Skip to main content

The FX benchmarks rigging case

On 2013 and 2014, a set of financial authorities (including the UK FCA, the US SEC, the Swiss FINMA and the Hong Kong SFC) launched a series of investigations related to trading in the FX market. The trigger to the investigation was the publication of a news article by Bloomberg on 11/Jun/2013 reporting that "traders at some of the world’s biggest banks manipulated benchmark foreign-exchange rates used to set the value of trillions of dollars of investments". Apparently, this practice had been occurring for several years.

It is worth to note that, as in many other financial wrongdoing cases, it was a set of journalists, and not the supervision authorities, who found out about what was happening.


What was being manipulated?

There were two FX benchmarks being rigged:
  • The 16:00 (UTC) World Market Reuters (WM/R) benchmark – set as the median price of all trades executed in the FX interbank market within a 60 seconds window, from 15:59:30 to 16:00:30;
  • The 14:15 (CET) ECB benchmark – set as the market price in the FX interbank market at exactly 14:15 (point-in-time snapshot).

These benchmarks are used to value a high number of investments in the worth of trillions of USD held by different types of investors, including pension funds and money market funds.

Many fund managers give orders to trade FX specifically at benchmark rates in order to eliminate tracking errors.


How the manipulation occurred?

This is a case in which the wrongdoers, who were FX traders from some of the world's largest banks, engaged in two types of market abuse:
  • Collusion, as the traders were sharing confidential information regarding client orders and aligning trading strategies;
  • Marking the price (with the particularity that the price that was being marked was used as a reference to a wide range of investments in the worth of trillions of dollars), as such traders were buying/selling aggressively during the fix period.

These traders were doing this in a very organized and conscient manner, to a point that they even gave themselves team-related nicknames, such as "the 3 musketeers", "1 team, 1 dream" and "co-operative".


Here is an example of how it worked:

1. A client tells a bank that it wants to buy a position of USD (in exchange for EUR) at the next USD/EUR fix rate.

The bank agrees, meaning that it will sell him USD (in exchange for EUR) at such price.

This means that the bank has to buy USD, to then be able to sell such USD to the client. Thus, the bank is now exposed to USD/EUR fluctuation risk and is wishing for the USD fix price to come as high as possible (i.e., the USD/EUR fix rate to come as high as possible). The bank profit (loss) in this operation (commissions aside) will come from the difference between the fix rate and the weighted average rate that the bank paid.

2. The bank’s trader tells other banks’ traders that he/she has a large order from a client to buy USD (in exchange for EUR). The other banks’ traders also share their client orders for trading USD/EUR at the fix.

Imagine that the 3 banks had client orders to buy USD, and only 1 had a client order to sell USD. Then the 3 banks decide to buy USD at the market during the fixing period, while the other bank makes an agreement to sell USD outside the market at an agreed rate (allowing the selling bank to eliminate its market risk whithout putting selling pressure in the market).

The banks also shared information about their clients' stop orders, so that they could better manipulate the market price. If these traders knew that there was a large stop order to buy USD (i.e., to close a short-position on USD) at a certain price, these traders might make an effort to trigger such order since it would then lead to further price increases (the triggering of a large buy stop order means that a new large market order is suddenly introduced, exaccerbating the upwards price movement).

3. The 3 banks would then concentrate all USD purchases in the fixing period and would put orders aggressively (namely by triggering stop loss orders) to try to move the USD price up.


Fines, arrests and other consequences

6 large banks settled to pay a total of about 10 Bln USD for charges of rigging the FX benchmark:
  • Barclays: 2,3 Bln USD;
  • Citigroup: 2,3 Bln USD;
  • JP Morgan: 1,9 Bln USD;
  • RBS: 1,3 Bln USD;
  • UBS: 1.2 Bln USD;
  • HSBC: 0.6 Bln USD;
  • BofA/ML: 0,5 Bln USD.
In addition, Paul Nash, a former RBS trader, was arrested.


What happened to the FX benchmark rates after the scandal?

Following a set of recommendations made by the Financial Stability Board (FSB) on 30 September 2014, both WM announced a set of changes on 15 February 2015 with the goal of making any market manipulation attempt harder to pursue, including:

The window fixing period was widened from 1 minute for ‘traded’ currencies and 2 minutes for ‘non-traded’ currencies, to 5 minutes for all currency pairs;
The number of data sources contributing to the calculation of the reference price was increased.

The ECB announced, on 7 December 2015, a change in the publication time of the rate from 14:30 to 16:00. Apparently, the goal of this change (of delaying significantly the publication) was focused on making this benchmark pretty much useless for trading purposes, making clear that it is an information-only purpose benchmark.

Comments