Pound Crashes in a Flash, Points to a Maturing Trend

British Pound Sterling experienced a moment of panic during a rare flash crash event on Friday, 7th Oct 2016 just after 1:05AM CET, entering the Asia hours.

Most prominent was the fact that the price lost 6% – 10% during 2 minutes of trading. Or of no trading, since the execution was quite latent. First order was filled within one second, the second one took nearly two seconds, while third order already took 4 seconds to fill. 4 seconds was required to fill the fourth order. After that things got worse.

Exactly 45 seconds into the event orders took 12 seconds to fill. Interestingly enough these were all buy orders and one would expect lower fill latency for bidding since the pound was crashing. Spreads were recorded at over 300 pips on GBP/JPY for example. Lows on platforms were recorded under 121,90 for GBP/JPY, however providers would not let buy entries bellow 124,97 for example. Speculation in the market was that some providers showed even significantly more aggressive spreads than that.

Human error or even fat finger was supposedly to be the cause, yet usual finger pointing to machine trading was not left out.



Reality is that any fat finger would not crash the entire GBP complex this way. GBP move was uniform across the board and there was no obvious accentuation of a particular pair. Would somebody move the market by up to 10% by mistakenly selling 10bn of currency? While everything is possible, systems must be very ancient with the big player that managed to do it. Or perhaps it was not a fat finger, but a fat bottom that sat on the keyboard to sell pound more substantially.

Usual pre-order check includes reserve margin check, market liquidity check, trade space check and order space check along with others. Any order of this size would get a lot of scrutiny from market liquidity check. Execution would potentially run into hundreds if not thousands of partial fills with increased latency and procedure risk. Urgent warning and confirmation request, plus pre-process reject are all a given. But margin filter definitely was passed.

Source probably runs higher than 1bn dollar equivalent margin or credit line or certainly at least in the region of that. This leaves regular pre-order procedures to filter out an obscenely large market order command that passes margin check and for procedural latency risk reasons to manage confirmation procedure. One of them is the one that checks if the book in effect has partial depth relative to market command size. This alone triggers the filter.

Fat finger suggestions therefore are probably 99% misinformed ones, hopefully. On the other hand media does not forget to associate algorithm trading with the move. Strangely it is always forgotten that there are algorithms stabilizing the crash. Like for example our ioNectar technology and surely there were other buyers of the British currency. Yet somehow, we had problems getting filled as well.



The Brexit story earned a lot of focus as well in relation to the event. The negotiations how Brexit would happen could send GBP panicking. Hard Brexit to happen in March 2017 as Theresa May advocates was pointed as another reason. The Telegraph mentions GBP also could become unnerved by François Hollande comments that European leaders should take a “firm” negotiating stance with the UK.

Market theory supports such notions, but it can also argue against them. Could some talk from a politician send the market in a pure vacuum tailspin, when Brexit vote itself did not take out all air from the market, even if it became thinner? Brexit talk therefore does not seem to be the actual point of the GBP move. It could play a part in it, but it is unlikely it was the trigger and the driver of the event all in one.



The bottom line is that markets are what they are anyway and actually the event provided a brief opportunity for conservative investors to exchange reserves at attractive rates. Mostly what is not in the forefront of mainstream is that if we consider the market to be a discounting mechanism and if the exit due in five months was actually the reason for the GBP flash crash then the Hard Brexit is mostly priced in around lows. Any Brexit story with longer time horizon will not be as effective any more in pushing GBP lower.

Therefore I expect strong bidding on any meaningful GBP downside and I suspect potential drops in the currency could be very good buying opportunities. The field shows already prices could not hold at the lows for very long. Unless the Hard Brexit progresses to some even harder version, some British Pound Sterling crosses could enter stabilization.

Gregor Kozelj

This entry has 0 replies