Written testimony has been lodged relating to former HSBC FX trading head Mark Johnson’s appeal against his conviction for wire fraud in the US and it contains a few surprises – some of which may indicate stronger implications for the FX industry than the original trial, as Colin Lambert discovers.
The original trial of HSBC’s former global head of FX trading Mark Johnson was widely seen in the FX industry as being, aside from the outcome on his personal circumstances, a single issue case – what would the US legal system say about pre-hedging? The initial outcome – which is now being appealed – was a conviction and a jail sentence under the very broad umbrella of “wire fraud”, a decision that was seen as potentially having an impact on the FX Global Code, which explicitly (under certain circumstances), endorses the practice.
The written appeal documents have now been lodged with the Second Circuit Court of Appeals in New York and the defence documents in particular appear to suggest that the potential impact of a failed appeal could be even broader, and deeper, than first thought. They also, potentially, have serious implications for the FX Global Code of Conduct – a document in which the FX industry is placing great trust to indicate that past indiscretions are behind it and that it operates in a fair and transparent manner.
The defence documents describe “…the latest flavour of the prosecution’s many ever-shifting, imprecise, and contradictory attempts to explain just what, exactly, the crime here was, and it exposes why there was none”. They go on to argue that the prosecution has accepted some of its original accusations have since been proven untrue and, more pertinently for the FX industry, that the prosecutors have now introduced new accusations, which the defence describes as “a litany of new, equally meritless theories of deceit and intent to harm.”
The defence continues, “These bogus theories are nowhere in the indictment, were not presented to the jury, are unsupported by the record, and cannot sustain the conviction. Not only do they contradict what the government said below, the Government’s brief repeatedly contradicts itself, typifying the lack of fair notice or coherent standard of criminality that has become the single defining feature of this prosecution.”
Three issues highlighted by the defence appeal brief explain the problem facing the FX industry should Johnson’s team fail to overcome his conviction. Firstly, the defence appeal brief notes that one of the prosecution’s examples of “mis-statements” which induced Cairn Energy, the customer with whom HSBC was trading, to enter into the trade (thus becoming victim to the alleged “fraudulent” activity), was fundamentally flawed.
“The government argues for the first time on appeal that other HSBC employees’ alleged misstatements [were material and capable of influencing Cairn’s decision to trade] even though they were made after the 3pm fix, because they supposedly induced Cairn to “settle” the transaction by formally exchanging the currency two days later,” the defence appeal brief states. “But this erroneously assumes that Cairn was entitled to withhold its dollars after placing the orders mandating HSBC to deliver £2.25 billion at the 3pm fix rate. Cairn had no such right.”
A senior banking source who was actively involved in the creation of the FX Global Code says that “at no time” did the subject of a customer not completing settlement come up during discussion, except in the more generic area of a bankruptcy and default. “The idea of someone not delivering payment because they didn’t like the execution simply wasn’t thought of,” the banking source says. “I honestly don’t know what would happen if this became a possibility – how could you accept an order from a customer in circumstances where, if they didn’t like how it went, they could withhold payment?”
In reality, notwithstanding the outcome of Johnson’s appeal, banks have already moved to protect themselves against such a circumstance, something that itself makes the prosecution’s argument somewhat weak. “Banks have really tightened up the language in their documentation and had their customers re-visit it and re-sign it,” explains a senior risk officer at a global bank. “The language is in line with the Global Code and the documents sit alongside the existing ISDAs. Effectively the customer accepts that we will make best efforts to execute their trade in the best fashion possible for them, but that strange things can happen in markets. By signing the document the customers are explicitly acknowledging that.
“Can they withhold payment if they think there has been serious impropriety? That’s a difficult one because not only do they have a legal obligation to settle the trade, but there is also the question of timing,” the senior risk officer continues. “When was the alleged misconduct found? If it was after the execution it’s irrelevant of course, but if it’s before? I still don’t think they could withhold payment, they would have to raise their issues through different channels.”
In response to the question, ‘can a bank price in the risk that customers will not settle?’ the answer is a simple “No” from the risk officer. “Thanks to CLS and more lately clearing, we don’t price credit risk in FX now, it’s too cheap,” the senior risk officer observes. “It’s a feature of other markets, but will it get into FX? I don’t think so.”
With the Global FX Committee working on deeper recommendations for market participants’ disclosure documents, the senior banking source involved in the creation of the Code believes that the industry is ensuring that it covers as many bases as possible – including quality of execution. “The disclosure documents are getting longer and longer as a result of bitter and painful experience,” the banking source says. “They are basically saying to clients ‘we’re going to do our best but [stuff] happens’ and that there can be no guarantees.”
How Much is Too Much?
Another factor introduced at the appeal stage in the Johnson trial is the prosecution’s apparent change of mood around the profit made by HSBC. As well as accepting (in a footnote) that its original assertion that Johnson and other HSBC traders personally profited from the trade was incorrect and that all profits went to the bank, the government case now states, somewhat contradictorily, that either the entire profit was “ill-gotten” or that HSBC “unnecessarily” profited from the deal.
Notwithstanding the apparent confusion over what the prosecution is now arguing, the idea that all trades have a normalised profit payout concerns traders.
“If you’re pre-hedging a large trade you need to be as careful as you can, but there is too much that’s beyond your control,” explains a senior FX trader at a bank in Asia. “What happens if, in the middle of the prehedging ahead of a fix, Trump tweets something, or an unexpected statement from a central banker comes out and impacts the market? You could be halfway through the pre-hedge, the market moves against the customer, who still only wants the fix and although you pre-hedged in good faith you’re going to make some serious money.
“We often offer an improvement to a client if we have pre-hedged well, but a lot are set on achieving the fix and don’t want it,” the senior trader continues. “And overriding all of this is the size of the trade. The larger the trade the longer the prehedging window has to be and therefore the higher the risk to both bank and customer from an event.
“Pre-hedging is already incredibly complex and difficult to do, because you have to take into account other customer orders and you are busy with the day-to-day pricing, and that’s without worrying about market impact,” the senior trader adds. “Throw in the need to limit your P&L and it becomes, frankly, something I wouldn’t like to do – there’s too much personal risk.”
An algo execution manager at a bank in Europe believes heightened legal risk could see banks insist more on their clients paying a fee for execution rather than using pre-hedging, observing, “At least with a fee it’s all transparent and upfront – the client knows what the actual execution is going to cost them. The problem is, however, that under this scenario the client is assuming the market risk and how many really want to do that, especially in volatile markets? But if there are questions going to be asked about how much is too much money to make by prehedging then the only solution I can see is an algo with a full, independent, TCA report – and that means a fixed window executions and fees.
“This would also have serious implication for the major fixes,” the algo execution manager continues. “Effectively everything would have to be done in whatever window is dictated by the fix owner or users would have to accept a random, point in time reference rate that would often see them miss the mark. If it’s the first one then we are going to see much bigger moves in the fixing windows; the latter means tracking error which would be unpopular especially amongst index tracking asset managers.”
One other issue with trying to establish a “normal” profit margin for a trade is the nature of the risk book, especially at the major banks. “Orders are not handled in isolation – even on a segregated desk,” explains the COO of FX at a bank in North America. “We handle portfolios of orders and trades and that means actually pinning down how much money you make out of a trade is incredibly difficult and complex. Customers have asked us how much money we make out of them and not only would we not tell them unless it’s a specific mark out, we can’t.”
Ultimately, as with so many things in the FX market, there has to be an element of trust in the relationship, and that includes how much money a service provider makes from a trade. “We do have internal parameters around how much can be made around fix orders, as an example,” the senior risk officer reveals.
The head of FX trading at the Asian bank observes, “A lot of the time it’s a matter of common sense and instinctively we know whether a P&L outcome is OK. In more volatile markets the P&L will likely be larger than in very calm conditions – and it should be stressed that this means the bank could face a larger loss from prehedging as well as profit.”
Pre-Hedging to a Schedule
A third element of potential concern to the FX industry is more a nuance on the original issue of pre-hedging larger fix orders. While the initial concern was that pre-hedging could be found to be front running by the US legal system, a lot of this was dismissed by senior sources, one of whom, heavily involved in the creation of the Code, told Profit & Loss that “it was more of an issue for the US FX market than the rest of the world”, the implication being – as was the case with the initial DoddFrank framework – that offshore banks would simply stop accepting US counterparties in the market.
In the appeal document lodged by Johnson’s defence team it argues that the prosecution’s argument has changed and now revolves around exactly when a trade should be pre-hedged. Again there does seem to be a contradiction in the prosecution’s case when it now acknowledges that, as the defence appeal brief states, “For the first time on appeal, the government claims that “there would be no dispute that th[e] profit was earned honestly” had the pounds been accumulated “in several tranches”.”
The method of buying in tranches, alongside another argument elsewhere in the prosecution case that HSBC could have “drip fed” the order into the market (an option that was apparently turned down by Cairn as it would reduce transparency), suggests that unlike during the original case where one expert witness for the prosecution argued that HSBC could have assumed the risk post-fix, the prosecution now accepts that pre-hedging is a prerequisite for sensible execution of a large ticket.
The problem is, however, that while its own witnesses could not come up with a credible alternative to pre-hedging, thus suggesting that it is the only sensible way to execute the trade, the prosecution is also questioning why HSBC allegedly bought GBP 2 billion of the GBP 2.25 billion order in the 10 minutes before the fix. The defence appeal refutes this, citing trade logs and evidence that highlight HSBC was actually buying for close to an hour before the fix, but it also raises questions over whether the prosecution is arguing for a formal framework for prehedging?
If it is, market sources say this would be very difficult to impose, although, at an extreme, it could be done. At the top level, some sources argue that because all executions are different, thanks mainly to market conditions, it would be impossible to mandate how a trade should be prehedged. “Liquidity conditions are very volatile, and it’s impossible to predict with any certainty how things will look in the next minute, let alone the next 10,” explains the senior FX trader at a bank in Asia. “Good pre-hedging takes into account market conditions and by its nature must be flexible – if we just execute to a prescribed pattern then there will be market impact, which is to the detriment of the client.”
The algo execution manager believes that one solution could be participation rate strategies, which ensure that the party executing the order make up no more than a certain percentage of market volume. “The problem is, however, this is a ‘look back’ strategy and if liquidity disappears for whatever reason, the algo is suddenly a big part of the market and if you want to know what that looks like then take a look at the original flash crash in equities in May 2010. Then a participation rate algo aimed at being 8% of volume was suddenly closer to 98%.”
A New Look at Principle 11?
In spite of what appears to be a backtrack on the part of the prosecution that now accepts that pre-hedging is required to ensure better outcomes for customers, the issue still remains a hot topic in the FX industry – and concerns that the waters may get muddied further will only increase the noise levels.
While most accept that the Global Code provides a degree of clarity over prehedging (i.e., that it is acceptable as long as it is for the benefit for the client and does not disrupt the market), there seems to be a growing school of opinion that the Global Foreign Exchange Committee could revisit Principle 11, which deals with pre-hedging.
It is not only Principle 11 that deals with execution quality, however, for Principle 10 also has guidelines for handling fix orders that are relevant to the Johnson case. In this Principle the Code makes the case for pre- (and post-) hedging by citing, “Transacting an order over time before, during, or after its fixing calculation window, so long as not to intentionally negatively impact the market price and outcome to the Client” is an “acceptable practice”.
It adds that another good practice is “collecting all Client interest and executing the net amount”, which highlights the need to approach fix orders on a portfolio basis. Interestingly, the Code also includes in Principle 10 a recommendation that would appear to back how Johnson and HSBC executed the Cairn trade, when it states that examples of inappropriate conduct include, “buying or selling a larger amount than the Client’s interest within seconds of the fixing calculation window with the intent of inflating or deflating the price against the Client; buying or selling an amount shortly before a fixing calculation window such that there is an intentionally negative impact on the market price and outcome to the Client; and showing large interest in the market during the fixing calculation window with the intent of manipulating the fixing price against the Client.”
As more than one source has pointed out to Profit & Loss, that effectively says that executing large tickets exclusively during the fix window itself is contrary to best practice. “Principle 10’s examples provide guidelines for pre-hedging fix orders and I can’t see how what HSBC did is much different to that,” argues the senior banking source involved with creating the Global Code.
Certainly it now seems as though the prosecution accepts that HSBC did appropriately disclose how it would execute the order ahead of the fix to Cairn, and therefore pre-hedging is out of the firing line, at least for now. That has not stopped some in the industry from expressing the view that Principle 11 needs to be re-visited. One senior industry source, who was also actively involved in the creation of the Code believes Principle 11 is “The biggest gap in the Global Code.”
The source points out that in most US states ignorance of basic laws is not a defence, however if an accused party can show they thought their action legal because “lots of respected people do it”, that is a defence. “A written code of conduct should support a reasonable defence, but I am not sure Principle 11 does in this case,” the source continues. “The grammar is confusing, especially the use of the word ‘anticipated’ when discussing client orders, as is the balance between the need ‘not to disadvantage the client’, but also to take into account the portfolio of orders a participant may be holding.”
Profit & Loss understands there is, at the moment, a lack of appetite to reappraise Principle 11, however the senior banking source involved in the Code acknowledges that could change following the conclusion of the Johnson case. “We will wait and see what happens there before discussing the impact on the Code,” the source says. “The GFXC will probably need to tighten up the language at some stage, especially around how a participant handles the trade-off between pre-hedging a large order and its broader client portfolio.”
This is welcome news to the senior FX trader at a bank in Asia, who says, “All help welcome on this subject! We need clarity and, more importantly, something in the Code that allows us to go to clients who may have been negatively impacted by a pre-hedging strategy and explain what happened was just one of those tings and that we acted within the Code’s guidelines.
“Of course, before that it would be helpful to know whether the pre-hedging should take into account the impact on client orders,” the trader continues. “We might be pre-hedging for the benefit of the client with the large order, but should that take preference over other existing – smaller – orders we have from other clients? At the moment size matters because we take the view that if the order is executed in one hour or one minute it will have an impact and that is irrespective of whether we do the trade or another bank. I would also argue that if you pre-hedge well and the market impact is limited, then you are actually helping those other clients who may be held in the market.”
Whether or not Principle 11 is revisited, the FX industry will await the outcome of the Mark Johnson appeal with great interest, for at an extreme it could have a lasting impact on how it operates. Although most industry sources regard the new claims by the prosecution in the case with disdain – they certainly seem to demonstrate a lack of understanding as to how OTC and FX markets in particular operate – there is no avoiding the fact that the conviction being upheld would create a stir. “I think Mark Johnson did nothing wrong,” says the head of FX trading at the bank in Asia. “But if the US legal system thinks otherwise then we have a lot of difficult questions. The first will be, ‘how do I, or even do I, trade with US clients?’ The second will be what risks am I running in being a signatory to a code of conduct that appears to go against US law?”
When asked about the possibility of having to mandate maximum profits from a trade and possibly having to impose a fixed execution framework on pre-hedging large tickets, the head of FX trading says with a meaningful sigh, “The only way out of such a mess would be to forget being a principal to any customer transaction and the greater use of algos. That comes with a health warning, though, for there is nothing in there that is beneficial to the majority of customers.”