Are pure high-frequency trading firms quietly dying?

High-frequency trading firms have spent the past decade growing from industry challengers to industry stalwarts, with the top firms having thousands of employees and widely coveted graduate and internship programs. However, they have a problem: HFT infrastructure is very expensive. Firms are therefore taking inspiration from their hedge fund competitors and slowing their strategies down.

Click here to follow our new WhatsApp channel, and get instant news updates straight to your phone 📱.

"There's quite literally hundreds of firms that do high-frequency trading to some capacity," one Hong Kong based quant headhunter, speaking anonymously, told us. "There's also an increasing number of firms opting for slower frequency trading strategies." He says they're doing so because it's becoming "increasingly costly for diminishing returns when trying to reduce latency" due to an excess of firms all pursuing similar simple low-latency strategies.

A recent study from Beijing Shaotong University tested high and low-frequency algorithms on historic data of the Shanghai Stock Exchange. It found that the HFT algorithms had an annual return of 28.5% compared to 22.7% for low frequency, but found that "sophisticated infrastructure" costs hampered long-term comparative success. The study advocated using a mixture of both algorithms instead.

A London headhunter, focused on the buildout of new teams, told us that many of the major firms have been building out mid-frequency trading (MFT) capabilities for a while. "They've made a lot of money, have excess capital sitting around, and are starting to spend it." Conversely, hedge funds have also been hiring more people from HFT to "increase their capability and execute faster," but most traditional funds aren't at "true" high-frequency yet.

Both industries have a reputation for eye-watering pay, so the headhunter says people "aren't necessarily moving for the money." Instead, the appeal is "building something new. It's appealing because it's challenging, easy to measure, and you have ownership of it." It's usually harder for someone in HFT to transition to a high-frequency team in a hedge fund: "they're more regulated and less nimble, so getting stuff built is harder."

One of the most notable algorithmic trading firms trading the lower frequencies is Jane Street. Jane Street has been known to hold positions for days at a time. Hudson River Trading also does some work in long-frequency trading as well as mid; HRT hired Max de Werd from Citadel in 2023 to work on both strategies across the FICC and macro spaces. It has been operating in mid-frequency commodities since at least 2020, hiring interns for the team.

For the HFT firms with established MFT teams, the new fascination is deep learning; the London headhunter says "really successful firms are making the most PnL by applying machine learning to trading." Jane Street and other major players like XTX Markets have been building out their machine-learning capabilities, while even more niche HFT firms like Radix Trading assert that "trading based on sheer speed of execution might have reached its limit."

The shifting focus to lower latency trades means a different sort of skillset is required. High frequency trading platforms are typically coded in C++, widely recognized as the best programming language for speed. MFT may open up the door to using other languages like Java and Rust, but the headhunter says "each firm has its preferred language of use" and adding MFT strategies likely won't change that. MFT is also a lot easier to do than HFT in Python, a language loved by hedge funds, especially as AI increases its prominence.

The Hong Kong headhunter said that hedge funds' interest in HFT varies on how much they're willing to invest. They have to "overcome the incremental costs associated with adapting infrastructure and hiring talent to build out varied strategies in that space." It's not uncommon for firms to get cold feet; low-latency FPGA teams are thought to be particularly susceptible to being dismantled as costs rise.

As the scope of HFT firms expands beyond pure latency, it's hard to argue that many of the top firms today are still 'pure' HFTs; time will tell if another major HFT will arise in what is now an oversaturated market.

Have a confidential story, tip, or comment you’d like to share? Contact: Telegram: @AlexMcMurray, WhatsApp: (+1 269 237 3950). Signal: @AlexMcMurrayEFC.88 Click here to fill in our anonymous form, or email editortips@efinancialcareers.com.

Bear with us if you leave a comment at the bottom of this article: all our comments are moderated by human beings. Sometimes these humans might be asleep, or away from their desks, so it may take a while for your comment to appear. Eventually it will – unless it’s offensive or libellous (in which case it won’t.)