HFT tech being transformed to provide new era analytics

Technology that was developed in order to meet demand for ever-lower latency as high-frequency trading took off is now being used to fuel new forms of analytics, according to technology experts.

By None

Technology that was developed in order to meet demand for ever-lower latency as high-frequency trading (HFT) took off is now being used to fuel new forms of analytics, according to technology experts.

The arms race for ultra-low latency technology is largely coming to and end, says Donal Byrne, CEO of network data specialist Corvil, but this technology is now being re-developed to provide new real-time services increasingly demanded by market participants including the buy-side.

“There has been a big shift away from focusing on speed of execution. Now trading technology development is much more about not getting caught out and safeguarding yourself,” explained Byrne. “This is fuelling the development of new IT infrastructure and new paradigms in the sector.”

The applications from this are quite wide, covering market surveillance, risk management and transaction cost analysis.

Spencer Mindlin, an analyst at consultancy Aite Group, added, “HFT is not generating the same returns it used to and the market has become crowded. This means the technology developers that previously focused on HFT are now looking at using some of the technologies, such as complex event processing, to give technologically savvy market participants a new type of competitive advantage.”

New regulations are also increasing the need for rapid processing of information.

“Rules covering risk management and cyber security mean it’s more important than ever to have real-time systems that can process information very quickly in place,” said Byrne.

“Regulators want firms to have a full audit trail in place and part of that is having proper event sequences in place, but with vast amounts of data being transferred in microseconds you need very sophisticated technology to get your timestamps right.”

Buy-side decision support

For the buy-side, perhaps the most interesting development in this space is the potential to integrate new datasets to help inform trading and investment decisions.

“There are a lot of startup firms around today that provide data from social media and sentiment analysis and offer those up as APIs for trading desks. They can then also integrate these with technologies that can process their own databases to develop new trading ideas internally,” explained Mindlin.

However, as with most emerging technologies, this kind of complex data analysis can come with a hefty price tag. Low-latency trading technology has become largely commoditised, lessening some of the advantages that HFT firms benefited from when it was relatively new and expensive, and the same could happen with these developing technologies.

“This could be something of a gold rush, and you have to wonder whether it’s better to be selling the shovels of digging for gold,” Mindlin added.

“A lot of players are springing up in this space right now but probably only a few will be able to deliver alpha for the trading desk so the buy-side needs to think carefully about whether these services really complement their strategy.”

However, he adds that using real-time analytics for risk management is far more likely to prove to be a long-term investment that will provide real benefits for institutional investors.

By being able to monitor activity in real-time, investors will be able to more effectively respond to potential risks such as rogue algorithms, fat finger trades and market glitches. Additionally, providing accurate timestamps for all events will give them the ability to fully audit and investigate problems, both to satisfy regulators and to prevent similar risks developing in the future.

«