To great cheers and joy unbridled, today’s column is not about last look! Instead, and I have
to stress I am unsure of the answer to this, I thought I would ask the question
– should the FX market shut down for a few minutes every day?

This would, of course, be a radical departure from the
historical norm, back in the day, once New York hit 5pm, all traders and
brokers changed the value date, and in the modern era where so much is handled
by machines, we are proud – as an industry – to state we are a “24/5.5” market.
The thing is though, the machines need to be reset – every day.

As I understand it currently, most venues shut down for a
few minutes (it varies by platform) between 5 and 5.30pm New York. This is
staggered – although I don’t think that is deliberate, it’s merely how things
have evolved as some take longer to reboot than others and some have more rolls
to calculate than others – to ensure there is always a market.

There is commercial sense in this for the multi-dealer
venues, of course, if they happen to be one of two or three platforms operating
if something happens (as opposed to the 10 or 12 there normally may be), they
will see more volume. There is also a valid argument that staggering the
closures will mean a market is always available for clients.

But would things be simpler if everything just stopped at,
for example, 5.15pm New York? It shouldn’t have to be for long – after all
everything else technology-related has got quicker over the past decade surely
the reboot process has as well? I would imagine if everyone knew the shut down
was coming and could do their roll calculations in advance, the closure may
only need to happen for a minute or two?

This question first popped into my head when I heard about
another strange trading episode around this time of day. One source tells me
they traded in Cable, on a public platform, at what they termed “a strange
rate” just after the New York close a month or two ago. Another source cannot
remember the details but confirms that someone hit through the bid aggressively
on Cable just after the New York close.

Details, as I say, are sketchy, but it seems someone put an
offer in an ECN more than 20 points below the current bid. They got filled and
there were still bids left over so the market quickly bounced back as the mean
reverting machines stepped in. All that was left was someone scratching their
head over getting Cable at, for example, 07, when, to all intents and purposes,
it didn’t trade much below 25.

Now this could be a simple case of an e-trader in New York
wanting to go home and getting an order from a client that they wanted filled
quickly, in which case I hope there were questions asked about firstly the
quality of the advice given to the client; secondly, the quality of the fill;
and lastly the client’s intentions.

It seems a silly time to trade in anything other than
one-off small amounts, so why did someone try to hit the market? I am told that
several platforms were in the reboot window so there were less sources of
liquidity (not that that matters – it would have been rejected on last look
anyway…sorry, can’t help myself!) and, it seems to me at least, there was a
risk of triggering a flash move.

Another issue concerns people wanting to trade on
announcements from the Reserve Bank of New Zealand. Obviously the Kiwi is not
one of the major markets, but the equivalent of $105 billion is traded in it
every day so it’s significant enough, and currently, the RBNZ monetary policy
announcements coincide with the reboot window.

It could be argued the central bank should change when it
makes these announcements but there is a routine of sorts to be gone through
and if there is a 30 minute window during which time some platforms will be out
of service that process may be messed up.

If there was a smaller window, during which time the foreign
exchange market is officially closed, the RBNZ could make the necessary
adjustments to ensure the smoother operating of FX markets around its
announcements.

I suppose the counter-argument to the points made above is
the question, “what happens if an event occurs during the closure?” It’s a
valid question but events also happen at weekends and traders have to wait
until 7am Monday morning in New Zealand (the official market opening is actually 5am Sydney) before trading (by which time the
market has already priced in the event), so the same could happen here.

In the pre-machine generated liquidity days I could see an
argument against this because there would typically be quite a few prices on
the way down (or up) to hit, but in the modern market, where there are less
resting orders, it seems to me if you’re on the wrong side of the event you’re
stuffed anyway so another minute or two will make little difference (it could,
very rarely, even be a benefit as it would take out the chances of an
over-reaction).

Whilst it is not a big issue at the top end of town, it
should also be noted that a large number of FX market participants do not have
access to aggregation or even multiple platforms on which to trade. It could be
argued that they should have more than one, but there is a cost involved in
aggregating streams and for many corporates and individual traders it simply
isn’t worth it – in the case of the former, the cost of aggregation for what is
a simple hedging process is likely to be one or two salaries.

If the FX market were to have an official close it would not
remove the opportunities for the type of risky behaviour I mentioned above –
there will always be that tricky two hour window – but it would perhaps allow
LPs to see who is trying to hit them across multiple streams. This might help
if, and we all hope not, such action does trigger a flash move.

As I wrote at the top of this piece, I am unsure if this is
a good idea or not, my instinct says not, but it might be one worth
investigating.

Off the top of my head – and given my preference for a
continuous market – I think my support would be for streamlining the existing
process into two windows, during which two distinct groups of platforms would
shut down. If these were clearly defined as in who shuts down in what window,
liquidity consumers could make their choice in selecting their platforms with
the intention of a constant supply of liquidity.

I don’t know to what extent this would impact the LPs
themselves but I would think at the moment they are dealing with a number of
different shut downs at different times so this could even simplify their
processes.

Technology has moved us on tremendously but it has also
raised issues, admittedly insignificant in this case. By having a shorter but
fixed window (or windows) for the necessary admin to be done, people will be
able to plan better and, hopefully, we can guarantee access for all, on a
24/5.5 basis, without creating too many overheads.

And now that I have added item number 197 to the tech teams’
to-do lists, I will sit back and await their thanks…

Colin_lambert@profit-loss.com

Twitter @lamboPnL

Twitter @Profit_and_Loss

Colin Lambert

Share This

Share on facebook
Facebook
Share on google
Google+
Share on twitter
Twitter
Share on linkedin
LinkedIn
Share on reddit
Reddit

Related Posts in

Profit & Loss is no longer publishing

Thank you for 21 great years of support