With apologies to those loyal readers who normally part with their hard-earned cash to read this column, today I am going to make it “free to air” – mainly because I feel there is a message that simply has to get out there regarding FX execution and liquidity.
I am referring, of course, to the ongoing Mark Johnson trial in New York, where last week we were treated to former colleagues and experts opining on how to execute a large trade in Cable. I would also apologise to regular readers for repeating myself, but, to reiterate, I think there are some seriously poor assumptions made about how the FX market operates.
It’s bad enough, in my eyes, that the Federal Reserve chose the middle of the trial to announce a fine on HSBC and explicitly link that fine with the trial, but reading through the accounts from last week at the trial I found myself amazed at some of the opinions of so-called experts – as well as the naivety of Johnson’s former colleagues and customers.
Obviously it is the job of the prosecution to put as strong a case forward as possible, and that means pushing facts to the limit in pursuit of their goal – I fully expect to hear just as strident views from the defence team when their time comes – but what I am reading and hearing about the case so far seems to seriously misjudge how the FX market really works.
There are going to be more factors than execution quality at stake in this trial – it is over allegations of front running of course, so technically the style of execution should have little or nothing to do with it. The prosecution seems to be taking the stance that all trades executed before the 3pm Fix are evidence of front running – and I shall get to that. There are also allegations that secret codes were used to coordinate activity and that a fake Russian order was invented to hide the fact that HSBC was buying.
The jury will decide on the allegations around the “Russian name” and it has to be said, in some ways it’s not a good look – unless of course, there was actually a Russian name buying. Regarding the code phrases, however, this is a tactic long used by FX trading desks.
In the past I have been party to the use of codes and uniformly it was for the purpose of ensuring that other desks or offices didn’t learn of sensitive information. Regular readers may recall my story regarding the period when I traded Cable according to the position of five cranes outside my window, well one of our code phrases around the desks was “the cranes are pointing left/right” which told those that needed to know we were selling/buying sterling.
The point of code phrases is about protecting the customer, but it’s also about protecting the trader. Without putting too fine a point on it, some sales desks leak like a sieve as the salespeople seek to ingratiate themselves with an important piece of information – this is all well and good, but if I am sitting a couple of hundred short, I don’t want my sales desk telling people! Equally, a lot of trading desks have “open boxes” to offshore offices and you just don’t know who is passing that speaker box the other end, who may hear what is going on and either not understand they can’t spread it, or do so deliberately.
Away from foreign exchange markets, the House of Commons in the UK, that country’s parliament, had, for centuries, the code phrase “I spy strangers” to alert members to the presence of an unauthorised person or persons. The point is, we all use code words and phrases – how did Cockney Rhyming Slang originate? To help London’s East Enders communicate without law enforcement officers understanding them.
So it strikes me that approach is easily refuted, whether it be true or not that the phrase was actually a part of a front running campaign.
But what about the method of execution? The prosecution seems to have focused in on this and as a key element of the trial the verdict could have a profound impact on the FX industry if it were to go against Johnson – especially those participants with large amounts to hedge.
The prosecution, and their expert witness, appear to be arguing that if an order is to be executed at the Fix, it must be thus – there is no room for pre-hedging (another factor of the case that has implications for the industry, not least the Global Code). This bothers me and it is here that a lack of understanding of the dynamics of the FX market comes to the fore.
I have noted before that if you take the UK trading day to be nine hours long, a rough calculation using FX Joint Standing Committee semi-annual turnover data gives you volume in Cable of GBP114 million per minute. Globally, using the BIS data for 2010 and 2013, average daily volume in Cable was around the GBP 120 billion mark.
Using these numbers, I make the following assumptions, which I know are rough and open to adjustment, but I think they tell enough of the story.
The BIS data, if you take out UK hours, equates to about GBP68 million per minute, obviously liquidity is lower when the UK is closed. Looking at data from the BIS report on the sterling flash crash last year and speaking to people in the industry, my best guess is that in the Australian morning ( i.e., before mainland Asia gets going), this number falls to something like GBP20 million per minute.
Therefore, and to be overly conservative on the Asian data, it can be assumed that turnover at 3pm in London is at least six times higher and probably closer to 10 times when depth of book is taken into account.
On October 7 last year, according to the BIS report, someone sold about GBP250 million in the Australian window – Cable proceeded to collapse 12 big figures before bouncing nine big figures back.
In December 2011, when the Cairn Energy deal took place (and liquidity would be below average in December), the prosecution seems to be saying the entire order should have been executed in one minute. That is roughly GBP2.5 billion, so 10 times the flash crash order, into a market that is, roughly, 10 times more liquid.
It is very easy to see a similar outcome isn’t it? And that is my problem with the prosecution focusing so much on the pre-hedging. The average amount traded in the UK day in Cable is about GBP1.9 million per second – to execute the Cairn order in one minute would mean buying GBP41.7 million per second.
The consequences of such an action would have meant, I am willing to argue, that we would not have had to wait until October 2016 to join the ranks of those markets to have experienced a flash move.
The fill for the customer would have been truly awful because this calculation doesn’t take into account that a lot more speculative dealers are operating in the UK afternoon. At midnight UK there aren’t too many involved apart from a few late US traders, at 3pm there would have been an army of prop traders ready to jump on the bandwagon.
So the message is simple – I don’t know whether there was a deliberate attempt to mislead the customer, or even if HSBC traders did have a bunch of positions to take advantage of the order. I am not willing to support or otherwise the assertion by one witness that they were told to “ramp it”.
But what I am willing to state is that trying to buy 2.5 yards in one minute is pure madness and would have destabilised the entire market and ensured the customer (or the bank if they had offered a limit price on the order) would have been seriously out of pocket.
I note also, an excerpt from Chris Salmon’s speech on Friday when the Bank of England’s head of Markets said, “Suppose, for example, that a future flash episode happened to coincide with benchmark fixings in foreign exchange markets…”
Should HSBC have executed the Cairn transaction as the prosecution seems to think it should, then we would have found out – although it has to be noted it wouldn’t have been the “big daddy” Fix at 4pm London.
So for the prosecution’s witnesses and experts to argue that pre-hedging was wrong means, I would humbly suggest, they either learn about the real foreign exchange market or take their expertise somewhere else. There may have indeed been wrongdoing, the jury will decide that in due course, but as far as I can tell, the most wrong thing so far in this trial is an assumption that FX is a nirvana where you can execute large sums at will, with no impact.
To paraphrase my favourite adage about liquidity – that will only happen when you are wrong.