What to Make of the HSBC-ECU Outcome?

A UK judge has thrown out claims brought against HSBC by ECU Group that its FX traders deliberately triggered stop losses and front ran the firm’s orders. The action which is one of several brought by ECU Group against multiple banks, throws interesting light upon the bank-customer relationship – light that remains relevant in today’s market.

Before getting to that though, a first observation is that ECU Group clearly hasn’t had much luck with its timing, as it went to the trouble of filing a similar claim against UBS (the other banks it has filed against are Barclays, Citi, Deutsche Bank, Goldman Sachs and Natwest Markets) on 29 October, only for Justice Clare Moulder to dismiss their claim against HSBC in a judgement published on 1 November. It remains to be seen how the erstwhile currency manager proceeds with the weight of a negative judgement against it.

As far as the claims against HSBC are concerned, reading the judgement the sense is that Judge Moulder believes that in 2006 the company might have had a case to bring, but in 2017 (when it first started proceedings) it had missed its chance. ECU claimed that it accepted HSBC’s assertion in 2006 that it had investigated the currency manager’s claims of front running and stop loss triggering and found nothing, but had been prompted by the events surrounding the chat room saga and the conviction in the US of former HSBC global head of FX Mark Johnson, to re-visit its claims.

As the judge rightly points out, Johnson was not even employed by HSBC at the time concerning the ECU orders, she also makes a point of observing that because the conduct for which HSBC accepted regulatory sanction took place between 2008-13 (as cited in the US and UK judgements against HSBC and others), this did not mean the practice was going on some five years before.

Best Practice?

Notwithstanding that the court did not make an actual judgement on whether the orders were front run and/or triggered, the heart of the issue back in 2004-06 is very relevant in today’s market – because it all comes down to pre-hedging. Putting aside the recent guidance from the Global Foreign Exchange Committee on what constitutes pre-hedging, there is little doubt that in the period concerned, activity ahead of a large stop loss order, as left by ECU, was commonly described as pre-hedging.

The first instinct, reading the judgement, is that HSBC was pre-hedging the orders. This, allegedly, according to ECU, was against its explicit instructions. The problem is – and again this is as relevant in today’s market as it was back then – it could be argued that the client was acting unfairly.

As highlighted in the judgement, ECU Group was leaving large stop loss orders – in the low hundreds of millions base currency – with multiple banks, probably the ones they are now suing. The judgement cites occasions when the currency manager complained to HSBC about the fills they got on their stops, noting they were often the worst, one being nine points away from the stop level.

It is reasonable to ask, however, exactly what did ECU expect when not only did it leave an order to buy (as was the case on one occasion) over 250 million USD/CHF, but it often left similar orders throughout the market? If a bank is going to wait for the level to be hit (and one USD/CHF order was executed around 3pm EST, when the market is thin – it was also post-FOMC), the nine points slippage would be a good outcome.

The hard fact is, especially with the data and analytics now available, a customer has to accept one of two undesirable outcomes –slippage on a stop loss, or the increased likelihood of execution due to pre-hedging.

Of course, by complaining about the slippage and telling HSBC it was doing a bad job comparatively, ECU was tipping the bank off to the fact that it was leaving multiple orders at the same level (no-one seems to have questioned this during the trial which is a little surprising, but may have become an issue had the case not been time barred). Given a customer complaining about slippage, and the knowledge that other banks had similar, if not identical orders, it is no surprise that the bank sought to improve the fills by pre-hedging (it could, of course, have refused to handle the orders).  This, inevitably, led the customer to complain about being triggered, proving once again that there are some occasions when, as a service provider, you just can’t win!

The hope is that customers active in the market today, take a look at this judgement and consider how they interact with their providers. The hard fact is, especially with the data and analytics now available, a customer has to accept one of two undesirable outcomes –slippage on a stop loss, or the increased likelihood of execution due to pre-hedging. In the first, they rely upon the skill and resources of the executing party (and bad executions will show up in the data); in the second they have to understand that while the executing party will (should) do their best, inevitably, every dollar bought or sold mathematically increases the chance of the market moving towards their order.

Throw in the fact that a number of other parties may also have orders  at the same level and the choice becomes even more stark – accept even greater slippage as multiple LPs hit the market at the same time; or see the order triggered even earlier because multiple LPs are pre-hedging.

ECU cited as evidence that the orders were being traded ahead of, or front run (their claim swung between the two phrases) and were triggered with only two points slippage – and that does indeed sound like an order being pre-hedged. If orders were left with multiple parties, however, how can they know who, if anyone, is pre-hedging?

In the strict letter of the law perhaps HSBC should have waited and executed with large slippage, however it is notable that one piece of evidence cited is an ECU Group manager observing that he expected the market to move “30, 50, or 60” pips when an order of a certain size was executed. One has to ask the question, if they are expecting such a move, how can they, in all decency, expect a fill within a few pips? It’s effectively saying that the bank – in particular the trader – should wear a massive loss on their behalf.

Remarkably, and highlighting how it had its suspicions in 2006, ECU left what was referred to as a “trap” for HSBC in handing out orders for CAD/JPY, but while they were effective immediately for the other parties, for HSBC they were “good from” a certain time. Within five minutes of that window opening the order was stopped, which is not, it has to be accepted, a good look for HSBC. Again though, if the bank is aware other parties have the order (and those parties themselves know this), what are they to do?

Strategy

It is a revealing insight into how ECU was operating back in 2004 that the firm had little or no access to live market data or trading technology. In a submission to the court, ECU CIO Michael Petley said that he relied upon his charting system to follow the market, although he did also have access to a Bloomberg Terminal, and 2005 was around the time that it started actively supporting FX trading. It seems remarkable that a firm trading such large amounts in FX did not have access to some form of market data, even if it was a single dealer platform. ECU could even have accessed the website of a platform like FXall, that used to publish indicative streaming rates on its homepage.

There is also the question of how unique its underlying FX trading strategy was. The firm was entirely systematic, relying upon technical analysis for its decision making. Aside from wondering what data populated the charts (and whether it could be used to monitor markets in real-time), there is a school of thought in FX markets that technical analyst-driven traders often leave themselves open to a stop-loss hunt.

With many using the same data, the argument is that stop levels become obvious to market participants, and indeed every FX player, including HSBC, would have had a technical analysis team providing thoughts and ideas to the trading team. In these circumstances, and this was a regular feature of FX markets for decades, whispers of stop loss levels circulate the market, often leading to an assault on the level concerned – critically, however, not by the bank holding the actual order. In such circumstances, it is probably very hard to pin down the stop-loss hunt on one or two players.

The Benefits of Transparency

If anything can be derived from a saga that, in all honesty, presents few of the protagonists in a good light, it is that the FX industry has, thanks to the Global Code and wider push for transparency of action, thrown more light upon behaviour around stop losses. Those who hold the order are aware they have to act fairly and in the interests of the client – who equally has to accept that there is a cost to be paid for any stop loss, it’s either money or time.

The case also highlights how there has been a breakdown of trust between banks and certain of their clients – in 2006 ECU Group seemed, at least some in the company did – to accept that HSBC had done the right thing. Through the prism of the chat room scandal that swung 180 degrees and the client sees nefarious activity in every deed.

The key takeaway from this case should be that responsibility goes both ways and that an open and honest dialogue, especially around how orders are going to be handled, can avert a lot of unnecessary conflict. If a client is unwilling to behave reasonably around large stops, then the executing party should have the right to refuse to watch them – and it is to be hoped that the vast majority would take the same path, thus forcing the customer to change their approach.

Equally, however, the executing parties need to be careful over how they pre-hedge, indeed some dealers spoken to by The Full FX observe that any (disclosed and agreed) pre-hedging is now done through a careful skew, rather than aggressing the market. This, they argue, makes the pre-hedging harder to spot.

One final observation on this is also a core plank of the Global Code. The judge makes clear that she thinks ECU should have pursued the issue in 2006 when market data would have been available to help an investigation. It’s hard to disagree with that observation. Therefore, what should be driven home from this outcome is clear – silence is or acceptance of behaviour you are uncertain about is unacceptable. If you see something, say something.