According to a rather in-depth analysis from Bloomberg this morning, some of the world’s largest banks may be manipulating yet another market for their own benefit. In light of the LIBOR fixing, aluminum warehousing, electricity manipulation, and a host of other allegations, it should come as little surprise that a market being dominated by only four of the world’s largest banks could be subject to manipulation. The top four banks, Deutsche Bank, Citigroup, Barclays, and UBS, account for 50% of global currency trading.
The alleged manipulation surrounds the final minute of trading. It is at this time, the MW/Reuters rates are set. This index is an important tracking figure for many mutual funds, pensions, and other asset managers.
WM/Reuters rates, which determine what many pension funds and money managers pay for their foreign exchange, are published hourly for 160 currencies and half-hourly for the 21 most-traded. The benchmarks are the median of all trades in a minute-long period starting 30 seconds before the beginning of each half-hour. Rates for less-widely traded currencies are based on quotes during a two-minute window.
As banks take client orders throughout the last hour of trading, they are able to see any dislocations that may occur from executing the orders. With that knowledge, it’s possible for traders to make their own side bets while batching together client orders in order to exacerbate the moves even further. Doing so in the last minute to effect the MW/Reuters rates can have such drastic effects as to shave five basis points off performance if a fund was trading only at the close.
The graphic below shows some of Bloomberg’s findings, and as you can see some of those last minute moves appear a bit curious. It’s also worth noting that Bloomberg competes with Thompson Reuters in providing this type of market data to participants, and stands to gain from any weakness at Reuters.
Bloomberg News counted how many times spikes of at least 0.2 percent occurred in the 30 minutes before 4 p.m. for 14 currency pairs on the last working day of each month from July 2011 through June 2013. To qualify, the move had to be one of the three biggest of the day and have reversed by at least half within four hours, to exclude any longer-lasting movements.
The sample was made up of currency pairs ranging from the most liquid, such as euro-dollar, to less-widely traded ones such as the euro to the Polish zloty