When discussing the future of the FX industry finding consensus amongst
market participants about what the market will look like and how it will
function can be challenging.
Yet one thing that appears to be broadly agreed upon is that the use of
algorithms for executing trades is likely to continue growing in the coming
years, as technology continues to evolve and firms look for new ways to
minimise their market impact when trading.
Indeed, the use of algos is often prescribed as the answer to a market
where it is becoming harder to execute in size and buy side firms are
increasingly concerned about this issue of market impact.
Despite the expectations of the market though, there are some potential
headwinds for the growth of algo trading amongst buy side firms. One potential
issue is that there is currently no real methodology for comparing the
effectiveness of the algos being provided to these firms, an issue that is
compounded when some algo providers are offering a large number of very similar
Part of this becomes a chicken and egg problem as algo adoption could be
hampered by a lack of tangible evidence and comparability between algos but
without significant adoption there won’t be enough data to conclusively prove
the value of one algo versus another.
As noted in the latest edition of Profit
& Loss magazine, another potential concern for buy side firms looking
to use algos to trade FX is whether there could be a conflict of interest for
principal trading banks to provide them with algos.
The concern here is that one bank now knows that firms’ entire order and
there exists the potential for information leakage. One buy side source
corroborates this concern, claiming that during a review of bank provided algos
two years ago they found that at a “surprising” number of banks the principal
desk was able to see the algo data come through. Although this may have changed
since, it is anecdotes such as this that could make some buy side firms
reluctant to use bank provided algos.
In addition to all this, there are the obvious anxieties about “rogue
algos”, with the experience of Knight Capital remaining as a case in point
about why such concerns might be valid.
Therefore, if algo usage is going to increase, the FX industry may need
to think about introducing kill switches to prevent a runaway algo from causing
too much damage. A key issue then becomes deciding where these kill switches
should be placed.
The issue of algo adoption in FX will be the subject of the second panel
at Forex Network Chicago on September 28th. To find out more
information about the panel, Algos: Silver Bullet or Accident Waiting to
Happen? Click here.