Goldman Sachs (GS) recently agreed to pay a $120 million penalty to settle allegations by the U.S. Commodity Futures Trading Commission (CFTC) that GS’s traders worked to manipulate benchmark rates used in derivatives markets. Specifically, the CFTC alleged that GS manipulated the U.S. Dollar International Swaps and Derivatives Association Fix (“ISDAfix”), a benchmark rate used to value a broad range of financial derivatives, including cash settlement options and interest rate swaps.
Similar to other benchmark rates, such as the London Interbank Offered Rate (LIBOR) and foreign exchange (FX) rates, the ISDAfix is determined based on inputs from market participants. ISDAfix submissions should reflect market-based actual rates offered in inter-dealer trades and executable inter-dealer bids at 11:00 am eastern time. A panel of banks, including GS, could then accept the market rate, submit a different rate, or take no action. (Since the time of the events in the CFTC complaint, the ISDAfix has been restructured and renamed the Intercontinental Exchange (ICE) Swap Rate.)
Allegations in other cases of manipulation of benchmark rates typically concerned concerted or coordinated efforts by a group of traders at multiple banks to manipulate the benchmark. The allegations settled here relate to GS traders’ unilateral efforts to influence the fix in two ways. First, GS allegedly timed bids, offers, and execution of contracts to move the ISDAfix in a way that favored its trading positions. Second, GS allegedly made false ISDAfix submissions that were not driven by its true willingness to bid or offer swaps, but instead by GS’s incentives in relation to its derivative positions. The misleading information allegedly was designed to benefit GS given its derivative positions priced in reference to the dollar ISDAfix. As a result, GS’s submissions were purportedly higher or lower depending on the bank’s then-current trading positions.
The conduct at issue in this case highlights the degree to which benchmark market rates are susceptible to manipulation. First, it shows rate-setting mechanisms that by design depend on the cooperative effort of multiple market participants still can be manipulated by individual participants. Second, it highlights the weakness of rate-setting mechanisms that are based on voluntary quotations by certain financial institutions rather than solely on transactional data. When relieved of the burden to tie their rate-setting submissions to actual transactions, participant banks can (and in some cases do) submit false or misleading information pertaining to hypothetical purchase and sell transactions affecting benchmark swaps.