Banking Soundness and Monetary Policy

13 Derivative Markets and Financial System Soundness

Charles Enoch, and J. Green
Published Date:
September 1997
  • ShareShare
Show Summary Details

Recent financial and exchange rate crises have highlighted the potential for derivative products to have macroeconomic consequences and to obscure the meaning of capital account categories in balance of payments data. Although the existence of large off-balance sheet derivative positions will not necessarily trigger a financial crisis, derivatives can affect the magnitude and dynamics of a crisis. Sales of a weak currency by domestic institutions to meet margin calls against offshore derivative positions, the dynamic hedging of derivative positions, expanded opportunities for hedging and speculation, and the general unification of markets that derivatives permit certainly exacerbated both the turbulence in Europe’s exchange rate mechanism and Mexico’s peso crisis. Moreover, the increased opaqueness of the national balance sheet—even to authorities—brought about by cross-border derivative usage causes crises to evolve in unexpected directions, sometimes defying classical remedies.

This paper reviews the ways that derivatives can affect macroeconomic stability—crisis dynamics, miscalculations of market risk, and excessive risk taking by the financial system through evasions of prudential regulations. From the IMF’s own operating perspective, we describe the possible implications of derivatives for interpreting capital account data. Finally, the paper outlines how authorities can adjust to the existence of derivatives in their policy formulations and determine how to increase official awareness of the presence of derivatives in their markets. Extensive examples from Mexico’s recent financial crisis are used to illustrate some of these points.

It is not easy to adjust surveillance operating procedures for the existence of derivative products, and it has proven difficult for regulators even to determine the amounts outstanding at any given time. The triennial surveys undertaken by the Bank for International Settlements (BIS) provide only snapshots of the amounts outstanding in the industrial countries and financial centers and only for broad categories of products. They do not indicate the risk position of the individual institutions that operate in this market. Indeed, regulators have changed their approach from trying to determine and explicitly control the position of key banks that operate these markets to one of allowing banks to operate their own risk models and relating capital coverage to value at risk. This may be sound for microprudential purposes, but it does not help a country determine its own aggregate risk position.

Extent of Derivative Markets

A recent BIS survey indicates that the notional value of over-the-counter (OTC) derivative products outstanding was $47 trillion in March 1995, about 55 percent of which were cross-border transactions.1 Most of this amount consisted of simple interest rate products, such as swaps, and most cross-border transactions occurred between industrial countries.2 For other derivative products, there are, nevertheless, large notional values outstanding in absolute terms—equity-based products and structured notes and options, which may be quite complex—and these are increasingly used in key emerging-market countries.

Specifically, of the $47.5 trillion in OTC notional values, 61 percent was in interest rate instruments, and 37 percent was in foreign exchange instruments, including outright forwards and swaps. Equity contracts amounted to 1.25 percent, and commodity-related instruments were 0.75 percent of the overall notional value. Exchange-traded contracts outstanding, net of double counting, amounted to $8 trillion, almost all of which were interest rate contracts. Gross market values (replacement costs) were $2.2 trillion for OTC contracts.

Of the interest rate products, 50 percent were cross border, while 56 percent of foreign exchange products were cross border.3 For equity products, cross-border position data are not reported by the BIS. The market or replacement value of outstanding derivatives is about 4.6 percent of the notional value. For comparison, the total stock of international securities in the OECD countries was $26.3 trillion, and international banking assets, excluding securities holdings, were $8.3 trillion in March 1995.4

Small Smorgasbord of Noteworthy Derivative Products

While the list of exotic derivatives products grows almost daily, most derivatives outstanding are quite simple, consisting mainly of forward contracts, swaps, and basic options, whose notional values are indicative of the magnitude of the market risks that are being acquired or hedged. Structured notes, however, are implicitly highly leveraged products whose notional values generally underestimate significantly the magnitude of the risks taken. A brief listing of products important in the Mexican currency and financial crisis, for example, would concentrate on a few types of swaps and structured notes; these products in their Mexican manifestations illustrate the important connection to crisis dynamics. Nevertheless, they are generally available in offshore OTC markets for a wide variety of other emerging markets.

A generic swap of yields is simply an exchange of the percentage return on one type of asset for the percentage return on another, multiplied by a predetermined notional value. The swap may involve a periodic exchange of yields of a fixed period of time and settlement of the net amount due. Specifically, for a given currency, an interest rate swap involves an exchange of a fixed interest return for a floating return, or perhaps one floating interest rate for another. An equity swap generally involves a periodic exchange of the return on a given equity or equity index, including dividends and capital gains, for some interest yield, multiplied by a notional value in a given currency.5 A structured note involves a deposit of a given amount of principal by the buyer with the seller. The payoff of either interest or principal is set as a function of some underlying market value, such as an exchange rate or interest rate.

In the Mexican context, these generic products took several forms, many of which were designed to circumvent various domestic prudential regulations.6 Similar products play the same role of avoiding regulation or taxes in other emerging and industrial country markets. They also are means of avoiding capital controls in the form of taxes on inflows or outflows involving individual transactions.

Tesobono Swaps

Tesobono swaps were offshore derivative operations used by Mexican banks as a means of leveraging tesobono holdings. In a tesobono swap, a Mexican bank would receive the yield earned on tesobonos and deliver LIBOR plus, multiplied by a notional amount of dollars. By this operation, Mexican institutions could circumvent regulations of the Comisión National de Valores that prevented the holding of financial assets on margin in Mexico. Industry sources in Mexico report that about US$16 billion of tesobonos were involved in tesobono swaps at the time of the devaluation (December 1994).7

The leverage involved in tesobono swaps can be most readily examined by considering first the nearly equivalent (in terms of payoff) tesobono repurchase agreement. As an example, consider a New York investment firm that is willing to lend dollars for one year against tesobono collateral. The firm engages in a repurchase agreement with a Mexican bank to buy tesobonos at some agreed price to resell them in a year at the original price plus a dollar interest rate.8 Suppose the dollar rate is LIBOR plus 100 basis points, with the interest rate to be settled and repriced quarterly.

In 1994, the typical tesobono repo had a maturity of one year and required a margin between 10 and 20 percent, which produces a leverage of between 9/1 and 4/1. The Mexican counterparty would, for example, buy $500 worth of tesobonos in Mexico and sell them to the New York firm for $400. The tesobonos would be delivered to the New York firm through its custodial account in Mexico. Note that the tesobonos would then have been held in the “foreign address” category, although their ultimate holder had a domestic Mexican address. The gain to the Mexican bank is that it pays LIBOR plus 100 basis points to finance tesobonos that may pay the equivalent of LIBOR plus 300 basis points. The gain to the U.S. lender is that it gets to place dollar funds at LIBOR plus 100 basis points against good collateral.

When the crisis arrived, the dollar market values of tesobonos suddenly fell. This resulted both indirectly from rumors that capital controls might be imposed and through the failed auction of January 1995 in which the government accepted an unfavorable yield. The fall in market value reduced the value of the collateral and triggered margin calls to deliver dollars or close out the positions.

Suppose, for example, that the typical tesobono fell by 15 percent in dollar value, to the extent that it could be valued at all. For the tesobono repo in the example above, collateral is now insufficient to the extent of $75, and a margin call to deliver this amount in cash is sent to the Mexican bank. The Mexican bank must now either go to the exchange market at the depreciated peso exchange rate to acquire the $75 or close out the position. To close the position requires the delivery of $400 in cash plus cumulated interest.

A tesobono swap places both parties in the same risk position as a repurchase. Suppose the New York firm swaps tesobono yield for LIBOR plus 100 basis points against a US$500 notional principal and requires US$100 as collateral from its Mexican counterparty. The payoffs to the two counterparties are identical to those of the repurchase. To hedge, the New York firm will purchase US$500 in tesobonos directly from the market, so once again the tesobonos will be held by “foreign addresses,” although Mexican domestic residents will bear the tesobono risk.9 Again, if tesobono prices jumped downward because of suddenly increased default risk, the New York counterparty would send out a margin call.

The scramble for dollars to cover such margin calls and position close-outs was associated with the currency turmoil of January and February 1995, and indeed with the large attack on the Banco de Mexico’s reserves on December 21, 1994.

Equity Swaps and Equity Repurchase Agreements

The market in equity swaps also existed to avoid financial market regulations: the regulation that prohibited buying securities on margin and the regulation that limited the possibility of short selling, regulations that appear in many emerging markets. Market participants have characterized the market in offshore Mexican equity swaps as very large, but they were not as explicit about orders of magnitude as in the case of tesobonos.10

The benefits to this market’s participants are obvious. Speculators can leverage and gain larger positions, and hedgers of long positions held either directly or implicitly in the form of options can short stock to cover their positions. The market arose to complete the liquidity of the rising domestic equity and warrant markets.

Technically, stock swaps might be swaps or repurchase agreements. The repurchase is easier to analyze directly, though swaps predominate. A repurchase transaction involves selling shares currently at a given price and repurchasing them for a fixed price one year later, for example. The interest rate payable to the purchaser of the stock, typically LIBOR plus 200 basis points, is settled and repriced quarterly. The initial purchase price might be $100 for $125 worth of shares, so there is a 20 percent margin.11

The offshore firm that initially purchases the shares has the right to dispose of the shares if it wishes. If it sells the shares on receipt, it has effectively used the repurchase to short Mexican shares because it is short shares through the forward leg, thereby circumventing the domestic constraints on short selling. If the initial seller of shares in the repurchase is an offshore subsidiary of a Mexican bank, the Mexican bank has circumvented the regulation against a Mexican bank’s lending equity for short selling. On maturity, the offshore lender may or may not resell actual shares to the borrower of funds. In 1994, many transactions were for physical delivery, but others might have allowed for cash settlement only. In many cases, individuals, especially those who were borrowing against controlling blocks of shares, wanted the stock back.

With the collapse of the Mexican peso and stock market, the margin in the equity repos was more than wiped out, triggering margin calls from New York in the form of cash or treasury bills. The response varied by client: some institutions claim that the margin calls triggered enormous sell orders in the stock market to acquire sufficient cash to close out positions while others report that their clients invariably delivered new funds rather than have positions closed. In either case, the Mexican institutions and individuals engaged in these repos had to sell pesos to get margin or close out their position, adding to the turmoil of the exchange and stock markets.

Again, an equity swap establishes leveraged positions equivalent to a repurchase agreement. Suppose that a New York firm agrees to swap the total return on a Mexican stock for LIBOR plus 200 basis points on a notional amount of US$125 and requires US$25 in collateral from its Mexican counterparty. To hedge its short equity position, the New York firm directly buys US$125 in the Mexican equity, thereby appearing as a foreign investor in Mexican shares. The risks borne by the two counterparties are the same as in the repurchase example above—the Mexican counterparty is taking a long position in Mexican shares and a short position in short-term dollar loans, while the New York counterparty has only a long position in short-term dollar loans.

It is a general view among market participants and bank regulators that prohibiting the holding of financial assets on margin in Mexico was a mistake. The effect was merely to drive such activity offshore rather than to limit it, and this made the magnitude and nature of the leveraging of positions opaque to the authorities.

Cetes Swaps

Cetes swaps and repos also were undertaken in large volumes with offshore counterparties.12 These deals were similar to the tesobono swaps and repos except that the collateral was cetes. These operations were regarded as more aggressive than the tesobono swaps because of the much greater currency risk. In return, however, there was a much greater spread between cetes and LIBOR interest rates. The typical deal might require a payment of LIBOR plus 500 basis points to the dollar lender, while cetes might yield 25 percent with an anticipated depreciation of 4 percent, for a spread to the Mexican bank entering the deal of 10 percent (based on LIBOR of 6 percent). Margin requirements were higher for cetes swaps than for tesobono swaps because of the currency risk, so the devaluation and surge in cetes interest rates triggered large margin calls.

Brady Bond Swaps

In addition to the swaps in short-term Mexican paper, there was a large volume of Brady bond swaps, which provided a means to Mexican banks of leveraging Brady bond positions. This allowed Mexican banks to reap a return both on the rising dollar yield curve and Mexico country risk. These were typically three-month contracts booked as repos with interest rates on the dollar legs set at LIBOR plus 100 basis points.

Structured Notes

Structured notes are investment vehicles with coupon payments and principal repayments that are driven by formulas that can vastly leverage the initial capital invested. Nevertheless, in value accounting systems they can be booked as normal investments and in the currency denominated in the prospectus. More than simply magnifying the usual market risks associated with investment positions, structured notes provide an easy method for circumventing prudential regulations on currency positions or interest rate mismatches.

During 1994, Mexican financial institutions took large positions in structured notes with investment houses in New York. Booked as claims with dollar principal and dollar payoffs, these notes were currency bets that allowed the banks to leverage their investment into a short-dollar and long-peso position to take advantage of the positive interest rate spreads between peso and dollar money markets.13 Because the notes were reported by the banks as dollar assets, however, the accounting rules in Mexico allowed them to be booked as a dollar position, so that they were not counted against the regulatory limit on net currency positions of 15 percent of capital. When the devaluation occurred, Mexican banks had much larger net short-dollar positions and losses than regulators had realized.

Structured notes exist in many forms. For example, a Mexican bank might buy a note with a one-year maturity from a New York investment house for $1. Most major New York financial engineering firms sold such products—including Bankers Trust; Merrill Lynch; Bear-Stearns; Donaldson, Lufkin & Jenrette; and Morgan Stanley. The coupon on the note and the principal on the note are payable in dollars. Suppose that the coupon is 85 percent. The principal repayment, however, depends negatively on the peso value of the dollar—suppose it is 1 + 5(s0s1)/s1, where s0 is the initial peso value of the dollar and s1 is the value at maturity. If the peso has depreciated by 50 percent at maturity—from, say, three to six pesos per dollar—the principal repayment will be -$1.50; the overall payout is then –$0.65. Note that this is the payoff structure of a position that is short $3 at a market dollar interest rate of 5 percent per year and long 12 pesos at a market peso interest rate of 25 percent per year (for example, in the form of cetes). Effectively, the initial $1 investment has been leveraged four-fold.14

Overall, the New York investment house would have, through the payoff formula, a position equivalent to being short 12 pesos worth of cetes and long $3 worth of treasury bills. In addition, it has the initial $1 from the sale of the note. To hedge, it could buy the peso by investing in cetes and short the dollars by borrowing them to buy cetes; it would then appear in the on-balance-sheet accounts as a foreign buyer of a peso-denominated asset rather than as a dollar-denominated lender.

Mexican banks face Banco de Mexico regulations that restrict foreign exchange net positions to a maximum of 15 percent of capital. According to the regulatory definitions of what constitutes foreign exchange—an asset or liability whose principal and coupon are denominated in a foreign currency—the $1 originally paid to acquire the structured note would enter the books as a long $1 position, even though its payoff implied a short $3 position. In addition, some banks could count it toward their liquidity coefficient, required for foreign currency-denominated liabilities, because its short maturity allowed it to be classified as a liquid deposit.

A bank that borrows the initial $1 to purchase the note would have a balanced foreign exchange position on net for regulatory purposes. When the Mexican exchange rate was devalued in December 1994, however, the asset value would have fallen to zero, leaving the bank with an unbalanced dollar liability. In scrambling to cover this imbalance, the Mexican banks had to sell pesos, contributing to the December attack on the Banco de Mexico.

This type of structured note was a financial engineering device to circumvent prudential regulation. Only the principal was booked, in accordance with value accounting principles. The structured note payoff formula component was not booked—it was an off-balance-sheet item. That was the accounting trick—one can alter the nature of the booking through a complicated payoff formula.

Accounting regulations for the determination of foreign exchange positions have recently been changed to be consistent with a risk accounting principle. Principles of risk accounting are not generally accepted or understood. It is still not clear when risks are being taken by a bank. Risk accounting is currently a subject of discussion among regulators in industrial countries, and they are coming down on the side of self-regulation because they feel they are unable to determine the risk posture of a derivative book in a timely manner. In the past, the long position in dollars was, through value accounting principles, defined as a dollar asset without taking into account the sensitivity of the asset to the payoff characteristics. The new definition of foreign exchange position leans on a risk accounting principle: the position is classified as unbalanced if it generates potential gains or losses from the movement of the exchange rate.

Cracking Capital Import Taxes

Taxes on the acquisition by foreign addresses of domestic securities have emerged in recent years as a means of stemming capital inflows. They sometimes have been imposed differentially by maturity of asset and type of asset. In general, such taxes have been successful in that they have placed a wedge between domestic and foreign yields on similar assets. Of course, they can be breached by the usual invoicing subterfuges, but market participants also have used financial engineering to circumvent the taxes. Specifically, let us presume that an enforceable tax is placed uniformly on all forms of gross inflows. Then, any positive net inflow will incur the tax, but gross transactions will move offshore. As an example, instead of acquiring an equity position directly, a foreign investor will buy an offshore equity swap from a domestic resident who can hedge without a tax. If the domestic resident has a lower credit rating, an export of capital in the form of margin will be recorded. There will be no taxable inflow, but foreigners can take risk positions in domestic assets.15

If the tax is differential across types of assets acquired from abroad, the net inflow will tend to take the form that incurs the lowest tax. The risk and maturity characteristics of the inflow can then be resculpted through offshore derivatives to a more desirable form. For instance, if equity investment is given a better treatment than short-term fixed interest securities or bank deposits, the inflow will take the form of a stock acquisition together with an equity swap that converts it on net into a floating interest loan of foreign currency. Even the maturity of the loan can be adjusted with an attachment of a stringent margining provision that permits the offshore creditor to realize cash on call.16

Effects of Derivatives on Balance of Payments Accounting

Among the rationales of balance of payments accounting is to ascertain the stability of capital flows of on-balance-sheet movements of assets. Typically, balance of payments accounting data are used to measure how long capital will remain in a country.17 Various categories of the capital accounts have been interpreted as indicative of the nature of capital inflows or outflows. Direct investment, for example, has been considered a more stable form of investment than portfolio investment or the foreign acquisition of bank claims. Foreign acquisition of short-term fixed interest products is generally regarded as a speculative flow. Balance of payments accounts are also used to measure the foreign exchange position of a country and, in times of crisis, to determine the potential outflow of foreign exchange through speculation or covering operations by holders of domestic liquid assets.

The revolution in global finance and notably the explosion in the use of derivative products have rendered the use of balance of payments capital account data even more problematic than it has been in the past.18 Balance of payments accounting data use on-balance-sheet categorizations, and they are based on value accounting principles to book and categorize asset values. They ignore almost completely the existence of derivatives and their role in reallocating who bears market risk. This problem has grown with a massive explosion in the use of derivative products and especially in the use of cross-border products.

For example, the acquisition of a large block of equity is classified as direct investment, but a foreign buyer may be acquiring the block simply to hedge a short position in equity established through a derivative position. In the case of the equity swap described above, the foreign investment firm that sells the swap must acquire the shares to form a hedge. If the swap is large enough, it may be booked as direct investment because the offshore swap position is not included in the capital accounts, although the investment house in fact is making a short-term floating-rate loan in foreign currency. Declines in equity values or the exchange rate will then generate instantaneous exchange market pressure as margin calls are made or positions are closed. If the buyer of the swap is a domestic resident, the capital import effectively takes the form of short-term foreign currency-denominated borrowing, but the leveraged equity risk, and even the long-term control, remains in the hands of the domestic resident. Thus, the “direct investment” turns into the hottest of money. In a similar manner, direct investment in the form of reinvestment of profits can be converted into short-term funding through an equity swap.

Alternatively, a foreign program trader may acquire the domestic stock index in the cash market while selling forward in the offshore OTC index market. On net, the trader has a zero position in equities but in the balance of payments accounts appears as a portfolio investor in domestic equities. If the opposite positions are taken by domestic residents—a sale of equities in the cash market and a forward purchase in the derivatives market—the net equity risk position for domestic residents is unchanged, though domestic residents are now in effect short-term foreign currency borrowers.

To the extent that they start with zero replacement values, as in the case of swaps and forwards, derivative products do not affect measured net capital inflows or outflows; but they blur the information in subcategories of the capital accounts.19 Specifically, they make a mockery of the use of capital account categories in measuring the aggregate short foreign currency position of an economy.

Some Policy Implications

From the explosion in the use of derivative products has emerged a blind spot in both national and international surveillance of capital markets. Through derivatives, both individual institutions and financial systems can be put at risk in magnitudes and from directions completely unknown to regulators. This problem arises because derivatives are ideal means of avoiding prudential regulations, given the universally slow adjustment of accounting principles to the advent of these products. Capital controls, the first cousin to prudential regulation, likewise are reduced in effectiveness. On a more parochial level, the accounting principles by which balance of payments data are gathered are being made increasingly obsolete. For each country, the extent of the problem is unknown because comprehensive data on derivatives are collected only at long intervals, and even the triennial BIS data are not broken down for emerging-market countries.

Given the data gaps emerging from the large-scale use of derivatives, it is tempting to restructure the data gathering of regulators and multilateral surveillance institutions to provide information about derivative positions on a systematic basis. Unfortunately, gross derivative positions alone are not sufficient statistics of risk positions, and for IMF surveillance purposes any attempt to establish aggregative net risk positions of the financial sector will fail.

It has proved impossible for bank supervisors in industrial countries to make sense of derivative positions, and it will likewise be impossible for supervisors in emerging-market countries to get a clear picture of the net risk positions of their financial institutions when such institutions are free to do cross-border business. Indeed, the G-10 banking supervisors are currently changing their approach to bank supervision—discarding compliance-based supervision in favor of supervising the quality of the banks’ risk management systems and relying on an agreed measure of capital at risk—because it is simply not realistic to get timely data on the net risk position of individual institutions, not to speak of getting such information on the entire financial sector. The problem is that gross financial positions are irrelevant. For example, why worry about a $3 billion long-dollar position in the international OTC options market held by an Indonesian bank if this position is offset elsewhere on its books? A complete picture of the bank’s on-and off-balance-sheet positions would be necessary almost hourly because such positions can change very fast.

Even more complicated, computing such net risk positions requires dealing with whether to count dynamically hedged positions that rely on continuous price movements or somehow adjust for price breaks that make these hedges useless. Finally, international financial sector surveillance would presumably be interested only in macroprudential issues—aggregative net risk positions—rather than in the risk positions of individual institutions. As such, it would be necessary to net open positions among all the major institutions, providing that a scheme can be hatched to avoid changes on the national treasury by making transfers from the winning institutions to the losers.

The experience of the G-10 supervisors could serve as a guide: forget about establishing data on net risk positions; instead make sure that institutions have good risk management systems in place. That translates into assuring that financial authorities have a good grasp of the activities of their financial institutions in the aggregate. Are many of them taking speculative positions on one side of the market, raising the possibility that losses or margin calls could have macroeconomic implications? In other words, are banks doing anything that in the aggregate could undermine macro stability if their bets do not pay off? Finally, authorities must ask themselves if they fully understand the implications of any liberalization measures that allow commercial banks to trade in OTC derivatives.

“The global nature of the markets is underlined by the large amount of business contracts with counterparties located abroad.” Bank of International Settlements, “Central Bank Survey of Foreign Exchange and Derivative Market Activity: 1995” (Basle: BIS, May 1996), p. 24.

Thus, the problem of inferring market risk from balance of payments data applies especially to the positions of industrial countries. Academic investigations into the lack of cross-border portfolio diversification based on capital account data are fatally compromised by this gap in the data.

See BIS, “Central Bank Survey,” p. 23, Table D3.

“Notional amounts do not reflect the payments obligations. They do reflect the price exposure in the underlying markets and they are useful for comparison with the underlying for the amounts outstanding. See BIS, “Central Bank Survey,” p. 24.

Interest and equity swaps do not involve initial and final payments of principal or notional value, although the counterparty with the greater credit risk may have to deliver some collateral. Currency and foreign exchange swaps do require initial and final delivery of principal. A foreign exchange swap, generally a very short-term deal, is a combination of a spot sale of currency and a forward purchase—it packages in a single deal both foreign exchange market legs of the familiar interest rate parity arbitrage operation. A currency swap similarly requires an initial and final exchange of principal amounts, but it is of longer maturity and involves periodic exchanges of interest on the principal amounts in the two currencies.

We adapt the following description of Mexican derivative products from Peter Garber, “Managing Risks to Financial Markets from Volatile Capital Flows: The Role of Prudential Regulation,” International Journal of Economics, Vol. 1, July 1996; and Peter M. Garber, and Subir Lall, “Derivative Products in Exchange Rate Crises,” paper prepared for the Federal Reserve Bank of San Francisco’s Conference on “Managing Capital Flows and Exchange Rates: Lessons from the Pacific Basin,” September 26-27, 1996.

About US$29 billion of tesobonos were outstanding at the end of December 1994. About US$16 billion of outstanding tesobonos on December 19, 1994 were held by “foreign addresses.”

In the swap form of the deal, only net amounts were due in each settlement period.

The tesobonos associated with these swaps were held by foreign addresses. If the swap position was suddenly closed, domestic addresses would have bought back the tesobonos at some predetermined price through a termination option that might have required a payment penalty of 10 or 20 basis points. Then the data should indicate a sudden shift in ownership from foreign to domestic addresses. Payment of margin shows up as an exchange market disturbance of a smaller magnitude, but it did not generate a shift in the ownership categories.

Offshore equity swap markets also exist for Malaysia and Thailand, among others, also to avoid curbs on short selling and leveraging.

The dollar purchase price is delivered to the initial seller of the shares—the funds are used to purchase $100 worth of shares. Thus, the initial seller of shares is long shares through the forward leg of the contract and has a 4 to 1 leverage of initial capital. The shares are delivered into accounts of the offshore financial institution that is lending funds; typically, these accounts are at Citibank, with Mexico acting as custodian. In turn, Citibank has stock accounts in Indeval, the Mexican securities depository.

These were undertaken for ajustabonos and pagares as well as for cetes as the supply of cetes declined in 1994.

In Malaysia, these instruments, known as “principal adjusted coupon notes,” serve the same purpose of providing leverage in acquiring domestic currency positions through foreign exchange financing.

As a safety feature for the buyer, however, such structured notes place a cap on the potential losses to the investor. For example, in no case could the principal redemption plus coupon payment be less than zero. This means that the structured note includes an option for the buyer to sell the implied short-dollar-long-peso position to the New York investment house for $0. Alternatively stated, the New York investment house is short a put option on a long-peso-short-dollar asset. Normally, this would be far out of the money and therefore require little delta-hedging. A large enough movement in the exchange rate, however, would require the New York firm to establish suddenly an appropriate delta hedge.

After losing the principal and coupon on the note, there are no further loss implications for the Mexican bank. The New York investment house, however, now has only the long-peso-short-dollar position used to hedge the original note. At this point, the foreign banks start taking losses. In preparation for the suddenly likely exercise by the Mexican bank of its put option, the New York investment house would normally want to delta hedge by shorting the peso, but it was difficult to take a short position in the peso during the crisis itself. Market participants argue that a close substitute was then to short “Mexico-like” currencies, such as those of Argentina, Brazil, or Venezuela. Shorting other currencies that would behave similarly to the peso would provide a cover, though there would still be basis risk. Such short selling to cover structured notes on the Mexican peso provides some linkage for the transfer of pressure on the Mexican peso to the other currencies.

In the case of Chilean equity, market sources report that offshore equity swaps are used regularly to permit trading in Chilean equity. They also report serious, though as yet unsuccessful, financial engineering research efforts to crack directly the Chilean tax on capital imports.

To the extent that there is differential income tax treatment for different types of income—capital gains, dividend, and interest—derivatives may lose some effectiveness circumventing the transactions taxes.

Although the balance of payments capital accounts are set up to measure cross-border changes in legal ownership of claims to assets and liabilities, the classification system for financial items is designed to bring out the motivation of creditors and debtors. See International Monetary Fund’s Balance of Payments Yearbook, Part I (Washington, 1994), p. xxii.

The usual problems concern omissions or miscategorizations of transactions. That these have been magnified in the presence of widespread use of derivatives has been duly recognized by authorities responsible for technical standards, as exemplified by the April 1996 meeting at the IMF of the Informal Group on Financial Derivatives. Nevertheless, technical discussions even now center on how to fit derivative-generated payments into standard categories such as interest versus capital gains, the treatment of margin flows, and how to book repurchase agreements. The undermining of the meaning of the various asset categories of the capital accounts in the presence of unrecorded derivative products is not an issue under discussion.

An exception arises it a deposit of margin is required by a foreign counterparty; the margin will be counted as a capital export.

    Other Resources Citing This Publication