The Dangers of Derivatives

Henry C K Liu

This article appeared in Asia Times Online on May 23, 2002

Recession in advanced economies, induced by the oil shock of 1973, pushed transnational banks to find borrowers in developing economies to accommodate petro-dollar recycling. That marked the beginning of finance globalization which, among other trends, replaced foreign aid with foreign loans to developing countries. In the beginning, the petro-dollar recycling was merely to compensate the developing nations for the sudden rise in oil prices.

Later, the surplus oil money not absorbed by Western markets was pushed on beguiled Third World governments as petro-dollar loans for development, leading the developing world into a bottomless abyss of foreign debt. Not only was the anticipated growth in the developing world not realized by foreign-debt-driven exports, debt repayment became increasingly punitive on the domestic economies as lender nations adopted anti-inflationary measures by the end of the 1970s.

Negotiations between borrowing countries and major international bank creditors were intermediated by International Monetary Fund (IMF) endorsement of structural adjustment (austerity) programs in borrowing countries that spelled reduced government social spending, currency devaluation and export promotion policies that distorted and reversed domestic development. Domestic austerity became the ticket to new foreign loans for servicing old foreign loans, and the servicing of the new loans in turn required more domestic austerity, driving Third World economies toward a downward spiral of accelerating contraction and deeper foreign indebtedness. But the oppressive pressure from the IMF in the 1980s was not anywhere near as severe as that after the financial crises of the 1990s.

The financial crises faced by newly industrialized economies (NIEs) in the 1990s were significantly different from the foreign debt crises in the developing countries in the previous decade. Different forms of foreign funds flowed to different recipients in developing countries during the two periods. More importantly, derivatives emerged as an integral part of funds flow in the 1990s.

Derivatives played an unprecedented key role in the Asian financial crisis of 1997, alongside the growth of fund flows to Asian NIEs, as part of financial globalization in unregulated global foreign exchange, capital and debt markets. Derivatives facilitate the growth in private fund flows by unbundling the risks associated with financial vehicles, such as bank loans, stocks, bonds and direct physical investment, and reallocating the risks more efficiently by expanding the distribution and the level of aggregate risk. They also facilitate efforts by many financial entities to raise their risk-to-capital ratios to dodge regulatory safeguards, manipulate accounting rules and evade taxation. Foreign exchange forwards and swaps are used to hedge against floating exchange rates as well as to speculate on fixed exchange rate vulnerability, while total return swaps (TRS) are used to capture "carry trade" profit from interest rate differential between pegged currencies.

Structured notes, also known as hybrid instruments, which are the combination of a credit market instrument, such as a bond or note, with a derivative such as an option or futures-like contract, are used to circumvent accounting rules and prudential regulations in order to offer investors higher, though riskier, returns. Viewed at the macroeconomic level, derivatives first make the economy more susceptible to financial crisis and then quicken and deepen the downturn once the crisis begins. Since investors can only be seduced to higher risk by raising the return on higher risk, the quest for high return raises the aggregate risk in the financial system. But investors always demand a profit above their risk exposure which will leave some residual risk unfunded in the financial system. It is in fact a socialization of unfunded risk with a privatization of the incremental commensurate returns.

The private global fund flows that led up to the crises of the 1980s were largely in the form of dollar denominated, variable interest rate, syndicated commercial bank loans to sovereign borrowers, recycling petro-dollar deposits from OPEC trade surpluses. The formation of syndicates to underwrite these loans helped to bind lenders together, and along with cross-default clauses in the loan contracts, it greatly reduced individual banks' credit risks by passing such risk to the banking system. Loan syndication amounted to a lender monopoly with open price-fixing between previously competing banks.

In order to reduce the banks' collective exposure to market risk, these loans were issued as adjustable interest rate loans (usually priced as a spread above LIBOR or some short-term interest rate that reflected banks' funding costs), and they were denominated mostly in dollars and otherwise in other G-5 currencies (which reflected the currency denomination of the lending banks' funding sources). Foreign fund flows in this form shifted most of the market risk to the borrowers who might not have fully understood that they bore both foreign exchange risk as well as interest rate risk, and the spiraling exacerbating effect of the two risks on each other, ie, rising dollar interest rates would devalue non-dollar local currencies which in turn would push up local interest rates.

Lending banks in the advanced financial markets of the 1980s, whose liabilities were mostly short-term and denominated in the same currencies as their loans to developing countries, bore little market risk. Their exposure was almost entirely credit risk, and this was substantially mitigated through the syndication of the loans and the inclusion of cross default clauses. Thus a supposedly "free" debt market transformed itself into a bilateral market between powerless individual borrowers and an all-powerful lending monopoly. It was the height of hypocrisy that in a era of blatant financial monopoly that neo-liberal finance market fundamentalism achieved its unprecedented intellectual ascendancy.

The change in the distribution of market risk to Third World sovereign borrowers laid the foundation for the crisis that began in August of 1982. The crisis began when the Federal Reserve raised short-term dollar interest rates to fight US run-away stagflation. Higher dollar interest rates, which served as the basis for payments on adjustable rate loans, both increased the dollar payments on loans and increased the cost in non-dollar local currencies for dollars. Debtor countries were forced to drastically increase their foreign borrowing in order to reduce the burden of servicing suddenly higher foreign debt costs, leading to inevitable crisis.

In August of 1982, the Mexican government announced its inability to make scheduled foreign loan payments. In response, the developing economy governments, major money center banks, and the IMF and World Bank began searching desperately for post-crisis recovery policies, initially by rescheduling existing debt, arranging new lending and requiring the developing economy governments to implement austere fiscal and monetary policies to make possible the eventual repayment of the continuously growing debt burden, but in effect foreclosing developing economies any prospect of growth with which to repay the still mounting foreign loans. The foreign creditors were protected; the debtor developing nations lost what little they had gained in the previous decade and then some, with no prospect of ever escaping from the tyranny of foreign debt in the foreseeable future. Neo-liberal economists cited Shakespeare: Better to have loved and lost, then not to have loved at all, while their paying clients laughed all the way to the bank.

The Asian financial crises that began in 1997 were a very different phenomena. They were caused by hot money (short-term foreign credit based on over-valued exchange rates that were defended beyond reason by Third World monetary authorities poisoned by neo-liberal advice). Derivatives trading in over-the-counter (OTC) markets were, and still are, neither registered nor systematically reported to the market. Thus the full risk exposure in the system is not known until the crisis hits. In macroeconomic terms, derivatives have the structural effect of privatizing the incremental efficiency (profits) while socializing the risk (losses) associated with such profits. This is done by extracting value through rising risk/return ratios by siphoning off part of the incremental return to known private parties while passing on the full incremental risk to faceless third parties spread throughout the system. Since the spread was minuscule, profit incentive pushed for massive increases in volume.

The foreign private fund flows that preceded the 1997 financial crises went to private entities in Asia and not just to sovereign borrowers as in the 1980s. Commercial bank loans in the 1990s, measured as a percentage of total foreign fund flows, were substantially less important than they were in the 1980s. Instead, fund flows to Asia ranged from foreign direct investment (FDI) to portfolio equity investment (meaning less than 10 percent ownership), corporate and sovereign bonds as well as structured notes, repurchase agreements, on top of traditional bank loans to public and private borrowers. This more diversified flow of funds generated a different distribution of risks towards global institution investors, mainly pension funds in the advanced capital and debt markets. Stocks and bonds investments in Asia shifted market risk and credit risk to foreign institution investors who bore the risk of changes in interest rates, securities prices and exchange rates.

FDI in physical capital and real estate similarly shifted market risk and credit risk to foreign institution investors. Even dollar denominated bonds issued by Asian governments shifted interest rate risk, as well as credit risk, to foreign institution investors. Socialized finance in the rich economies, what the Wall Street Journal fondly referred to as mass capitalism, was called on to finance old-fashioned compradore capitalism in the NIEs. The effect was to expose the NIEs to the risk of changes in US interest rates or the exchange value of the dollar, not as economic fundamentals, but as technical trends perceived by the herd instincts of fund managers in the advanced markets. Thus the neo-liberal focus on the need to resolve the national banks' domestic non-performing loans (NPL) as a prerequisite for generating growth is, to say the least, misplaced. The NPL problem in Asia is a fiction invented by the Bank of International Settlement (BIS) to prepare national private banks as ripe targets for predatory acquisition by Western large, complex banking organizations (LCBO).

Derivatives trading grew up alongside new forms of capital flows as part of an effort to better manage the risks of global investing. Derivatives facilitate new composition of capital flows by unbundling risk and redistributing it in commensurate return/risk ratios to a broad market. At the same time, derivatives create new systemic risks that are potentially destabilizing for both developing and advanced economies. While the risk shifting function of derivatives initially served the useful role of hedging and thereby facilitating fund flows, the prevalence of derivatives is now threatening the stability of the global economy as a whole.

Derivatives are unlike securities and other assets because no principal or title is exchanged. In their essence, nothing is owned but pure price exposure based on "notional" values. They are merely pricing contracts. Their price is derived from an underlying commodity, asset, rate, index or event, and this malleability allows them to be used to create leverage and to change the appearance of transactions. Derivatives can be used to restructure transactions so that positions can be moved off balance sheet, floating rates can be changed into fixed rates (and vice versa), currency denominations can be changed, interest or dividend income can become capital gains (and vice versa), liability can be turned into assets or revenue, payments can be moved into different periods in order to manipulate tax liabilities and earnings reports, and high yield securities can be made to look like convention AAA investments. The accounting of IRS (Indefeasible Right of Use) of communication network capacity, particularly "dark trades" of unused fiber-optics capacity widely employed in the telecommunication sector, has inflated revenue for the entire sector, by booking future revenue as current income and related future liabilities as off-balance-sheet future capital expenditure. Forwards and foreign exchange swaps are not always highly collateralized (measured as a percentage of the principal). Collateral is less likely to be used for trading between the major market dealers or participants, and collateral is lower for less volatile financial instruments such as currency. This exposes foreign exchange derivatives counterparts to credit risk.

The largest source of credit risk losses in the derivatives markets in recent years, up to the recent fiasco in telecom and energy trades, had been due to defaults on foreign currency forwards in Asia and Russia. Foreign exchange forwards and swaps were used in Asia by both foreign and domestic investors to hedge foreign exchange risk. Foreign investors from advanced capital markets who purchased securities denominated in local Asian currencies could use foreign exchange forwards and swaps to hedge their long local currency exposure. Similarly, FDI in real estate, physical plants or equipment were exposed to the risk of local currency depreciation. Asian investors who borrowed in dollars, yen or European currencies and invested in local currency assets were also exposed to foreign exchange risk. Foreign exchange forwards and swaps were also used for speculation in Asian currencies. Derivatives enabled speculators to leverage their capital in order to take larger positions in the value of local currencies. It also meant that Asian central banks had to watch the exchange rate in two markets, the spot and forward, in order to maintain their fixed exchange rates.

The term swap is sometimes used to refer to OTC derivatives in general. This no doubt arose from the fact that the vast majority of OTC derivatives are swaps of one form or another. A TRS is a contract in which at least one series of payments is based on the total rate of return (change in market price plus interest or dividend payments) on some underlying asset, security or security index. The other leg of the swap is typically based on a variable interest rate such as LIBOR, but may be a fixed rate or the total rate of return on some other financial instrument. Pre-crisis data on Asia show that TRSs in those situations usually swapped LIBOR against the TRS on a government security. A TRS replicates the position of borrowing at LIBOR in order to finance the holding a security or security index. The returns are the same, but unlike the actual cash market transaction, it does not involve ownership or debt. Instead the only capital involved in a TRS is the posting of collateral. In addition to the reduction in the need to commit capital to the transaction, a TRS also has no impact on a firm's balance sheet and not subject to regulatory restrictions on foreign exchange exposure.

It would incur a capital charge only if it were to move into the money. In short, TRS allow financial institutions and investors to raise their risks, and potential returns, relative to capital. One of the uses of TRS in Asia was to capture the gains from the carry trade or carry business. A profitable carry trade exists where exchange rates are fixed and interest rate differentials persist between the two economies. Then it is possible to borrow in the low interest rate currency and lend in the high interest rate currency with no risk other than that of a failure in the fixed exchange rate. Yet when large numbers of market participants catch onto the game, the fixed exchange rate cannot hold. When a central bank defends fixed exchange rate under these conditions, it is essentially giving money free to all comers.

The foreign money center banks were eager to lend the G5 currencies and the Asian local banks were eager to capture carry profits from the interest rate differentials. The profitability and ultimately the solvency of the financial institutions conducting this carry business depended on the Asian central banks' ability to maintain the exchange rate peg or at least preventing it from depreciating by more than the interest rate differential. But the combination of foreign borrowing and domestic lending created a foreign exchange mismatch on balance sheets. Financial market regulations either limit the extent of this mismatch or they require additional capital in proportion to the exchange rate exposure. Foreign banks also had country limits and lending limits to any one Asian financial institution, which created disincentives for borrowers to swell their balance sheets (which lowered the reported earnings on assets). In this context, TRS were tailor-made for financial institutions wishing to avoid prudential regulations by taking their carry positions off balance sheet.

The use of TRS altered the form of fund flows to developing countries. If banks engaged in the carry trade by borrowing abroad, then the capital flows were in the form of short-term, hard currency bank loans. If banks pursued the same profit opportunities by using TRS, then it would generate indirect capital flows as swaps counter parties, usually advanced capital market swaps dealers, bought the underlying asset as a hedge against their own position in the TRS. As a result, the fund flow was in the form of a local currency denominated security instead of a dollar bank loan. However, the local currency security did not have the effect of shifting foreign exchange risk to advanced capital market investors. Instead, it functioned in conjunction with the TRS to leave the local developing country investors holding the foreign exchange risk (the short dollar position) much like a hard currency bank loan. Not only is the developing economy's foreign exchange exposure the same, but the TRS carry strategy exposes it to even greater surges in foreign currency than with short-term bank loans. The bank loans, except when there are attached puts, are certain for the maturity of the loan. However the collateral arrangements on the TRS can result in a large immediate surge, overnight if not intraday, in foreign currency transfers. If short-term bank loans are considered hot money, then payments to meet margin and collateral requirements are microwave money.

The use of TRS also increase the likelihood of contagion. They often involve cross-currency assets and payments and are therefore more likely to transfer disruptions from one market to another. One reason that Korean banks engaged in so many Indonesian TRS was that they were seeking higher rates of return in response to a rise in their funding costs. At the end of this process, Korean banks were exposed to Indonesian credit risk. This however, is not visible on their balance sheets. This situation not only creates the possibility for contagion, but may also make the contagion unpredictable and severe.

The same situation applied to Brazilian investors in Korean TRS. The Korean debt and currency crisis in 1998 was solved by the US Treasury and the State Department requesting the Federal Reserve to allow the US banks to roll over pending Korean non-performing loans (NPLs) into longer term performing loans without required writeoff provisions and capital adjustments. The New York Times reported that the administrative decision was reached by Treasury Secretary Robert Rubin in the last minute before a Korean default because of the surprised discovery that Brazilian banks were holding a lot of Korean bonds and TRS contracts. A Korean default would quickly spread to South America with more direct impact on the US economy than previously realized. International banks and regulators in lender countries colluded on the classification of non-performing loans in Korea for regulatory purposes. But the local banks in Korea enjoyed no such flexibility.

Hybrid instruments include such conventional securities as convertible stock options, convertible bonds and callable bonds. These have long been among the set of traditional securities regularly issued and traded in US financial markets. The structured notes used in Asia were usually structured so that their yield was linked to the value of one or more of the currencies or stock indices in the developing economies. The issuers of these structured notes were financial institutions from advanced capital markets and the investors were often Asian financial institutions and investors who are more willing to hold their own exchange rate risk or that of their neighboring developing countries. One reason given for this is that they are more knowledgeable of their economies and markets than investors from advanced capital markets. An additional reason is that they are inherently long in their own currency in the sense that if the fixed exchange rate regime were to collapse, then the whole economy would contract and therefore the perceived cost of additional foreign exchange exposure is small in the overall scheme. When the whole house is on fire, who cares if the kitchen is uncomfortably warm.

Financial institutions from advanced capital markets were interested in issuing these instruments in order to create long-dated futures and options positions in developing country currencies and securities. Foreign exchange forward and swap markets are short-term markets; the vast majority of transactions in these markets have a maturity of one year or less, the majority are for seven days or less. Faced with the absence of other alternatives, these hybrid instruments were designed to create such a multiyear foreign exchange derivative. The issuer held a long-term short position in a developing country currency for the cost of the 100 or 200 basis points per year on the principal. The most widely known of these derivatives is called a PERL - principal exchange rate linked note. These instruments were denominated in dollars, but the value of their payments were linked to a long position in the value of a developing country currency. The compensation or premium for holding this exchange rate exposure was a higher than normal yield in comparison to a similarly rated dollar denominated note. If the foreign currency exchange rate remained fixed, or did not decline too far in value, then the higher yield would be realized. A devaluation or a substantial depreciation, however, could cause the return of the note to fall below the norm and in the event of a major depreciation the structured note might realize a negative return.

Financial institutions in developing countries were interested in buying these hard currency denominated assets because they needed hard currency assets to match their hard currency liabilities on their balance sheets. While it has been noted that derivatives have generally facilitated fund flows to developing economies, in this case they encouraged the reverse. When a developing economy investor or financial institution purchases a high yielding structured note such as the above example of a PERL, then the fund flow is reversed and principal is transferred to the advanced capital markets.

Derivatives lead to transparency problems in two basic ways. One, they distort the meaning of balance sheets as the basis for measuring the risk profile of firms, central banks and nation accounts. Two, when traded over-the-counter, derivatives lack adequate reporting requirements and government surveillance. The proposed BIS Basel II Accord aims to address these problems.

Accounting rules are used to calculate profits and loses, designate assets and liabilities, and determine tax liabilities and capital requirements. A survey of US businesses before the Enron fiasco revealed that 42 percent used derivatives primarily to "manage reported earnings" by moving income from one period to another. In this case, the lack of transparency results in distorted market information.

The lack of transparency caused by off-balance sheet positions is also a problem for the market participants in their efforts to assess a central banks' ability to intervene in the foreign exchange market. The ability to intervene is critical in the context of a fixed exchange rate regime, but it is also important in the context of a floating rate system in order to stabilize the economy following a speculative attack or other financial market disruption. The problem arises when a central bank accurately reports the value of its foreign reserves, but does not report the amount they have contracted to sell in the future through foreign exchange forward and swaps contracts. This delinking of total risk exposure from balance sheets also occurs in regards to a nation's balance sheet, ie, their balance of payments accounts.

A country's actual exposure to market risk was once reflected in the maturity and currency denomination of its foreign assets and liabilities in its capital account. That is no longer the case. The currency denomination of assets and liabilities such as foreign loans can be changed with foreign exchange derivatives. Interest rate swaps can alter the interest rate exposure on assets and liabilities. Long-term loans can become short-term ones if attached "put" options are exercised. Even the form of capital or the investment vehicle can be transformed with derivatives. Total return swaps can make short-term dollar loans (liabilities) appear as portfolio investments. Also, the requirement to meet margin or collateral calls on derivatives may generate sudden, large foreign exchange flows that would not be indicated by the amount of foreign debt and securities in a nation's balance of payments accounts. As a result, the balance of payments accounts no longer serve as well to assess country risk.

A main challenge facing the IMF due to the spread of derivatives is how to restructure the balance of payments accounting systems of its major member countries. Cross-country derivatives positions have played havoc with the balance of payments data. One internal IMF estimate has off-balance positions potentially warping emerging market economic data by as much as 25 percent.

Derivatives also create transparency problems in other ways. The lack of reporting and government surveillance limits the government's and market participants' ability to assess the amount of open interest in the market, large positions held by single entities and the adequacy of collateral and capital. This prevents a dependable assessment of the stability of these markets as well as the markets to which they are linked. In these ways the presence of derivatives can make it difficult for firms to make an accurate assessment of their counterparts' credit worthiness. Similarly, the lack of information and data on OTC derivatives means that regulatory authorities cannot detect and deter manipulation in the immediate or related markets. In addition, the regulatory authorities cannot know outstanding positions - whether measured gross or net - of their financial sectors or major participants in the financial sector. Thus they cannot know how much risk their financial markets are exposed to in comparison to the capital on hand. As a result, it is difficult for government regulators or supervisors to track the sensitivity of the economy to changes in certain key market variables such as interest rates and exchange rates. The increased use of derivatives in developing economies is increasingly making the full disclosure of relevant information, or at least the full interpretation of the disclosed information, even more difficult.

Accounting rules distinguish different types of assets and receivable by their credit worthiness or credit rating. These rules assign capital charges according to their credit risk, and in some cases government regulations prohibit financial institutions from holding certain classes of assets. For example, structured notes are sometimes designed to manipulate accounting rules so that high yield notes can be treated like top rated credit instruments for the purpose of assigning capital charges. The result is that foreign exchange exposure is not treated as such for regulatory purposes. These types of structured notes have also been used to outflank US regulations that prohibit institutional investors such as pension funds and insurance companies from holding foreign currency assets.

Some tax provisions are designed to enhance regulatory safeguards by raising the relative costs of certain financial activities deemed to be less productive. For instance, long-term capital gains may be taxed at a lower rate than short-term gains in order to raise the reward on long-term investing. Using derivatives, payments, receipts and income can be shifted from one period to another. Transactions can be restructured so that they appear to occur as capital gains instead of interest payments (or vices versa), or as long-term capital gains instead of short term ones. This problem is doubly important for developing economies whose tax bases are not well established, and where threats to their tax base can put fiscal pressures on the government that can lead to monetary expansion or greater foreign borrowing. The use of derivatives to circumvent or outflank prudential regulation has been acknowledged by the IMF, World Bank and the OECD, among others.

Another problem is the tendency for derivatives to be used to raise the level of risk relative to capital. This can occur even within the regulatory structure. Derivatives are designed to create price exposure so that risk can be transferred from one party to another without the expense of transferring title and principal as required to buy or short-sell. If the initial payment is thought of as equity or capital, then the size of the notional principal in comparison to capital is the degree of leverage in the derivative instrument. This leverage allows investors to assume far greater degrees of risk per dollar of capital than would be available by purchasing the asset outright or even borrowing in order to purchase the asset outright.

Leverage is a double-edged sword. Leverage enables derivatives to offer a more efficient use of capital for hedging or investing, and at the same time it reduces the amount of capital backing a given amount of price exposure (ie, the size of a position). Raising the risk-capital ratio weakens the stability of each investor, and in turn it increases everyone else's exposure to the repercussions from investor failure. In short, it increases system risk by socializing individual risk while privatizing the rewards.

The presence of a market for foreign exchange derivatives can undermine the stability of a fixed exchange rate system in several ways. Derivatives provide greater leverage (lower capital costs) and lower transactions costs for investors taking a position against the success of the fixed exchange rate. Such investors are often referred to as speculators, attackers or hedge fund operators. Lowering the costs of betting against the fixed exchange rate will only encourage this behavior and strengthen the effects of those efforts. The greater the volume of positions that are short the currency, ie, against the fixed rate regime, the greater the necessary size of central bank intervention or interest rate hikes needed to defend the currency peg.

The presence of foreign exchange markets means that the central bank is faced with the greater task of having to peg the exchange rate in two markets: the spot markets; and the forward or swap market for foreign currency. Whereas the spot market is generally large in relation to the amount of foreign reserves at the central bank, and thus the central bank's potential for intervention is small in regards to the overall size of the market, the size of the derivatives market is unlimited. Together they increase the critical size for a successful central bank intervention.

Another problem posed by the presence of foreign exchange derivatives markets is that price discovery process in those markets will, under many circumstances, indicate a future devaluation. There are two reasons for this. First, interest rates in developing countries are in most circumstances higher than in advanced capital markets or developed economies. This interest rate differential means that the equilibrium forward or swap rate will always be higher than the spot rate - thus indicating that the currency will depreciate at the rate as the interest rate differential. Second, if the credit market in the developing country is not perfectly efficient, then foreign exchange market makers will not provide forward and swap contracts at rates that do not include a market risk premium. If a market risk premium is added to the interest rate differential, then the forward and swap rates will indicate a greater rate of depreciation.

In the event of a devaluation or a sharp downturn in securities prices, derivatives such as foreign exchange forwards and swaps and TRS functioned to quicken the pace and deepen the impact of the crisis.

Derivatives transactions with emerging market financial institutions generally involve strict collateral or margin requirements. Asian firms swapping the TRS on a local security against LIBOR were posting US dollars or Treasury securities as collateral; the rate of collateralization was estimated at around 20 percent of the national principal of the swap.

If the market value of the swap position were to decline, then the East Asian firm would have to add to its collateral in order to bring it up the required maintenance level. Thus a sharp fall in the price of the underlying security, such as would occur at the beginning of a devaluation or broader financial crisis, would require the Asian firm to immediately add dollar assets to their collateral in proportion to the loss in the present value of their swap position. This would trigger an immediate outflow of foreign currency reserves as local currency and other assets were exchanged into dollars in order to meet their collateral requirements. This would not only quicken the pace of the crisis, it would also deepen the impact of the crisis by putting further downward pressure on the exchange rate and asset prices thus increasing the losses to the financial sector.

The BIS "Lamfalussy Report" defined systemic risk as "the risk that the illiquidity or failure of one institution, and its resulting inability to meet its obligations when due, will lead to the illiquidity or failure of other institutions". Similarly, contagion is the term established in the wake of the Asian financial crisis of 1997 to describe the tendency of a financial crisis in one country to adversely affect the financial markets in other, and sometimes seemingly unrelated, economies. It is the notion of systemic risk taken to the level of national and international markets.

The presence of a large volume of derivatives transactions in an economy creates the possibility of a rapid expansion of counterparty credit risk during periods of economic stress. These credit risks might then become actual delinquent counterparty debts and obligations during an economic crisis. The implication is that even if derivatives are used to reduce exposure to market risk, they might still lead to an increase in credit risk. For example, a bank lending through variable rate loans might decide to reduce its exposure to short-term interest rate variability, thus the volatility of its income, by entering into an interest rate swap as the variable rate receiver. If short-term rates were to rise, then the fair market value of the bank's swap position would rise, and thus would increase the bank's gross counterparty credit exposure above that already associated with the loans which were being hedged.

In so far that derivatives increase counterparty credit exposure throughout the economy, they increase the impact of one entity becoming unable to fulfill its obligations. And to the extent that derivatives are not used to reduce firms' exposure in the market, then the greater leverage brought to speculative investments increases the likelihood of such a failure. In this way, derivatives contribute to the level of systemic risk in the financial system.

The presence of derivatives can also increase the global financial system's exposure to contagion by the international nature of markets. Many derivatives involve cross-bred counterparts and thus losses of market value, and credit rating in one country will affect counterparts in other countries. The second channel of contagion comes from the practice of financial institutions responding to a downturn in one market by selling in another. This demand for collateral assets can be sudden and sizable when there are large swings in financial markets, and thus this source of contagion can be especially fast and strong.

The process of policy formation was much more straightforward in the wake of the 1980s' debt crisis. The borrowers were mostly governments, and the private lenders were the large money center banks. This meant a single representative borrower for each debtor country was therefore represented by a single borrower, and the key lenders could be gathered into a single room. Together with the relevant multilateral institutions, all the parties could negotiate a plan to restructure debt payments.

The policy making process became much more complicated in the 1990s. There were many different private and public debtors and issuers of securities. There were many investors and many different types of claims on parties in the affected developing countries. Capital flows in the form of stocks, bonds and structured notes meant that there were hundreds of major investors and millions of lesser investors. These claims were all the more complicated because of derivative contracts. Derivatives added to both the number of potential counterparts and raised problems as to who held the first claim on outstanding debts and other obligations.

Moreover, losses on derivatives are not the same as late payments on loans. Debt payment problems do not necessarily have to result in loses to either side. Some loan payment problems are short-term liquidity problems that can be solved by merely restructuring the loan payment schedule. Derivatives losses, in contrast, are already lost and cannot be mitigated. These loses must be paid immediately, although the payments can feasibly be financed by acquiring additional debt. What is more, changes in market interest rates and exchange rates can cause derivatives losses to occur more suddenly, accumulate more quickly and sum to greater magnitudes than the losses associated with dollar denominated variable rate bank loans.

In Turkey, one of the midsize banks, Demirbank, had been taking business from Turkey's big established banks, betting big on the government's anti-inflation program to bring interest rates down still further. At one point, Demirbank held 10 percent of Turkey's domestic debt. But it was funding its operations from Turkey's short-term money markets, which were supplied by the same big-money banks it had alienated. These Turkish money center banks in turn got their funds from the short term US repo market, profiting from the interest rate spread of the two currencies. When delays hit a big Demirbank foreign loan syndication in December 2000, the bank suddenly found its lines of credit cut by the money center banks. Demirbank was forced to dump its Turkish treasury bills at a loss to meet margin calls and other funding obligations. A classic fire sale began, not much different than the situation faced by Long Term Capital Management (LTCM) in the US a year earlier.

Normally, the Turkish central bank would have stepped in to ease Demirbank over its liquidity crisis, as the NY Fed did with LTCM, by easing up on regulatory rules. But a key condition of IMF support for the anti-inflation program was a cap on the total foreign and local currency in circulation in Turkey. So, when the Demirbank crisis triggered a small rush to buy dollars from the central bank, it drained Turkish lira out of circulation just when it was most needed to ease lending between banks. Already spooked by trouble brewing in Argentina, the poster boy of neo-liberal monetarist economics, investors stampeded out of Turkey on November 22, 2000, thanklessly just before the long US Thanksgiving holiday weekend. As yields on Turkish domestic assets slid from 35 percent to 4.5 percent in dollar terms, many investors started selling Turkish treasury bills to cut their losses. They sought safety in dollars, sucking the central bank's currency reserves down further. Deutsche Bank alone sold US$700 million worth of Turkish treasury bills in a day, mostly on behalf of clients (read hedge funds) but also for its own proprietary trading accounts. Briefly breaking with the IMF plan, the Turkish central bank supplied local currency to the banks. But it was too late. Turkish markets stalled and plunged into a panicky tailspin. Everybody lost, including clients of international banks, some of whom (hedge funds) managed to pass the pain onto the financial system, but the international banks themselves continued to collect handsome fees.

While derivatives threaten the global economy, their effect on emerging economies is catastrophic.