Digital Assets Report


Like this article?

Sign up to our free newsletter

The onward march of the machines

Related Topics

A revolution in technology is transforming the way hedge funds operate, offering managers advantages – as well as challenges – in tumultuous times.

A revolution in technology is transforming the way hedge funds operate, offering managers advantages – as well as challenges – in tumultuous times.

The hedge fund sector has undergone a “technological revolution” during the past decade, which has disrupted and reshaped an assortment of areas and business functions, according to market participants, with the pace of change continuing to accelerate. 

Recent years have been underpinned by “huge leaps forward” in the volume and range of data used by hedge funds and the availability and flexibility of technology infrastructure within firms, says Alex Nixon, chief technology officer at Maniyar Capital, the London-based macro-focused firm established by ex-Brevan Howard Asset Management and Tudor Investment Corporation manager Dharmesh Maniyar. 

“In the past, successful asset managers, and hedge funds managers in particular, needed to invest in significant resources across the firm to acquire data and operate large middle and back-office teams,” says Nixon. 

“Over the last 5 – 10 years high-end computing power has become significantly cheaper and more accessible. Additionally, with the growth in quality of third-party cloud-based services, managers can now outsource many functions and run leaner and more agile teams. This incredible industry transition has levelled the investment field and the advantages enjoyed by the super large hedge funds can now be enjoyed by even mid-sized firms.” 

At the same time, this tech revolution has also heralded a rapid expansion in the number, availability and flexibility of different techniques for signal processing, machine learning and AI that help fuel quantitative hedge fund strategies’ investment processes. 

“Computer-based strategies benefit from the strength computers can bring – they can compute predefined indicators on a large amount of data, optimise some criteria and do all this systematically, with no errors,” observes Nicolas Guassel, CEO and co-CIO of Paris-based Metori Capital Management. 


However, he concedes that systematic hedge funds suffer from “the intrinsic limits of computers.” “No computer has ever created something new. The conception of an algorithm remains something which requires a lot of practical experience and creativity – a typical human quality,” he tells Hedgeweek. 

“Practically, on liquid markets, where data is available and public, computer-based strategies might have an edge. On the opposite, it remains very difficult to ‘automatise’ strategies based on less quantified indicators, such as concentrated stock picking or private debt investing.” 

That potential advantage is reflected in recent industry performance stats, which indicate that hedge funds which use a machine-based investment approach have outflanked those firms with more traditional discretionary trading styles in the past year, reversing performance trends of recent years. 

According to Barclays’ 2022 Global Hedge Fund Industry Outlook, quantitative equity managers, for instance, advanced 11 per cent on average in 2021, while computer-based CTAs rose 9 per cent last year. 

By comparison, traditional discretionary long/short equity hedge funds gained 10 per cent last year, while discretionary macro funds were up just 1 per cent. Delving deeper into the numbers, quant managers are also outflanking their discretionary counterparts when it comes to generating alpha – a core barometer of success for the hedge fund industry. 

Annualised alpha among quant equity funds topped 9.1 per cent in 2021, with a Sharpe ratio of 2.2 thanks to lower volatility, compared to discretionary equity managers’ -0.3 per cent annualised alpha and 1.0 Sharpe ratio the same year.

Reflecting on the relative merits of a computer-based investment approach, Laurent Laloux, chief product officer at CFM, explains how quantitative investment managers have the ability to amass and aggregate a large quantity of different sources of data and alpha, which in turn potentially gives them the edge against competitors. 

“The good thing about being systematic and automated is that you remove a lot of the emotion from the process,” Laloux tells Hedgeweek. 

“From there, you can then really focus on assessing the quality of data, and risk-controlling what you are doing.” 

He continues: “It’s not because you crunch a lot of data and you use a computer that you’re necessarily any better at predicting alpha. It’s an investment style like any other, so you should be reasonable about it and not have your hopes up to the sky. 

“But instead of having just one, or a few handfuls of bets like a human being, you can have thousands and thousands of tiny bets which on their own may be barely statistically relevant, but by the sheer force of statistical aggregation, you get something which is robust and works. This is really the job we are doing – we are, if you want, a vacuum cleaner of tiny alpha remaining everywhere. But because of the scale of what we do, we get something which is really relevant in the end.” 

Quant advocates readily admit that for all the state-of-the-art technology underpinning their investment processes, the traditional reliance on historic trading patterns means that rapid changes in market regimes and investor sentiment can – and do – send systematic strategies into a tailspin.

It’s a question that has come into ever-sharper focus given the current backdrop of geopolitical turmoil, soaring commodities prices, shifting inflationary patterns and investment uncertainty. 


To counter such potential risks, managers such as Aspect Capital, the London-based quantitative hedge fund pioneer which runs a range of systematic funds across the managed futures and global macro spectrum, utilise combinations of models that trade with and counter to the prevailing trends within certain asset classes. 

“Often, in the broad scheme of things over the order of months, when everyone gets too bullish it can be a very strong sell signal; when everyone gets too panicky, it’s usually a good time to buy,” says Razvan Remsing, Aspect’s director of investment solutions, adding that despite the rise of alternative data in recent years, there is still considerable merit in long-established systematic models utilising traditional datasets that have stood the test of time. 

“Last year, for example, some alternative datasets actually had very little traction in being able to navigate the way the Fed was trying to message that inflation is transitory. Every few weeks sentiment would swing from one side to the other – it’s not always the case that more frequent sentiment information gives you a better quality of information. You have to combine that with models that look at what’s actually priced in, via traditional sets.” 

He adds: “Traditional data forms the backbone of our strategies, and it is quite likely still the mainstay of most significant managers out there. The reason is that it’s the most reliable, and accurate, and irrefutable data.” Laloux says: “We know that when there is a big regime change – it could be the Trump election, or Covid, or unfortunately a war in Europe – there’s always a risk that you have taken on the wrong side of things. 

But that’s also the case for human beings. 

“As a quant, you know that your system will take a little bit of time to adapt to the new environment, and recover potential losses you may have suffered. Discretionary managers might be faster to react in a regime shift, and are able to assess, based on hypothesis, what should be the right way forward.” 

According to an industry survey carried out by Hedgeweek for this report, just 13 per cent of respondents believe hedge funds have fully captured the potential opportunity provided by new technology, including software, compared with some 60 per cent who said they have not, with close to 27 per cent still unsure. 

One sector rapidly opening up to new opportunities in systematic trading is emerging markets, says Paulo Remião, partner and portfolio manager at London-based Broad Reach Investment Management. 

Launched by former Spinnaker Capital manager Bradley Wickens in 2016, emerging markets-focused Broad Reach is “at the frontier” of applying a fully systematic process to emerging markets, explains Remião. 

“20 years ago, emerging markets had some credit trading, some FX trading, and a relatively limited number of countries. Now, emerging markets have increased massively in breadth, in terms of the number of countries that can be accessed, and the number of instruments available across countries. You can also trade credit, FX, rates, equities, commodities,” he notes. 


“One way that we figured out how we can monitor all the opportunities that exist across all these countries and all these asset classes is to have a much more systematic approach to the way we deal with data and the investment process.” 

As artificial intelligence and machine learning functions continue to spread into all corners of financial markets and various points of the investment process, fund managers acknowledge the increased risk of events similar to the 2015 ‘flash crash’, as well as other assorted code blips, which in turn could risk renewed episodic and unpredictable bouts of volatility. 

“Behind the machines, humans are coding, and people are likely to think in the same way,” Julien Messias, co-founder and head of R&D at Paris-based Quantology Capital Management tells Hedgeweek. 

To address this, Messias, whose firm manages several systematic equity strategies, urges quant-based managers to carve out a unique approach and, ultimately, “think by yourself”. 

“We are convinced by the advantage of systematic, ruled-based strategies over discretionary ‘human’ approach. This difference occurs mainly during fast markets, in which humans see their rationality overcome by their feelings – anger, fear, greed. This is when they commit errors – and big ones,” Messias says. 

“Try not to copy off your neighbour – AI and machine learning being more prevalent means people using the same techniques which means more crowded trades. Try to build in as autonomous a way as you can your processes, and your strategies, from A to Z.”

Read the full A Tech Revolution: How machines are reshaping hedge fund investment Insight Report here.

Like this article? Sign up to our free newsletter

Most Popular

Further Reading