Since 2013, regulators and government agencies in several countries have investigated the allegation that large financial institutions manipulated the $5.3 trillion-a-day foreign exchange (FX) market. Separate from these investigations, an antitrust lawsuit was filed against 12 major financial institutions for allegedly rigging prices in the FX market. The conduct at issue involves dealers colluding to move spot FX rates and affect the determination of key FX benchmarks based on these rates. Currency dealers allegedly took advantage of their knowledge of pending client orders, which referenced certain future benchmark rates, and traded to their own benefit (“front-running” of client orders). In an effort to move FX rates to their advantage, dealers allegedly colluded to concentrate trades just before benchmark rates were set with the intent of manipulating spot FX benchmarks.
The FX market for trading currencies operates continuously and is concentrated in major financial centers around the world. No single exchange rate exists between any two currencies: there is a bid and an offer price, and these may vary across banks and clients and over time. Transacting clients typically rely on a large financial intermediary to find the best execution for their FX trades, but such executions may not be fully transparent. A practical alternative is to rely on published spot benchmark foreign exchange rates (known as “fixes”), which reflect market rates at certain times. The WM/Reuters benchmark rates, in particular, are computed (“fixed”) for 160 currencies, and are set every half hour for the most heavily traded currencies.
On November 12, 2014, five major banks entered into settlement agreements with U.S., U.K., and Swiss financial regulators over foreign exchange rate manipulation and agreed to pay $3.4 billion in fines. The charges involved “attempted manipulation of, and. . . aiding and abetting other banks’ attempts to manipulate, global foreign exchange (FX) benchmark rates to benefit the positions of certain traders,” more specifically claiming that “certain FX traders at the Banks coordinated their trading with traders at other banks in their attempts to manipulate the FX benchmark rates.”
The alleged misconduct requires that dealers be able to affect FX rates, which in turn requires that a single dealer or a group of dealers account for a sufficiently large share of the market. In addition, according to the investigating agencies, collusion was facilitated by participation in online chat rooms. Through these communications, traders shared information about clients’ orders and coordinated their trades to affect exchange rates.
Assuming that the necessary conditions for collusion are in place, rate manipulation may be implemented as follows. Consider a client that transacts with a bank at 3:30 pm to sell British pounds and buy $100 million at the 4:00 pm fix rate. Since the fix is uncertain at the time of the agreement, the bank’s purchase of dollars to deliver to its client after the fix implies that the bank has assumed a risk associated with rate movements between 3:30 pm and 4:00 pm. To manage this risk, the bank may pair off with a counterparty holding a selling interest at the fix and thereby eliminate its exposure.
The bank, alternatively, may assume the exchange rate risk by completing a proprietary transaction. For example, say the bank purchases the $100 million at 3:30 pm for 66 million pounds ($1 for 0.66p). The bank profits if the average price at which the bank buys the currency in the market (dollars in this example) is lower than the 4:00 pm fix rate at which it has agreed to exchange currency with its client. These profits or others resulting from an effort to find the best execution may be associated with legitimate risk management activities.
Profits may, however, also result from market manipulation. The bank may effectively eliminate (or significantly reduce) the exchange rate risk it assumes between 3:30 pm and 4:00 pm if it colludes with traders at other institutions and succeeds in favorably moving the price of the currency at the 4:00 pm fix. For example, if the colluding institutions account for a sufficiently large volume of the transactions around the time the fix is determined, and they concentrate large transactions at this time (“banging the close”), they may collectively drive up the value of the dollar relative to the pound. Benchmark rates, in turn, would be driven up because they would be based on the higher rates for these transactions.
A successful effort to affect the price of a currency when a benchmark rate is set will result in a benchmark that is different from the value absent the manipulation. A party that transacts at the benchmark rate suffers economic harm if the price of the currency it buys (sells) is higher (lower) than the price it would have paid had the currency manipulation not occurred.
When benchmark rates are manipulated, measuring damages requires determining the benchmark rate that would have been in effect absent the manipulation. Benchmark rates may be recalculated using an alternative sample of untainted transactions that provide a reliable “but for” measure of the exchange rates in the market. To isolate the effect of the price manipulation, economists typically rely on economic models that account for the collusive behavior and other factors driving prices. Such models often are based on the identification of periods that are not affected by collusion. Singular changes in exchange rates during the fix period can be identified relative to periods with no collusion. Alternatively, a court may order disgorgement, in which case damages may be measured by the benefits that colluding banks derive from manipulating the market.
Ultimately, qualifying the economic harm also will depend on how rate manipulation affects a specific transaction. The most immediately injured parties from rate manipulation are the direct victims of front running. Third parties also may be harmed by the artificial concentration of orders around the time of the fix. In the absence of significant arbitrage opportunities, trades executed at times when market prices are manipulated are likely to affect the exchange rates at which other market participants trade. If the rate these parties pay is increased, they are harmed in the amount of the overpayment – the amount traded times the difference between the market exchange rate and the one that would have prevailed had market orders not been artificially concentrated.
As with the manipulation of LIBOR rates, the alleged conduct in FX markets involves benchmark rate manipulation, which extends economic harm well beyond the direct client submitting the initial FX order. The impact of FX manipulation, for example, extends to derivative instruments, such as options, futures, and swaps. These contractual agreements may define their value, in whole or in part, by reference to an FX benchmark. If the value of that benchmark is manipulated, then the value of the derivative instrument will be affected to the detriment of one of the contracting parties. Finally, FX manipulation also can exert broader economic harm by distorting key market prices. Such costs may be difficult to measure, but they are distinct economic harms.