Recent market events have again stirred the investor community’s ire on the world’s banking institutions. While the financial crisis wounds are still very fresh, the media is again littered with stories of unscrupulous behavior: Libor setting scandals, insider trading convictions, and data releases being sold to the HFT speculators early without proper disclosure to the broader market. Confidence in the concept of fairness rings hollow in the minds of institutional and retail investors alike.
The currency market has been shaken as well by the headlines. Tuesday evening Bloomberg released a story detailing alleged market manipulation on some of the most widely used “price fixings” in the institutional FX markets: the WM/Reuters (WMR) benchmarks.
The story details that the UK Financial Conduct Authority (FCA) is investigating allegations from an unnamed European Investment manager. The allegations hit two major areas of concern:
• bank traders are colluding on large pending fixing orders
• “banging the close” – essentially aggressively front running client orders to influence the setting of the WMR price benchmark to the banks benefit at the expense of clients
The purpose of this article is not to opine on the legitimacy of the claims, but rather to shine a light on the reality of the WMR benchmarks, their use by buy-side investors, and how the sell-side manages the market risk created by these price benchmarks. FX Transparency believes that if properly utilized, the WMR price benchmark can be part of a best execution strategy for foreign exchange. Additionally FX Transparency feels that it would be difficult for market-making dealers to consistently influence benchmark by “banging the close”.
Like all Over-The-Counter (OTC) markets, sell-side risk management of FX trading positions (both proprietary and client related) involves active trading. This may involve buying ahead of orders in anticipation of price moves and anticipated investor demand. The term “front running” while an anathema to the heavily regulated equity markets is common risk management practice amongst all banks involved in the OTC FX markets that are trading as principal.
Despite being owned by a State Street subsidiary, the WMR fixings are considered by the vast majority of participants as an independent pricing mechanism. These fixings are tied into many investment mandates and are utilized to set the Net Asset Value (NAV) for many funds. The most popular use of these price benchmarks comes at month end when funds allow money to enter and exit, and passive investment strategies typically hedge to maintain proper ratios relative to their asset benchmark.
Broadly WMR has a good process to determine currency rates at set times. Because benchmark rates are being set by actual trades, it is by definition superior to a “survey”, such as Libor. Additionally, if compared to executing FX trades via standing instructions with your custodian bank, again, the WMR fix is a much better trading practice, even if there is collusion to push the rate around. The reason is because trades executed via standing instructions leave the custodian bank a 24-hour window from which to choose the worst rate, versus a 60 second window. Put another way, would you rather be short a floating rate option with a 60 second duration, or one with 24 hours of duration? Again, the WMR fixing is superior.
The downside for the WMR, as with any point in time price benchmark, is that you may execute at an unfavorable rate relative to the day’s range from time to time. But conversely, you will also execute at favorable times as well. If a consistent point-in-time execution strategy is applied during liquid times with oversight, execution cost results are typically good, and very good for those who execute at the mid-rate benchmark.
While it is possible that a group of dealers could have influenced the fixing process on limited occasions by attempting to “bang the close”, in practice it is much more difficult than what was reported in the Bloomberg story.
Dealers are putting up risk capital (i.e. they can lose money) each time they agree to execute trades for clients at the mid-market WMR fix. In attempting to manipulate the WMR benchmark, dealers must have strong confidence that they can push the market their desired direction. Given the depth of liquidity and breadth of participants in the OTC FX market this is never assured. Let’s play one possible the collusion / “bang the close” scenario out:
• Assume several banks have wind at this month’s end fixing that the passive US international money managers will be large sellers of EURUSD as the European countries in the MSCI EAFE Indices have performed well. After talking with several bank EUR dealers they collectively have orders to sell 5 billion EURUSD at 4 pm London. Banks typically accept orders up to 15 minutes before the respective fix. EURUSD is 1.3300
• The colluding traders decide to wait until 3:55pm London time to begin selling (to minimize the length of time unwanted EUR positive event/news items can hit the tape and influence their plan for the fix). They plan to sell most aggressively during the 3:59.30 to 4:00.30 window where the benchmark is set. They start at 3:55 and push rates down to 1.3280 as the fixing process starts. They have collectively sold 2.5 billion EURUSD, (2 billion for client flow & the other 500 million for their own proprietary positions).
• Unbeknownst to these dealers at the same time across 24 other banks there is a cumulative 8 billion of EURUSD buy orders. These have come from various sources including European fixed income investment managers who are hedging recent US treasury purchases, several smaller European bank treasury desks who are hedging their balance sheet risks, and 3 cash related merger and acquisition deals. All operating individually they have all decided to match their flow with WMR fixing window (3:59.30 to 4:00.30). They begin buying at 1.3280.
• The dealers attempting manipulate the market in EUR lower have 2.5 billion remaining and this gets easily absorbed by the incoming 8 billion buy orders. Now the market rapidly rises through 1.3300 and finishes at 1.3325 at 4:00.30 pm London.
• The fix is set at 1.3303. The dealers looking to “game” the fix end with a 1.3293 average, realizing a 10 tick loss on their substantial risk positions.
While this example shows the pitfall of attempting to game the WMR fixing, it is fair to say there is no easy money even when large orders are involved. It is impossible for even the largest banks to know consistently now much of any fix their net position entails even when enlisting information from other major dealers. History has shown us many examples where market participants of all stripes typically feel that they have market information cornered and the unforgiving financial markets deliver a large dose of humility though substantial losses. Additionally, many proprietary traders on both the sell and buy-sides of the market have attempted to create HFT alpha strategies around the WM/Reuters fixings for years with the vast majority proving unsuccessful.
In summary the allegations are serious and disturbing to read about. The influence and size of the WMR benchmarks makes this front page news but a dose of reality needs to be taken when examining this issue. The regulators will undoubtedly investigate and make determinations and suggestions to improve the process. In the meantime, the institutional market is still predominately a principal market with little to no regulation. The Ronald Reagan adage “trust but verify” applies. For institutional investors in international markets, this means independent FX transaction cost analysis (FX TCA).