HFT's mixed signals
Contrary to common thought, high-frequency
trading (HFT) isn’t on the rise, but firms must remain vigilant in arming algos
to avoid predatory gaming strategies, according to Paul Daley, managing
director and head of product development at Fox River.
Fox River runs SunGard’s algo development
team, after being acquired by the trading technology outfit in 2010, and Daley
believes the increasing chatter about HFT – which is thought to account for
upwards of 50% of US equity trading – isn’t backed by market activity.
“There’s always predatory situations you have to be
concerned about, but we haven’t seen HFT on the rise. There was so much
competition in the HFT world that its profitability has gone down on a
per-trade basis, and they’re now trying to access markets other than the US
equities,” he said, adding that executing alongside HFT firms has been an
integral part of building Fox River’s algos.
“You can never let your guard down and never stop
worrying about these things, so we spend a lot of time measuring reversion
statistics to get a sense of toxicity of displayed and dark venues to use in building
more intelligent algos,” he said.
Fox River entered the algo development
space after deciding it could build more efficient algos, according to Daley. With the recent uptake of switching engines that
choose the most optimal strategy for an order, Fox River has focused on
developing the signals used by algos with the aim of giving traders a broader execution
“Developing signals and their logic, so algos know
when to be aggressive and when to be passive or when you should be in dark
pools, is key to our strategy.
“The signals must be tailored to the individual stock
level, so algos work differently for a small-cap, low volume stock with wide
bid offer spread, than for a large cap high volume stock with a tight bid offer
spread,” Daley said.
Instead of focusing on algos that
chase benchmarks, Fox River’s trader logic technology is designed to seek and
secure the best possible price under current market conditions with the aim of
helping asset managers reduce information leakage, improve trade fill rates and
achieve best execution.
The same philosophy has been embedded in the recent
development of Fox River’s algo tailored to the exchange-traded fund (ETF)
market, to capture alpha using real-time signals.
“Our ETF algo is net asset value (NAV) sensitive, so
it knows where the NAV and the underlying is as its trading, to make that
process smarter,” Daley said, adding, “we’re using information from dark pools
to determine how we trade on lit venues and vice versa, because they are 100%
connected and have been for a long time.”
The firm has recently embarked on a move into the
Canadian market, which Daley explains was driven by clients, who wanted to use
Fox River technology to trade in US and Canadian venues.
“We’ve build algos and signals specific for the
Canadian market place, we didn’t just import them from our US markets, and
we’ve experienced significant growth because of that,” Daley said.