current parities were established. Overall, the evolution of the real exchange rate does not point to the pound (or the franc) Movements in the Real as a candidate for a devaluation.21 There was Exchange Rates nothing like the usual pattern of an increasing overvaluation due to a persistent inflation rate Figure 1 shows the real mark/pound exchange differential. Still, theme was the discrepancy be- rate when Britain participated in the ERM. tween the cyclical needs of the British economy Nominal exchange rates, the margins of the and the high interest rates imposed on the ERM exchange rate band, and cost or price differen- by Germany. The next section examines the per- tials are given for convenience. There are vari- ception of the exchange rate band’s credibility ous ways to calculate real exchange rates. Here, before the speculative currency attacks actually two indexes were calculated. The first is based forced the withdrawal of the pound (and the on unit labor cost, and the second on consumer lira) from the ERM. prices?9 Note that unit labor cost data are quarterly (ending in the second quarter of 1992), Perceptions of Britain’s Credibility while consumer price data are monthly (ending in the ERM in August 1992). Hence, the top graph of figure 1 (quarterly data) does not have the pound fall- A simple way to assess an exchange rate ing to the lower margin at the very end of the band’s credibility is based on uncovered interest period. Both indexes show that cumulated infla- rate parity.zz Uncovered interest rate parity tion differentials were very small on the eve of states that under perfect international capital the collapse (0.05 percent for consumer prices mobility and risk-neutral speculation, the and 1.4 percent for unit labor cost). Since the differential between nominal domestic and for- nominal exchange rate of the pound depreciated eign interest rates is equal to the anticipated inside the band, both measures of the real rate of depreciation of the domestic currency. exchange rate indicate that the mark price of So, given the current exchange rate and domes- the pound was lower in real terms compared tic and foreign interest rates for various maturi- to its entry level. ties, the expected exchange rate for these Figures 2-4 repeat this exercise for France, maturities can be calculated. Then, the band is said to be credible if the expected exchange rate Italy and Spain, the three other, major, non- German countries participating in the ERM. is within the margins of the band.23 Only the real exchange rates based on consumer Figure 5 shows the results for the mark/pound prices are shown. The starting dates differ ac- exchange rate when Britain participated in the cording to the preceding realignment (January ERM. The time horizons are three months, 12 21 ~Consumerprices excluding mortgage rates are used for There may have been real factors leading to a lower real Britain in this calculation, but not in table 2. The former mark/pound equilibrium rate, however. Also, the central par- measure is widely seen as a better indicator of core infla- ity may have been too ambitious from the beginning. Note tion. If a tight monetary policy raises interest rates, includ- that the real Deutsche mark/pound exchange rate was low- ing mortgage rates, the overall price index may actually er in October 1990 than during most of the 1980s. show an acceleration of inflation, while the brakes on infla- 22This test was originated by Svensson (1991). tion are already in place. Britain’s inflation would be lower (and the real depreciation of the pound even more 2Slf the assumption of risk-neutrality is dropped, a term for pronounced) if the overall consumer price index were risk premia must be included in the interest rate parity taken, since interest rates were falling during much of the condition. Svensson (1992) showed that these premia are period under review. very small for narrow exchange rate bands. 20 Again, the realignment of January 8, 1990, caused by the switch of the Italian lira to a narrow band is ignored. See footnote 4 for explanations.

SEPTEMBER/OCTOBER 1993 48

Figure 1 Nominal and Real DM/Pound Exchange Rates

Index (1990.3=100) 115

110’ Upper Margin for Nominal Exchange Rate 105 Cumulative Unit Labor Cost Differential 100-

95 Lower Margin for Nominal Exchange Rate 90

80 Ill IV IIIIII IV I II 1990 1991 1992

Nominal and Real DM/Pound Exchange Rates

Index (1990.10=100) Monthly Data 115

110— Upper Margin for Nominal Exchange Rate

105- Cumulative Consumer Real Price Differential 100- ~nal 95- Lower Margin for Nominal Exchange Rate 90-

85-

80— I I II I ON D J FM AM J JASON D J F MA M JJA 19901991 1992

~n~DAI Dretouc DAMIt flC I flflIO 49

Figure 2 Nominal and Real DM/French Franc Exchange Rates Index (1987.01=100) Monthly Data (January 1987 to August 1992) 115-

110-

Upper Margin for Nominal Real 105-

100-

Lower Margin for Nomtnal Cumulative Consumer 95 - Exchange Rate Price Differential

90-

85-

80- II I 1987 88 89 90 91 1992

Figure 3 Nominal and Real DM/ltalian Lira Exchange Rates Index (1987.01=100) Monthly Data (January 1987 to August 1992) 115—

110— Upper Margin for Nominal Exchange Rater 105-

100- I

95 ~r~nfoTomina~~ 90-

85- CumulatIve Consumer Price Differential 80 I 91 1992 1987 88 I 89 90 50

Figure 4 Nominal and Real DM/Spanish Peseta Exchange Rates Index (1 990.01 =1 00) Monthly Data (January 1987 to August 1992) 115

110 Upper Margin for Nominal Exchange Rate / 105

100 Cumuiati.ve Consumer ‘~¾. Puce Dtff*ranvat

95 - Lower Margin for Nominal 90- Exchange Rate

85-

80 1990 91 1992

Figure 5 DM1 Pound Exchange Rate (as Deviation from Central Rate) Percent October 8, 1990, to September 16,1992

1990 91 1992

CflCO A OtOCO~IC C AMU flt 0T fl* HO 51 months and five years, hased on interest rate within the band (conditional upon no realign- differentials. Interest rates for three and 12 ment) is estimated under the assumption of months are Euromarket rates, guaranteeing that rational expectations using realized exchange the deposits are comparable in every respect ex- rate data. The result is deducted from the in- cept currency denomination. The interest rates terest rate differential to get the expected rate for five years are based on government bond of realignment.26 Rose (1993) recently used this yields, since no Euromarket rates are readily method to calculate expected realignment rates available for maturities of more than 12 months. for the mark!pound exchange rate. He shows The data are daily. that expectations of a pound realignment were low throughout most of 1992. The pound’s The results indicate that the expected exchange credibility was not in reasonable doubt until rate always moved within the margins of the mid-August 1992, at the earliest. band for maturities up to 12 months until mid-August 1992. For maturities of five years, Overall, the credibility measures indicate thai however, the expected exchange rate most of the the successful speculative attack on the pound time was outside the margins. The latter result was not anticipated well in advance. The diver- implies that there were some lingering doubts gent economic forces stemming from German about the long-term credibility of the mark! unification—well known for quite some time— pound exchange rate band. Nevertheless, a crisis did not result in a prompt decline in the ex- was not perceived as particularly likely. Doubts change rate band’s credibility Signs of an ap- about the short- or medium-term credibility of proaching crisis were strikingly rare. To gain a the band arose only immediately before the par- broader picture of the timing and the dynamics ity actually was attacked. of the speculative attack, the next section con- siders the basic model of a balance-of-payments Using monthly data for 12 months, the same crisis. exercise was applied to the franc, the lira and the peseta during the periods these currencies were participating in the ERM. The results, THE ECONOMIC ANALYSIS OF A summarized in figures 6-8, show that the SPECULATIVE ATTACK credibility of the band derived for the pound cannot be generalized to apply to other EMS To clarify Britain’s problem in the ERM, an countries throughout the period of their par- outline of the basic model of speculative curren- ticipation. The expected future exchange rates cy attacks that has dominated the recent litera- of both the franc and the lira mostly were ture follows. Also reviewed are some extensions outside the margins of the bands during 1979-89. motivated by the preceding discussion of the Since 1989, the expected future exchange rates British episode. These extensions include the of ERM currencies usually were within the roles of interest rate pegging, borrowing and margins.2~ capital controls.27 Rose and Svensson (1991) proposed a slightly Some Theoretical Considerations more elaborate technique to assess the credibility of exchange rate bands.25 They split the total Consider a situation of a currency whose rate of depreciation implied by the interest rate value is pegged to a foreign currency. The differential into the expected rate of deprecia- foreign interest rate is assumed to be constant. tion within the band and the expected rate of There is no commercial banking sector, and the realignment. The expected rate of depreciation supply of money is equal to the total of both

4 27 ‘ See Frankel and Phillips (1991) for an investigation of credi- The literature on speculative currency attacks starts with bility in the EMS using survey data on expected exchange Krugman (1979) and Flood and Garber (1984b). Flood and rates in addition to interest rate differentials. Garber (1984a) and Obstfeld (1986) analyzed self-fulling at- 25 3ee Svensson (1993) for an application on EMS data. tacks and multiple equilibria. Agënor et al. (1992) and Blackburn and Sola (1993) provide selective reviews of the 26lt is assumed that the position of the exchange rate inside fast-growing literature. For an excellent non-technical the band is the same before and after the realignment. description, see Goldstein et al. (1993), Annex V. While this makes it possible to interpret the result as the expected rate of realignment, it is somewhat at odds with the facts. After most realignments, the exchange rate lumped toward the upper margin of the band. 52

Figure 6 DM1 French Franc Exchange Rate, 12 Months Expectations (as Deviation from Central Rate) Percent 10

-5

-10

-15

-20

Figure 7 DM1 Italian Lira Exchange Rate, 12 Months Expectations (as Deviation from Central Rate) Percent 53

Figure 8 DM1 Spanish Peseta Exchange Rate, 12 Months Expectations (as Deviation from Central Rate) Percent 10- Upper Margin 5—

-5- Lower Margin

-10-

•-15

-20 IIIIIIIIIIIII 19798081828384858687888990911992

domestic credit and foreign assets held on the rate falls discretely. Consequently, forward- central bank’s balance sheet. Now, imagine that looking speculators, foreseeing the imminent col- all domestic assets are increasing at a steady lapse, would try to sell domestic currency and pace, in part, to finance a chronic government buy foreign currency before the central bank budget deficit. Since the exchange rate (defined runs out of reserves. A speculative attack on the as the foreign currency price of one unit of foreign exchange reserves of the central bank home currency) is fixed, a steady outflow of takes place, and a collapse occurs. With perfect foreign exchange reserves occurs if the growth foresight, no discrete drop in the exchange rate of domestic credit exceeds the growth of money occurs at the time of the collapse. demand. In this case, foreign exchange reserves eventually are exhausted and the central bank For assessing the timing of the attack, it is useful to look at the so-called shadow exchange has to withdraw from foreign exchange inter- 28 vention. Once exchange rates at-c flexible, the rate. ‘i’he shadow exchange rate is the exchange rate that would result if the speculative price level will rise and the exchange rate will fall at the same pace as domestic credit growth attack that exhausts all foreign exchange reserve in excess of money demand. holdings of the central bank would take place today. The speculative attack takes place when There is a problem, however. At the moment the shadow exchange rate is equal to or below of the switch to a flexible exchange rate, inves- the current exchange rate. As long as the tors with assets denominated in domestic cur- shadow exchange rate is higher than the current rency suffer a capital loss since the exchange (pegged) exchange rate, a speculative attack

28 5ee Flood and Garber (1984b). 29This is known in the literature as the peso problem. 54 would bring losses for the speculator. On the of the central bank. In contrast to the two cases other side, a speculative attack is profitable if discussed above, the policy of the central bank the shadow exchange rate is lower than the cur- in this case is not exogenous to the speculative rent (pegged) exchange rate. Competition among currency attack. As a result, there may he sever- speculators reduces expected losses and profits al equilibria, depending on the expectations of down to zero. This story provides an explana- the market. If there is no attack, the exchange tion of why a speculative attack may take place rate is perfectly viable forever. If there is an even when the central bank still has sizable attack, the central bank gives in and accepts a reserves. The ultimate reason for the collapse is depreciation of its currency. the inconsistency of the pegged exchange rate with the growth of domestic credit. But the This model of a self-fulfilling speculative at- point to remember is the pivotal role of expecta- tack is quite attractive in the context of the Brit- tions in bringing the collapse forward. ish case. Remember that the main reason for joining the ERM was the credibility that the Perfect foresight, however, is not a realistic ERM seemed to provide for monetary policy assumption. What happens when this assump- Later, realignments were rejected repeatedly tion is dropped? Imagine that there is uncer- because of the suspected risks to the credibility tainty either about domestic credit growth or of the system. So, when the speculative attack about the minimum size of foreign exchange on the pound revealed that the established reserves that would cause the central bank to exchange rate parities actually had lost their adopt flexible exchange rates. Again, the credibility, the situation had to be reevaluated. speculative attack takes place when the shadow The costs of defending the established parities exchange rate is equal to or below the actual had increased and it was quite possible that rate. Now, however, uncertainty implies that monetary authorities would regard these costs there is always a probability greater than zero as being too high. for the policy switch to take place in the next period. Moreovet this probability increases over The British Experience time, since the decline in reserves makes it more likely that the next realization of credit The main characteristic of the defense of the growth forces the switch to a free float. As a pound in September 1992 was the reluctance of result, there is a positive (and, over time, British monetary authorities to raise interest increasing) interest rate differential, implying an rates. What are the implications of this policy expected depreciation of the home currency for speculative currency attacks?~°Consider the exchange rate.29 In addition, there may be a general case with a pegged exchange rate and discrete drop in the exchange rate at the selling pressure against the home currency. So moment of the policy switch. These regular long as interest rates are free to move, competi- features of a balance-of-payments crisis are not tion among speculators for expected profits captured in the perfect-foresight case. from an attack on the home currency drives up A chronic fiscal defIcit financed by the central domestic interest rates. If the central bank pre- bank is the driving force of the collapse described vents these interest rates from rising, that brake above. This assumption does not meet the on the demand for foreign exchange disappears. experience of Britain in 1992. The Bank of As a result, the central hank will have to absorb England did not have to finance a budget the larger demand for foreign exchange. In the deficit, domestic credit growth was low, and end, the central bank, by pegging the interest there was no continuous outflow of reserves. rate, offers the market a favorable opportunity So, an alternative conceptual analysis would con- for a run on Britain’s reserves. Investors may sider a situation in which a country initially speculate at very low costs, and the speculative adopts fiscal and monetary policies in line with currency attack thus easily may involve huge the maintenance of the pegged exchange rate. amounts of funds. Nevertheless, suppose the market expects the central bank to switch to a free float and that A second characteristic of the British episode it would ease monetary policy if a speculative was the government’s borrowing on the interna- attack exhausted the foreign exchange reserves tional capital market to increase foreign exchange 30 Goldstein et al. (1993) first examined this extension of the basic model.

FEDERAL RESERVE BANK OF St LOUIS 55

change reserves. As emphasized above, a reserves, in turn, raises the shadow exchange speculative attack occurs when the shadow rate in much the same way as in the example in exchange rate is equal to or below the current which reserves are increased by borrowing. A pegged exchange rate. The size of the fot-eign tax on foreign interest earnings, by pushing up exchange reserves, in turn, has an effect on the the shadow exchange rate, may postpone a shadow exchange rate. Since a run on foreign speculative attack. To put it another way, the exchange reserves jeduces the domestic money removal of capital controls in the EC removed a 3 supply (assuming the effect is not sterilized), the shelter for weak ERM currencies. ’ money supply reduction associated with running Over’all, the distaste of British monetary out of a larger stock of foreign reserves is also larger. The potentially larger reduction in the authorities for allowing short-teim interest rates to rise made the defense of the pound in Sep- domestic money supply, in turn, raises the tember 1992 more difficult. The effects on shadow exchange rate. speculation were particularly grave since there Another device to postpone speculative cur- were no restrictions on capital movements. rency attacks is capital controls. While Britain Borrowing foreign exchange reserves by the did not use them at all in 1992, the ERM crisis authorities—short of unliniited borrowing— in September 1992 was the first, since virtually was no serious substitute under these circum- all capital controls in the EC had been removed. stances. Thus, it may be useful to take a look at how capital controls work in a speculative currency attack. The simplest way to model capital con- CONCLUSIONS troLs is to treat them like a tax on foreign The near collapse of the ERM in 1992-93 interest earnings. lb simplify the account, let reflected the vulnerability of pegged exchange the starting point he a situation in which the rate systems. This feature of pegged exchange fundamentals aie in order: i.e., no chronic fiscal rates is hardly new. Standard theory in interna- deficit or other influence is causing a gradual tional macroeconomics teaches that monetary depletion of the central bank’s foreign exchange autonomy and pegged exchange m-ates are incom- reserves. The pre-tax foreign interest rate con- patible in the absence of capital controls. Diver- tinues to he assumed as constant throughout gent econonuc forces or simple political events the analysis. may trigger a speculative attack, leading to the collapse of the pegged exchange rate. Since a speculative attack takes place when the shadow exchange rate is equal to or below For policymakers, the possibility that specula- the current pegged exchange i-ate, we direct our live currency attacks may be self-fulfilling is attention again to the effects of the instrument particularly troublesome. In Europe, proponents on the shadow exchange rate. A tax on foreign of pegged exchange rates have argued for years interest earnings reduces the net return on that exchange rate pegging to the mark provides foreign assets. As a result, an excess supply of a way to import the reputation of the Bundes- foreign currency emerges at the current ex- hank and get a credible anchor for monetary change rate. Since the exchange rate is pegged, policy. For obvious reasons, this approach had a the excess supply must he absorbed by the cen- special appeal to countries lacking a credible tral bank. This leads to highet- foreign exchange monetary policy. Yet the argument is less con- reserves, a higher money supply, and a reduc- vincing if speculative attacks are self-fulfilling tion in the domestic interest rate (reflecting the and the credibility of a country’s exchange rate reduced level of the net return rate on foreign commitment can vanish as quickly and unex- assets). The increase in foreign exchange pectedly as it did in September 1992.

310f course, capital controls are not as efficient as true taxes on foreign exchange. More importantly, capital controls typically are not introduced when there is no pressure on foreign exchange reserves. Nevertheless, the analysis here clarifies that capital controls can reduce the prevailing excess demand for foreign currency, boosting foreign exchange reserves and the shadow exchange rate. This analysis ignores the reduction in incentives to invest in a country that imposes capital controls as well, and this influence coutd also bode ill for any actual boost in the shadow price. 56

A necessary condition for such an attack to Fratianni, Michele, and JUrgen von Hagen. The European occur is that the markets expect the central Monetary System and European Monetary Union (Westview Press, 1992). bank to shift policy as a result of the attack. If Giavazzi, Francesco, and Alberto Giovannini. Limiting the markets have reasons to believe that a coun- Exchange Rate Flexibility: The European Monetary System try will relax monetary policy once a specula- (MIT Press, 1989). tive attack has exhausted the central bank’s Goldstein, Morris, David Folkerts-Landau, Peter Garber, reserves, an attack is more likely. In the case of Liliana Rojas-Suárez and Michael Spencer. “International Capital Markets: Part I. Exchange Rate Management and Britain, a persistent recession prepared the way International Capital Flows?’ World Economic and Financial for such beliefs. Uncertainties about the pros- Surveys, International Monetary Fund (April 1993). pects for EMU and the reluctance of British Gros, Daniel, and Niels Thygesen. European Monetary authorities to allow short-term interest rates to Integration (Longman, 1992). rise in defense of the pound subsequently ac- Krugman, Paul. “A Model of Balance-of-Payments Crisis?’ celerated the attack and reinforced a realign- Journal of Money, Credit, and Banking (August 1979), ment of the pound. In short, the United Kingdom pp. 311-25. could not convince the markets of its commit- Obstfeld, Maurice. “Rational and Self-Fulfilling Balance-of- Payments Crises?’ American Economic Review (March ment to a fixed exchange rate. This credibility is 1986), pp. 72—81. an essential factor in maintaining an effective Organization for Economic Co-operation and Development. exchange regime. OECD Economic Surveys: United Kingdom (various issues). Pringle, Robert, ed. “The Currency Crisis: A Seven Part Special Feature?’ Central Banking (Vol. 3, No. 2, autumn REFERENCES 1992), pp. 12—21. Agénor, Pierre-Richard, Jagdeep S. Bhandari and Robert Rose, Andrew K. “European Exchange Rate Credibility P. Flood. “Speculative Attacks and Models of Balance of Before the Fall: The Case of Sterling?’ Federal Reserve Payments Crises?’ IMF Staff Papers (June 1992), Bank of San Francisco Weekly Letter (May 7, 1993). pp. 357—94. ______and Lars ED. Svensson. “Expected and Predicted Bank of England Quarterly Bulletin (various issues, 1992). Realignments: The FFIDM Exchange Rate During the Batten, Dallas S., Michael P Blackwell, ln-Su Kim, Simon E. EMS?’ Working Paper Na 3685, National Bureau of Nocera and Yuzuru Ozeki. “The Conduct of Monetary Economic Research (April 1991). Policy in the Malor Industrial Countries: Instruments and Svensson, Lars E. 0. “Assessing Target Zone Credibility: Operating Procedures?’ Occasional Paper Na 70, Mean Reversion and Devaluation Expectations in the ERM, International Monetary Fund (July 1990). 1979—1992?’ European Economic Review (May 1993), Belongia, Michael T., and K. Alec Chrystal. “The Pitfalls of pp. 763—94. Exchange Rate Targeting: A Case Study from the United ______- “The Foreign Exchange Risk Premium in a Target Kingdom,” this Review (September/October 1990), Zone with Devaluation Risk,” Journal of International pp. 15—24. Economics (August 1992), pp. 21—40. Blackburn, Keith, and Martin Sola. “Speculative Currency ______- “The Simplest Test of Target Zone Credibility?’ IMF Attacks and Balance of Payments Crises?’ Journal of Staff Papers (September 1991), pp. 655—65. Economic Surveys (June 1993), pp. 119—44. Ungerer, Horst, Owen Evans and Peter Nyberg. “The Edwards, Sebastian. Real Exchange Rates, Devaluation, and European Monetary System: The Experience, 1979-82,” Adjustment (MIT Press, 1989). Occasional Paper Na 19, International Monetary Fund Flood, Robert P., and Peter M. Garber. “Gold Monetization (May 1983). and Gold Discipline?’ Journal of Political Economy (February ______Owen Evans, Thomas Mayer, and Philip Young. 1984a), pp. 90—107. “The European Monetary System: Recent Developments?’ ______- “Collapsing Exchange-Rate Regimes: Some Linear Occasional Paper Na 48, International Monetary Fund Examples?’ Journal of International Economics (August (December 1986). 1984b), pp. 1—13. ______Jouko J. Hauvonen, Augusto Lopez-Claros and Frankel, Jeffrey, and Steven Phillips. “The European Thomas Mayer. “The European Monetary System: Monetary System: Credible at Last?” Working Paper Na Developments and Perspectives?’ Occasional Paper Na 73, 3819, National Bureau of Economic Research (August 1991). International Monetary Fund (November 1990).