Monday’s column got me thinking about
another aspect of the changing FX market, driven by the continued shift in
balance from human to machine-based liquidity provision. We have spent a lot of
time talking about the equitisation of the FX market structure, but I wonder if
the increased automation in FX markets is leading to the equitisation of FX market
Obviously a changing market structure will
lead to changing market behaviour in some ways, but I sense that the FX market
is stopping short of fully embracing the equity market model. Recent Bank for
International Settlements data shows that in spite of the reform process and
some regulators’ desire for an equities model, when it comes to derivatives
trading, customers are voting with their business and staying with OTC markets.
Of course they are able to do this by
virtue of the OTC industry’s efforts to reform, especially around the clearing
piece, but they also clearly like the OTC model. I found it wryly amusing the
other day to read a suggestion that the FX market had reformed by adopting
equity models such as ECNs, after all, Matching and EBS have only been around for
24 and 23 years respectively!
Such an observation highlights the lack of
understanding that drove the broader financial reform process. As I, and many
others, stated incessantly in the hope the regulators would listen, FX is
different. It is not a threat to the global financial system. Additionally we
pointed out that at the less exotic end of the derivatives space, interest rate
swaps etc, customers need bespoke pricing – the type of which is rarely
available on a standardised model such as an exchange.
The BIS data proves to me, if it ever
needed proving, that, as someone told me last week, the “Singapore Model” of
market reform has won over that from the CFTC in the US. In Singapore the MAS
stated on the day Dodd-Frank was enacted that whilst it fully agreed with the
need to clear certain products, it would not mandate where or how those
products were traded. Looking at the volumes and the data from the BIS it seems
to me that is what people want.
But even while the equitisation of the
market structure may have stalled, should the focus now be on the markets
themselves? Is FX market behaviour starting to mirror that in equities? And if
I don’t have empirical data on this, it is
more an estimate based upon my years in the game, but price action does seem to
have changed, most probably triggered by the SNB event. General activity levels
are lower and it has become a much more event-driven market, something I would
suggest also reflects the reduction in human traders.
We seem to have long periods of not a lot
happening following by intense bursts of activity, and to me this signifies the
domination of the machine generated, data-led, liquidity model. It also, of
course, is a reflection on the growth of the agency model in FX amongst banks,
as well as the changing risk management profile of LPs – something I
discussed at length in Monday’s column.
It is hard to know whether such a shift in
behaviour is either permanent or good for the industry. On one hand you can
look at the growing geo-political risk and uncertain macro-economic outlook,
overlay it with central bank policy divergence and the increased chances of a
move away from zero interest rates and see this behaviour being very much short
The problem with that argument is we have
existed in such a condition for quite some time now, and market behaviour
hasn’t followed the traditional path. It was observed to me last week that people
trained in traditional market and finance theory will be lost in the current
market because nothing is what it seems and markets react in very different
fashion to how traditional theory dictates.
This is, I believe a direct impact of the
data-led market making model. Put simply, in the past when there were more
human traders (who were taking and holding risk), they would have been
anticipating some of these events and moves would have been longer developing
and smoothed out. Now, one day no-one is paying attention, the next, an
announcement is made or an event occurs and everyone tries to jump through the
same (pretty small) window at the same time.
One aspect of this is the growth of
modelling in markets around events – highlighted during Brexit. As someone
noted to me this week – it’s been a long time since people were able to make
serious money in options around events. The point being the options desks are
modelling market behaviour, using the same data the pricing engines are using,
and pricing accordingly. So when an event happens, the cost of the option
accurately reflects the market outcome.
This is potentially worrying for the FX
options market because if it gets too good at pricing events where is the value
in using the product? Traders might just as well try to jump through the window
with everyone else and take their chances there, meaning options desks will be
reduced to the much lower volume business of structuring.
Technology has been a great enabler of the
FX market and will continue to be so, however an ideal market structure
probably has a balance. Someone said to me last week their ideal trader was
like Robocop (yes that is a first mention of that film in this column), the
human brain, able to understand when things have gone too far, allied with the
best possible technology to help them action their ideas.
I think that has been a view long held by
many in the industry but due to events in all sorts of areas, mainly legal and
regulatory, we have seen an imbalance grow. To create a healthy market
ecosystem I would humbly suggest that we need to restock the human resources,
especially in the risk taking function.
Whether we can or not is an open question