In recent years the sell side has justifiably been criticised for its behaviour in the FX market. But should regulators and market participants be taking a closer look at how the buy side operates in this market? Galen Stops reports.
The FX industry has been rocked by a number of scandals in recent years and in many cases the implications of these scandals is only now coming home to roost.
Two of the largest custodian banks in the world, BNY Mellon and State Street, have agreed $714 million and $530 million settlements, respectively, related to allegations they systematically set disadvantageous rates for their customers in contrast to their claims to be achieving best execution for them.
Meanwhile certain investment banks have collectively been fined billions for allegedly colluding to manipulate the WMR benchmark, they have paid hundreds of millions to settle civil court claims making the same allegations in the US and could still be facing similar class action claims in London.
Then this summer the US Department of Justice (DoJ) arrested HSBC’s global head of FX cash trading, Mark Johnson, and issued a warrant for the arrest of his former colleague at the bank, Stuart Scott. The DoJ alleges that the pair are guilty of front-running a large client order and deliberately misleading the client about how the bank was interacting with their order.
As a result of all these scandals, the FX industry has been working on the much talked about Global Code of Conduct, the first part of which was released in May this year.
Thus far the investigations and court cases have all been directed at sell side institutions, with the buy side interested onlookers. The Code of Conduct, however, embraces all market participants. That said, in each of the scandals the argument has been made that ultimately end-investors were materially disadvantaged by the unethical behaviour of sell side FX market participants.
But in some cases, it has also highlighted the degree to which these end investors have been badly serviced by the firms that are responsible for investing on their behalf. At best this can be viewed as a lack of thoroughness on behalf of investors, at worst it can be taken as willful negligence.
Part of the problem is simply the way that some fund managers view the FX market, given that often the FX trades are just a necessary byproduct of their main investment objective.
“In the early days of electronic FX trading, the execution was considered more of a ‘nuisance trade’ relative to the underlying position, the FX was only meant to cover or hedge and was often just an afterthought. And for that reason, the asset management and real money community in particular, didn’t focus as much on the FX part of the trade because it was a residual of their equity or fixed income position and historically executed by their custodian as a ‘service’.
“That’s why standing instructions were taken advantage of for so long, because no one was paying attention to that cost, they were only concerned about their equity execution cost at the time,” explains Christopher Matsko, head of FX trading services at Portware.
He adds: “Once all the commission was squeezed from equities people started paying closer attention to their costs in other asset classes and that’s when they started shining a light on the standing instructions trade and the executions by the custodians. That was the first scandal, and then the second was when we found out about the bank “cartel” chatrooms and the subsequent manipulation of the Libor and WMR rates in an effort to improve the bank’s market position over its clients.”
Likewise, Alex Dunegan, CEO of Lumint, a currency management firmt, says: “In some cases, it’s clear that fund managers have not been paying enough attention and haven’t had the right policies and procedures in place. The result of this is that they’ve been giving away money, and in the end fund performance, which is what hurts their clients.”
Tax on Performance
A paper written in 2012, “Asymmetric Information and the Foreign Exchange Trades of Global Custody Banks”, argues that fund managers were actually aware that they were overpaying for their FX transactions when they left it up to their custodian banks to execute for them.
Using the complete trading record of a mid-sized global custody bank during 2006 to study how markups at these at custodian banks were implemented, the paper notes that the average markups on standard custodial trades “are indeed quite high”.
Despite this, the authors find that although the custodial banks’ client funds could choose to trade OTC with their custodian and pay smaller markups, “they do so only 3% of the time”.
Running through a gamut of data, the authors “test and reject the hypothesis that funds are ignorant of the relative markups by showing that funds are more likely to trade OTC for larger trades and for currencies facing the biggest markup differential”.
In other words, when the amounts were small, the fund managers were willing to overpay for their FX transactions because the difference was less noticeable. But over time these amounts could add up to a significant quantity and, given that smaller amounts are traded more commonly than larger amounts, this overcharging acts like a perpetual tax on performance.
Deciding that some orders are too small to be worth actively executing themselves and that custodian banks are providing a service that warrants accepting in some cases marginally less optimal pricing could perhaps be viewed as a reasonable approach by a fund manager.
But a number of market sources question whether many of the fund managers that have taken this approach would be able to provide adequate analysis to actually justify it if they were required to do so.
Paul Aston, CEO of Tixall Global Advisors, goes a little further in his assessment of the situation.
“Most international investment managers don’t allocate the dedicated and specialised resources necessary to ensure best execution in FX. This is because in many international mandates, currency is rarely part of the performance competency set for which a manager is selected and hired. The result is that you tend to get currency execution policies that are suboptimal and chosen for the convenience of the manager rather than the benefit of the underlying investor. In essence a ‘fiduciary lapse’ can get inadvertently built into the investment and trading process itself as a defect of how the mandate is defined and awarded in the first place.”
Aston argues that the role that the WMR Fixing rate plays in many investment mandates is a case in point. The WMR Fix is the mark-to-market valuation rate used to value many of the benchmark indices used to evaluate manager performance. Managers tend to be measured and compensated against these benchmarks in terms of tracking error, meaning that any deviation they make away from the benchmark affects their relative performance. But if a manager is not explicitly awarded for making a currency decision, even for best execution, then there is no incentive for a manager to do anything except execute FX transactions at the WMR Fixing rate, a service provided by many dealers.
But according to Aston, “This breeds all kinds of problems from a fiduciary perspective.
“Firstly, by leaving standing instructions to simply be filled at the Fix, a manager is implicitly transferring their fiduciary responsibilities for best execution to the sell side. This situation is not only fraught with conflicts of interest it is also highly suboptimal, because the manager has transmitted the details of their order to the counterparty well before a firm price has been defined or negotiated. This results in information leakage that can be exploited by the market and can generate adverse price action. It represents the type of problem that led to the WMR Fixing scandal in the first place.”
“Secondly, although execution at the Fix is convenient, it eschews the opportunity to pursue better execution using other simple strategies.”
For example, Aston says that the WMR Fixing price is more likely to reside in the extreme tail of the day’s price distribution than the middle and that executing FX transactions throughout the liquid hours of the day using a simple TWAP or VWAP strategy would be more effective in producing prices that are much closer to the middle of the days pricing distribution.
He continues: “The difference between the two approaches can be substantial, as much as 20-60 bps, depending on the currency. Thus, if we assume that the WMR price tends to move adversely away from the center of the day’s price distribution, pursing a WMR execution strategy vs. a TWAP/VWAP strategy can be leaving 20-60bps of performance on the table. If a manager rebalances 12x per year, this can add up to represent 243-744 bps of uncaptured performance.
“To me, the whole idea behind ‘fiduciary responsibility’ is to always act in the best interests of the client and to place the client’s interests before one’s own. When I look at the practice of executing at the WMR Fix, I see an execution policy that may be convenient and beneficial for an asset manager, but that may not be in the underlying investors best interest if they had a choice.”
Combined with the idea of fiduciary responsibility, Aston makes the case that best execution implies a set of systematic policies, protocols and procedures that are devised to protect and enhance the net asset value of an investor’s portfolio. Taking the most convenient option or relegating responsibility for doing this to another party, he says, going against these principles.
“At the end of the day, we have teachers, firefighters, police and municipal workers entrusting their future welfare to the system of investment policies and practices we have in place and we have to ensure that these procedures are carried out and scrutinised at the highest fiduciary standards,” he concludes.
For all the talk about restructuring the WMR Fix, lengthening the window or providing an alternative benchmark, Dunegan says that the notion that there should be just one benchmark for market participants to adhere to is misguided.
He points out that an equity managers have different funds that are benchmarked depending on the focus of that fund, so that what they compare themselves to is based on the type of strategy that they’re deploying.
Dunegan argues that managers should employ a similar approach in FX, so that if a firm is trading in Japan and doing all their FX trading at the end of the market close in Japan it should be benchmarked relative to when it executes rather than at the WMR. “The root issue that’s causing a lot of problems on the buy side is an educational one. These firms need to have a better understanding of how to produce benchmarks that are relevant to them and then how to evaluate them,” he says.
Despite this, Michael DuCharme, head of currency strategy at Russell Investments, argues that the WMR does still serve an important purpose for many buy side firms. “The word is out about trading at benchmark times. However, trading at the fixes became popular because having a published rate at a particular time met critical investment needs such as fairly managing investor subscriptions and redemptions.
“The fix also allows the investor to value international portfolios and then compare them to a benchmark or other portfolios. Those needs still exist. Thus, investors still trade at the Fix even though other times might provide better trading opportunities,” he adds.
When considering if there has been a fiduciary lapse on the buy side another issue to consider is that the banks view investment managers, not their underlying investors, as their clients and this can potentially lead to incentives between all three becoming misaligned.
This is particularly relevant when considering the concept of “cross subsidy”. There was a race to zero around WMR fees – something that anecdotally a number of market participants say is beginning to happen again – because the buy side firms were constantly demanding lower rates for benchmark execution from the banks.
One senior banker explains that some buy side firms would either dangle the carrot of offering to trade more of its other fund’s FX with a particular bank, or the stick of threatening to trade less of its fund’s FX with the bank, in order to secure these lower prices.
So an investment manager might have some FX that it trades passively, because it is doing index replication, and other parts of the portfolio that it trades actively. But offering the active flow, which is valuable to the bank, in order to get better prices for its passively traded FX can be bad for investors of the underlying actively traded fund, because they now might not be getting the best price that they could.
In effect, the actively traded funds have just been used to subsidise the passively traded ones.
Not everyone is so quick to say that the buy side is at fault for failing to get best execution for their clients in FX, with DuCharme making the case that the real issue is the fundamentally problematic structure of the FX market. “It would be easy to lay blame, but we think the real fault lies in the structure of the foreign exchange market,” he argues. “It’s an over-the-counter structure involving many simultaneous bilateral deals in a largely unregulated (although that’s changing) and decentralised market.
“No institution, such as an exchange, exists to collect and post trade information. As a result, some investors find it challenging to identify “where the market is” when they want to trade,” he continues. “Additionally, the FX market lacked transaction cost analysis capabilities so investors had little information with which to evaluate their transactions and to modify their trading behaviour.”
The solution, according to DuCharme, is to move away from the more opaque bilateral trading model and towards electronic trading venues that can provide investors with “vast liquidity, visibility to current exchange rates, and access to ‘depth of book’ prices”.
Algos for the buy side are another solution to some of these execution concerns that are frequently touted by some market participants. But, just as with bank provided transaction cost analysis (TCA), there are potential conflicts of interest with regards to principal trading banks providing algorithms to their clients.
“One of the biggest concerns for large buy side traders today is market impact, that’s why algos are getting a lot of take-up right now. But at the large global buy side firms there seems to be an even divide between the heads of desks who like bank algos and are happy to trade them all day and those that think bank algos are the next fixing scandal because now one bank knows their entire order and there exists the potential for information leakage,” says Matsko.
Although he describes that the growing usage of algos by buy side firms is “a step in the right direction”, Dunegan also expresses unease about whether principal banks will be able to see the algo trade flow and questions how buy side firms are meant to effectively evaluate all the different algo offerings being provided by the banks.
He also cautions: “Different trade scenarios require different trading strategies, sometimes algos will be the best solution and sometimes they won’t. But the firms automatically gravitating towards algos is a similar situation to firms choosing the WMR when they didn’t have a benchmark before, it seems like the easy option. The problem is that FX is a cost centre for most buy side firms and as a result they’re looking for a one- size-fits-all solution.”
If anything is likely to really change buy side behaviour as it pertains to achieving best execution in FX for their clients, it almost certaintly will have to be regulatory driven. Aston and DuCharme claim that momentum is already building on this front.
“I think that not only regulators but opportunistic litigators are looking at this issue as well,” says Aston. “Some of them are going through prospectuses and statements that have been made to investors claiming to have gotten best execution, but the problem they’ve got is that it’s difficult to get empirical evidence showing that they definitively didn’t get best execution.”
DuCharme adds: ”The importance of the fiduciary aspect is this: regulators around the world have become much less patient with funds and investment firms who do not manage all their costs – including FX costs – effectively.
“Basically, they have decided to make funds and their managers responsible for every critical judgment—that each decision is defensible as expert, prudent, and without conflict at all times. Investors can manage this fiduciary requirement if they can show that they’ve delegated responsibility to a service provider who is also a fiduciary.”
In contrast, one banker expresses a much less optimistic outlook regarding buy side behaviour in FX. “For the buy side to change these firms first have to care, and I don’t see them caring,” the banker says, agreeing that a regulatory push is needed. The banker adds, however, that they don’t see any such push on the horizon as regulators simply have too many other pressing issues stemming from Dodd-Frank and the swathe of regulations currently being introduced in Europe for this to be a focus for them.
If no significant regulatory scrutiny is forthcoming in the near future it will be left to the FX industry to try and educate investors and place greater pressure on fund managers to clarify and justify their best execution policies.
As an industry-led solution to ethical problems in the FX market, the Global Code of Conduct has been largely well received, although many still question its enforceability.
If the code is successfully implemented then it could represent a blueprint by which to help ensure that buy side firms are fulfilling their obligation to achieve best execution for their clients in FX.