Back Matter

Back Matter

International Monetary Fund
Published Date:
January 1993
    • ShareShare
    Show Summary Details


    January 1988International Capital Markets: Developments and Prospects, by Maxwell Watson, Donald Mathieson, Russell Kincaid, David Folkerts-Landau, Klaus Regling, and Caroline Atkinson.
    February 1988Officially Supported Export Credits: Developments and Prospects, by K. Burke Dillon and Luis Duran-Downing, with Miranda Xafa.
    April 1988World Economic Outlook: A Survey by the Staff of the International Monetary Fund.
    May 1988Multilateral Official Debt Rescheduling: Recent Experience, by Peter M. Keller, with Nissanke E. Weerasinghe.
    May 1988Primary Commodities: Market Developments and Outlook, by the Commodities Division of the Research Department.
    July 1988Staff Studies for the World Economic Outlook, by the Research Department of the International Monetary Fund.
    October 1988World Economic Outlook: Revised Projections, by the Staff of the International Monetary Fund.
    April 1989World Economic Outlook: A Survey by the Staff of the International Monetary Fund.
    April 1989International Capital Markets: Developments and Prospects, by a Staff Team from the Exchange and Trade Relations and Research Departments.
    July 1989Primary Commodities: Market Developments and Outlook, by the Commodities Division of the Research Department.
    August 1989Staff Studies for the World Economic Outlook, by the Research Department of the International Monetary Fund.
    September 1989Developments in International Exchange and Trade Systems, by a Staff Team from the Exchange and Trade Relations Department.
    October 1989World Economic Outlook: A Survey by the Staff of the International Monetary Fund.
    April 1990International Capital Markets: Developments and Prospects, by a Staff Team from the Exchange and Trade Relations and Research Departments.
    May 1990Officially Supported Export Credits: Developments and Prospects, by G.G. Johnson, Matthew Fisher, and Elliott Harris.
    May 1990World Economic Outlook: A Survey by the Staff of the International Monetary Fund.
    July 1990Primary Commodities: Market Developments and Outlook, by the Commodities Division of the Research Department.
    September 1990Staff Studies for the World Economic Outlook, by the Research Department of the International Monetary Fund.
    October 1990World Economic Outlook: A Survey by the Staff of the International Monetary Fund.
    November 1990Multilateral Official Debt Rescheduling: Recent Experience, by Michael G. Kuhn with Jorge P. Guzman.
    May 1991International Capital Markets: Developments and Prospects, by a Staff Team from the Exchange and Trade Relations and Research Departments.
    May 1991World Economic Outlook: A Survey by the Staff of the International Monetary Fund.
    October 1991World Economic Outlook: A Survey by the Staff of the International Monetary Fund.
    December 1991Private Market Financing for Developing Countries, by a Staff Team from the Exchange and Trade Relations Department.
    May 1992World Economic Outlook: A Survey by the Staff of the International Monetary Fund.
    May 1992Developments in International Exchange and Payments Systems, by a Staff Team from the Exchange and Trade Relations Department.
    August 1992Issues and Developments in International Trade Policy, by a Staff Team led by Margaret Kelly and Anne Kenny McGuirk.
    September 1992International Capital Markets: Developments, Prospects, and Policy Issues, by Morris Goldstein, David Folkerts-Landau, Mohamed El-Erian, Steven Fries, and Liliana Rojas-Suárez.
    October 1992World Economic Outlook: A Survey by the Staff of the International Monetary Fund.
    December 1992Private Market Financing for Developing Countries, by a Staff Team from the Policy Development and Review Department led by Charles Collyns.
    January 1993World Economic Outlook—Interim Assessment: A Survey by the Staff of the International Monetary Fund.
    April 1993International Capital Markets: Part I. Exchange Rate Management and International Capital Flows, by Morris Goldstein, David Folkerts-Landau, Peter Garber, Liliana Rojas-Suárez, and Michael Spencer.
    Note: For information on the titles and availability of World Economic and Financial Surveys published prior to 1988, please consult the most recent IMF Publications Catalog or contact IMF Publication Services.

    The progressive liberalization and internationalization of financial markets has produced, inter alia, significant efficiency gains, including higher returns for savers, lower commissions for security transactions, and lower overhead costs on organized exchanges. To some extent, the financial sector has been undergoing a restructuring similar to that in many other industries.

    SICAVs are French money market mutual funds that are managed primarily by banks; they are similar to open-end investment funds in the United States.

    In some instances, governments have sought to broaden their international investor base. For example, in late 1988 the French Government successfully listed two issues of government bonds, OATs (Obligations Assimilables du Trésor), in the form of American Depositary Receipts at the New York Stock Exchange.

    This figure for worldwide turnover is an estimate based on turnover in the markets surveyed. The rapid growth in foreign exchange market turnover and changes in its composition are discussed further in Annex I.

    Hedge funds are typically incorporated in lightly regulated offshore jurisdictions; in addition, they are not constrained by standard investor-protection regulations (e.g., United States Security Exchange Commission (SEC) registration) because the number of investors is kept small enough, and the minimum investment large enough, to be exempt from such provisions.

    For example, the short-term external funding of Italian deposit money banks has doubled to $160 billion since 1987.

    As is well known, there are considerable problems in accurately measuring international capital flows; see, in particular, International Monetary Fund (1992b) and the recommendations (contained therein) for improving these crucial data.

    In some way, the convergence play is another version of the “peso problem.” In the mid-1970s, the Mexican peso had exchanged for the U.S. dollar at the same rate for two decades. The Mexican interest rate was significantly higher than dollar interest rates, year after year. This phenomenon was dubbed the “peso problem.” The interpretation in 1975, which is now commonplace, was that the probability of a large devaluation was low because empirically the event had not occurred in a long run of data. The devaluation, once it occurred, would be large because of the large divergence in interest rates. The game for any market participant long in the peso was to time the conversion of funds back to dollars before the devaluation and obtain higher than the market return on dollars.

    The harder or more expensive it was in a given market to obtain forward cover during the crisis, the more likely it was that institutional investors would resort to selling of assets.

    Similarly, if a fund had promised always to hold at least x percent of its total assets in foreign-currency-denominated instruments, it could only rebalance currency composition subject to that x percent floor.

    These transactions share an important feature: they all require bank financing—either directly or indirectly. Bank credit is required for short selling, by either borrowing currency or selling it in a forward exchange contract. Hedging utilizes a forward exchange contract with a bank. And liquidating a long security position in a crisis typically requires bank funding of the market maker. A technical analysis of the bank financing requirements of these operations is presented in Annex IV.

    However, this discount was relatively short lived, and the ECU had returned to its theoretical value in both currency and interest rate terms by early 1993. The ECU market began to recover slowly toward the end of 1992, and early in 1993 ECU bond issuance rose.

    Of which, in billions of U.S. dollars—France: 29; Germany: 61; Ireland: 5; Italy: 20; Portugal: 27; Spain: 69; Sweden: 17; and United Kingdom: 40.

    A central bank might also consider undertaking forward sales of foreign exchange. Just as with a private forward contract, the counterparty bank grants credit—but in this case, to the central bank. Although there is no direct limit on the amount that major domestic banks can lend to their own central bank, they will likely be constrained in practice because foreign commercial banks have both capital requirements and credit-line limits on lending to banks in other countries. Annex IV describes in some detail the nature of bank credit operations that would be associated with such a forward contract involving a central bank.

    This disproportionality of available resources between the private financial sector and even groupings of major countries is reminiscent of an earlier epoch in which major countries based exchange rate defenses on private bank funding. For example, through the Belmont-Morgan loan of 1895, the U.S. Treasury defended the dollar gold standard. Loan contracts were written to ensure that supporting lending syndicates did not simultaneously abet the run on the currency. Thus, the Belmont-Morgan contract required the underwriters to support the exchange rate through their own interventions. This provision led J.P. Morgan to organize a syndicate of all the major banks in New York and London to freeze out financing to anyone speculating against the dollar.

    In addition to the VSTFF, central banks may also borrow on a bilateral basis from other central banks for marginal intervention; such loans need not be denominated in ECUs, nor need they be subject to a rigid repayment schedule.

    By Article 16.1 of the articles of agreement establishing the EMS, settlement of VSTFF loans must be carried out first by settling the VSTFF’s ECU claim on the borrower in the creditor’s currency and then in ECUs. However, the Basle/Nyborg agreement allows the creditor country to demand reimbursement of the VSTFF in its currency rather than in ECUs for loans granted to finance intramarginal intervention.

    See Bundesbank (1992), p. 16. In addition to the VSTFF, central banks may, of course, enter into bilateral credit arrangements with each other.

    For example, repurchases by the Bundesbank fell from DM 147 billion to DM 68.5 billion at the beginning of October; and from September 7 to September 30, holdings of domestic bills fell from DM 57 billion to DM 51 billion; net holdings of external assets rose from DM 104 billion to DM 181 billion; and total assets rose from DM 346 billion to DM 380 billion.

    Alternatively, it may be forced to impose capital controls.

    Annex IV explains the mechanics of a squeeze and its consequences for other financial markets.

    This effect is most pronounced in systems where securities markets are made by specialized dealers that are heavily dependent on overnight financing. In universal banking systems a credit squeeze would also have an impact, albeit smaller, on the subsidiaries or departments of banks that specialize in securities trading. These differences in financial structure are discussed in Goldstein and others (1992).

    In some countries, the issue of central bank independence also extends to responsibility for policy decisions on exchange market intervention.

    For example, if holders of foreign currency securities hedge by creating a synthetic put (as do many pension funds and wholesale banks offering over-the-counter derivative products), they commit to a dynamic trading strategy as outlined above.

    Annex VI discusses in greater depth the course of events in each country during the crisis.

    BOT denotes Buoni Ordinari del Tesoro; CCT, Certificati di Credito del Tesoro; and BTP, Buoni del Tesoro Poliennali.

    The squeeze also adversely affected dealers in longer-term treasury securities who were using overnight funding to carry their positions. To alleviate this difficulty, paper eligible for collateral at the refinancing rate was extended to include OATs during the crisis.

    The Bundesbank lowered the official discount rate and the Lombard rate by 50 and 25 basis points, respectively, on September 15; in addition, it permitted a decline in money market rates by almost a percentage point by not fully sterilizing the intervention reflow.

    Empirical evidence on the effects of capital controls is presented in Mathieson and Rojas-Suarez (1993).

    The proposals discussed are not meant to be a comprehensive list of all policy options—nor does the discussion in this section seek to evaluate the merits of these proposals.

    Indeed, the Maastricht Treaty explicitly asks ERM member states to ensure that national central banks are not subject to instructions from national authorities by the time the European System of Central Banks (ESCB) is established.

    The role of uncertainty in determining the timing and dynamics of a speculative attack is analyzed in Annex V below.

    The recent crisis also raises the thorny issue of whether large attacks against currencies change the fundamentals; this possibility is explored in Annex V. For example, sequential attacks that lead to the devaluation of a number of currencies early in the sequence can erode the competitive position of countries whose exchange rates have remained fixed, thereby possibly increasing their own susceptibility to attack. In the context of the recent ERM crisis, such considerations may have played a role in the attack on the French franc; that attack can also be viewed as merely a case of “parity busting” that would never have taken place without the profits made by trading rooms in earlier successful attacks.

    The April 1992 survey results are still incomplete. The countries listed—excluding Germany, which was not included in the 1989 survey—accounted for nearly 90 percent of total net turnover in 1989.

    This total is adjusted for double counting of transactions within national markets, but not for double counting of international transactions.

    Worldwide net turnover is net of double counting of international transactions.

    The wholesale classification in Table 7 corresponds to the “interbank” classification provided by the authorities’ reports on the survey results. This therefore combines dealers’ trade both with other dealers and with nondealing banks. This category also includes transactions in organized derivatives exchanges and OTC options markets that are dominated by transactions between banks and nonbanks that deal in foreign currencies. “Retail” business therefore includes only transactions with nonbank financial institutions that are not foreign exchange dealers, and with nonfinancial customers.

    The 1992 data are not directly comparable, since they give the share of wholesale trade in total net turnover and disaggregate derivatives transactions between dealers and customers, which was not possible in earlier years.

    For example, hedging a deutsche mark/pound sterling forward exposure can require four interbank transactions: U.S. dollar/deutsche mark and U.S. dollar/pound sterling spots and forwards. See Healey (1992), p. 412; and Walmsley (1992), p. 43.

    In this context, automated dealing systems refer to the interdealer dealing systems marketed by Reuters, Quotron, and Telerate, for example, and not to automated dealing systems provided by banks to their clients to facilitate the handling of retail orders. According to the 1992 surveys of activity, 24 percent of gross turnover in the United Kingdom and 32 percent of gross turnover in the United States was traded through these proprietary systems.

    Hedge funds are privately subscribed funds that take highly leveraged speculative positions or that engage in arbitrage. Most of these are largely unregulated offshore funds. No precise figures for total assets of offshore funds—of which hedge funds are the majority—are available, but estimates run as high as $250 billion.

    Japanese funds managers were not included in the surveys.

    Ferrill Capital Management, as quoted in Institutional Investor (December 1992), p. 36.

    A.M. Best Co. (1992), p. 11. Information on foreign investments of other types of U.S. insurance firms is also unavailable.

    More detail on the investment practices of Japanese life insurers is presented in Table A3.

    The data suggest that in France the FCPs (fonds commun de placement) invest a much greater proportion of their assets abroad than do the SICAVs (sociétés d’investissement à capital variable). The second group of institutions held less than 1 percent of their assets abroad in September 1992, but the SICAVs and FCPs together held over 4 percent abroad.

    The two-day convention was established in the days when confirmation and settlement required the transfer of documents by mail. For some bilateral trades (for example, the U.S. dollar/Canadian dollar and U.S. dollar/Mexican peso), the value date of a spot trade is now only one day.

    Since the spot value date is two working days later, there are also two days before the spot date during which currency could be delivered: today and tomorrow. Same-day delivery is generally only practical when the time zone of the foreign currency being delivered is behind that in which the deal is struck. For example, dollar/pound deals can be struck in London because the time difference allows instructions to be sent to and processed in New York on the same day. Deliveries before the spot date are generally made on a swap basis with the subsequent exchange occurring on the next day. Thus, there are today/tomorrow (overnight, O/N) and tomorrow/next swaps (T/N). Usually, the T/N swap is completed on the original spot value date, and these swaps are generally used to roll over spot positions.

    The specifications can be relaxed in variants of the forward contract. For example, an option forward contract has no specific delivery date. Instead, the holder of the contract may conclude the exchange at any time before the contract expires.

    The value date of a one-month forward contract will generally be the same date as the current spot value date in the following month, but this is complicated when the calendar date is not a working day (when the forward value date is usually moved forward a day) or when the spot value date is the last day of the month (when the forward value date is the last working day of the following month). See Walmsley (1992), pp. 193-94, for a discussion of the calculation of forward value dates.

    A11 options traded in the United States are cleared through the Options Clearing Corporation. Options and futures traded in the United Kingdom are cleared through the International Commodities Clearing House.

    CHIPS volume currently exceeds Fedwire volume, which is the U.S. domestic electronic payments system.

    A net due-to position means that a bank owes more funds to other participants on the payments system than it is owed by the other participants.

    CHAPS is scheduled to move away from net settlement over the next few years toward a system of gross settlement with collateralized overdrafts permitted.

    The guarantee on CHIPS is backed by member collateral. Each member has a credit limit that is the sum of bilateral credit limits permitted by each of the other members.

    In Belgium, the Banking Commission has the authority to set statutory limits on net foreign exchange positions but has not exercised this regulatory power.

    This fraction is delta, which is a measure of the probability that an option will be exercised. See the discussion of dynamic hedging in Annex II.

    If the option is out of the money, this requirement is reduced, subject to a minimum requirement of the premium plus 0.75 percent of the value of the underlying currency.

    To allow dealers the freedom to take speculative positions during trading hours, the intraday position limit is generally much higher than the overnight position limit, since only overnight positions must be financed by borrowing or asset sales.

    Multinational banks will often impose global trading limits, constraining the activity of dealing rooms in different locations.

    Counterparty credit limits would be set in conjunction with the bank’s loan department.

    Annex V, which surveys economic analysis of the timing and dynamics of an attack, complements this discussion.

    This is for relatively short-dated swaps—longer-maturity swaps will involve the exchange of a series of interest payments in the two currencies as well as the swap of principal.

    This behavior by the large banks primarily reflects internal risk management policies, which, in turn, have been shaped by the weakened financial position of the major commercial banks and the market discipline imposed on the investment banks.

    Retail in this sense means to banks and investors that do not have access to large-scale funds in a given currency.

    Even the quoted bid-ask for customers may not be indicative of the true bid-ask at which the dealer will trade: in a crisis, it may be reserved for special customers in small amounts, so the true bid-ask spread may be much wider.

    The pension fund will also pay a bid-ask spread on the purchase and sale of deutsche-mark-denominated assets.

    A dealer will price the pension fund’s securities based on the price that he will eventually receive less his net funding costs while he holds the securities in inventory.

    The specific instruments that speculators use, and the associated costs are discussed in Annex IV.

    The ideas examined here were developed in Salant and Henderson (1978); Krugman (1979); and Flood and Garber (1984b).

    For simplicity, the domestic banking system is not considered in this argument.

    The discussion so far has concentrated on speculative attacks based on market fundamentals. But this is not the only kind of speculative attack treated in the recent literature. A second kind is known as an arbitrary speculative attack or a pure speculative attack, and this kind results from self-fulfilling speculative prophesy concerning the shadow exchange rate. In these models, all speculative attacks are based on the difference between the controlled exchange rate and the shadow rate. If the shadow rate contains an arbitrary speculative element, the timing of an attack on a controlled parity is arbitrary also.

    At the time of an attack, an attack based on “fundamentals” and an arbitrary attack look the same. Afterward, however, the exchange value of the currency attacked arbitrarily is predicted to depreciate indefinitely in excess of the depreciation that would be appropriate based on fundamentals.

    This policy involves continuous sterilization of reserve losses by the central bank whose currency is being attacked.

    The possibility of such multiple equilibria was developed in Flood and Garber (1984a).

    For a discussion of the macroeconomic developments leading up to the crisis, see International Monetary Fund (1993).

    These figures refer to investment in Italy, Portugal, Spain, Sweden, and the United Kingdom.

    Although this section focuses on the currency crises in Italy, the United Kingdom, France, and Sweden, Table A11 summarizes the central bank policies of all European countries involved in the exchange rate turmoil during the fall of 1992.

    The lira crisis, the first crisis in Italy since the liberalization of capital markets in 1988, occurred in an environment in which capital markets have been modernized with screen-based trading for government paper; interest rate futures contracts on lira bonds are now traded on LIFFE, on Marché à Terme International de France (MATIF), and on Mercato Italiano dei Futures (MIF); and reduced market segmentation and improvements in transactions technology have made possible much larger and more rapid movements of funds.

    Data on gross reserves reported in this section are taken from International Financial Statistics. These figures understate the decline in net reserves held by weak-currency countries owing to the large-scale foreign borrowings undertaken by these countries.

    The discount rate is applicable to “ordinary” advances to banks from the Bank of Italy—this rate is charged on drawings on the normal credit line that the Bank of Italy extends to banks. In addition, there are special advances whose rates are markups on the discount rate.

    BOT denotes Buoni Ordinari del Tesoro; CCT, Certificati di Credito del Tesoro; and BTP, Buoni del Tesoro Poliennali.

    In the United Kingdom, discount houses intermediate between the Bank of England and the commercial banks in the channeling of central bank liquidity. To obtain good funds–deposits in the central bank—the discount houses offer the Bank of England bills of maturity up to three months. The minimum dealing rate is determined by the stop rate or lowest rate that the Bank of England will accept on paper. Occurring infrequently, changes in the minimum dealing rate are taken as signals of Bank of England policy changes and therefore are reflected throughout the interbank market and in bank base rates. As banks set their base rates in lockstep with the minimum dealing rate, and mortgage rates tend to move closely in line with base rates, an increase would almost certainly have boosted interest rates on millions of mortgages.

    For an explanation of synthetic puts, see the final section of Annex II.

    Dealers in longer-term government securities, who carry positions through overnight funding, were also harmed by the squeeze. To alleviate this difficulty, paper eligible as collateral at the Bank of France’s refinancing rate was extended to include OATs (Obligations Assimilables du Trésor) during the crisis.

    Banks can borrow from the Riksbank in unlimited amounts, though the interest cost of the funds rises incrementally on a scale depending on the amount borrowed, the bank’s equity capital, and overall borrowing of the banking system. Plotting the amounts borrowed against the rising incremental steps generates a supply function of good funds. The marginal rate at any time is the interest rate at which the overall demands for these funds cross the step supply function. The marginal rate might rise by raising the steps in the supply function across all steps or for the single step at which demand crosses the function. Alternatively, of course, the marginal rate might rise with a shift in funds demanded and with the step supply function held fixed. Funds from the Riksbank are also provided through repurchase agreements and open market operations.

    Losses had arisen particularly in real estate lending. By the end of 1991, bank credit losses had reached SKr 45 billion, and in October 1992 losses of a further SKr 60 billion were anticipated for 1992.

    The crisis revealed some of the remaining segmentation in European financial markets. For example, rates for the deutsche mark London interbank offered rate (LIBOR) fell well below the deutsche mark Frankfurt interbank offered rate (FIBOR). However, in the “gentleman’s agreement” between the Bundesbank and the German banks, only a limited amount of deutsche mark deposits can be loaned from German banks in London to their branches in Germany without triggering a reserve requirement.

    A review of some industrial countries’ experience with capital controls is contained in Mathieson and Rojas-Suárez (1993).

    Two changes in the CHIPS network took effect on October 1, 1992. First, the monthly charge for use of the system was reduced from $1,600 to $1,500, and the charges per transmission were reduced from 20¢ to 18¢ for amounts under $80,000 and from 15¢ to 130 for amounts above $80,000. Second, the CHIPS format was made compatible with the SWIFT format. Neither of these changes is likely to have affected the volume of transmissions significantly, however—and if they had, the volume of transmissions would not have fallen to pre-September levels as it did in December.

    There was an anomaly in the one-month forward spreads for Sweden, where the ask-bid spread rose to 8 percent in September while the spread on one-month interbank loans had jumped to 20 percentage points.

    The PIBOR futures contract reached its intra-day limit price change of 60 bp twice during the week ended September 18, leading to temporary suspensions of trading. At the end of the week the limiting price change was increased to 120 bp.

    While the OTC currency options market essentially dried up during the week ended September 18, the OTC interest rate options market fared somewhat better, especially in the shorter maturities. Implied volatilities and bid-ask spreads were high. However, most of the activity in interest rate options was directed through the exchanges.

      You are not logged in and do not have access to this content. Please login or, to subscribe to IMF eLibrary, please click here

      Other Resources Citing This Publication