Role of Institutional Investors
Global Hedge Funds
As noted earlier, hedge funds—private, closed-end investment funds—assemble pools of private capital to engage in highly leveraged position-taking. Their moves last summer to position themselves favorably for possible exchange rate realignments in the ERM apparently served as a signal for other institutional funds managers to re-examine their own (convergence play) positions. Thus, although hedge funds have less than $10 billion in capital, their potential influence on forex markets is larger.
The hedge funds appear to have taken two types of positions during the crisis. First, with respect to foreign exchange, they typically sold forward the vulnerable currency, usually using longer-dated contracts (one to two months) and employing their capital to finance and to maintain margin requirements. Second, they also established interest rate positions in that currency (for example, by buying interest rate futures on the London and Paris futures exchanges) in anticipation of a rate decline after the devaluation. Intrinsic to this strategy was the expectation that authorities in the weak-currency country would raise interest rates to defend the parity—but also that this period of high interest rates would be short lived. To close out their short currency positions, the hedge funds bought the devalued currency in the spot market before the forward contracts expired.
Institutional Investors and Corporations
While the hedge funds acted as market leaders, the real financial muscle was provided by institutional investors (mutual funds, pension funds, insurance companies) and by nonfinancial corporations. These market participants sought to undo as quickly as possible their increased holdings of assets denominated in currencies with high interest rates and/or to cover their deutsche mark funding. The mandates of most mutual funds are broad enough to allow a rebalancing of portfolios in anticipation of exchange rate movements. Anecdotal evidence suggests that funds sold their foreign assets, hedged their exposure, and sold the vulnerable currencies short (using their assets as collateral—in roughly equal proportion (although outright selling of assets seems to have predominated in Italy, whereas hedging and short sales were probably relied on more in the United Kingdom).9
In some isolated cases, either the prospectus of the fund or the regulatory constraints on permissible financial operations limited the room for maneuver. For example, if a fund was established as a peseta bond fund, it could not cease to hold peseta-denominated securities even if it anticipated a peseta devaluation.10
In some countries (e.g., the United Kingdom), mutual funds can short a currency only up to the amount that they hold securities denominated in that currency. In other countries, mutual funds are limited in their ability to leverage their investments and, in effect, to short a currency. Corporate and financial sector borrowers from the high-yield ERM countries who borrowed extensively in low-cost ERM currencies undertook massive hedging operations during the summer and early autumn. U.S. corporations appear to have engaged in proxy hedging of their European currency exposures, using deutsche mark and French franc positions to hedge exposures in other European currencies. These corporations sought to unwind these hedges as the crisis unfolded.
Role of the Banking System
The international banking system itself does not appear to have taken unusually large net open positions in foreign exchange during the crisis. “Appear” is used because the information available does not permit determination of the intraday net open positions of banks—nor are the data on foreign exchange payments comprehensive enough to calculate how bank earnings during the crisis were apportioned between market making and other activities. Nevertheless, a robust conclusion is that the main role played by banks (and securities houses) during the crisis was to provide credit to those institutions seeking to liquidate long positions in vulnerable currencies, hedge such long positions, or establish open short positions in these currencies, and to make markets in foreign exchange.11
Both of those activities proved profitable. In the early summer of 1992, immediately after the Danish vote, U.S. and European commercial banks and securities firms—many of which had in their capacity as market makers in foreign exchange witnessed firsthand the flow of funds into the convergence plays—began to negotiate credit lines in the currencies that were perceived to be vulnerable—lira, sterling, peseta, escudo, but not yet the French franc. By building up cash reserves and arranging credit lines (at fixed rates) in the vulnerable currencies before the crisis, they were subsequently able both to lend at a premium above their borrowing costs in the weak currencies and to buy forward the currencies under attack. Several banks have argued that such “liquidity plays” made more sense for banks than outright currency plays because they were less risky and allowed banks to employ their comparative advantage, namely, their access to large credit lines (relative to those available to other types of financial institutions). As the crisis was also characterized both by a large widening of spreads and by extremely high volumes, simple market making likewise paid handsome returns.
If banks play a key role during a currency attack by providing liquidity to other participants, where do the banks, in turn, obtain that liquidity? It is available only if central banks feel constrained in allowing interest rates to rise enough to choke off the increased demand for credit in the currency under attack. And taking the argument one step further, interest rates may not rise much during the attack because central banks typically sterilize the (contractionary) monetary effects of their exchange market intervention.
Liquidity Problems During the Crisis
According to most observers, the forex market-as well as domestic money markets in ERM countries—generally worked well during the crisis. Given the huge volumes of securities and currencies traded during that period—as well as attempts to ration liquidity to those taking positions against existing parities—that outcome was hardly preordained. There were indeed strains. While forex spot markets operated continuously, spreads at times widened from five to ten times the norm in most of the ERM cross-rates. The size of trades also declined at times. For example, the normal size of a lira-deutsche mark trade is DM 50 million, but this size was not available in mid-September. Dealers in lire hesitated to quote forward rates because of the great volatility in short-term interest rates, and this market did not operate for two weeks after the devaluation. OTC option markets suffered, as interest rate volatility many times the norm made quotation of prices hazardous. As credit-line limits were reached, those institutions with lower credit ratings lost their access to interbank markets and had to move to derivatives exchanges to hedge their positions. The greatest liquidity problems actually surfaced in the European currency unit (ECU) market, where for a period in September and October it was not possible to complete wholesale transactions (as market makers would no longer bear the risk of holding an ECU trading portfolio). A significant spread (250 basis points) also developed between the official ECU and the private ECU (ECU-denominated bank deposits at ECU clearing banks) when some central banks tried to use their ECU reserves to support their currencies.12 Trading in the private ECU against the basket came to a halt for about a week immediately after the onset of the crisis. Presumably, the political events that raised uncertainty about the future of EMU also created increased uncertainty about the value of the private ECU in terms of the official basket.