Skip to content

‘Computers don’t go to drinks parties’: Where next for Luke Ellis’s quant quest?

Luke Ellis has just finished fielding an intense Q&A session when he sits down with Citywire Selector. The Man Group CEO was giving a virtual presentation to schoolchildren in East London which sparked considerable interest.

The discussion focused on children interested in pursuing a career in either maths or technology who might never have considered quantitative investing as a viable route.

Given that Ellis has grown Man Group and its five investment arms into a collective entity with $151bn (€143.6bn) in assets, up from $52bn almost a decade ago, there is definitely a way forward for systematic thinking.

Man Group is often prefaced with ‘hedge fund’, but Ellis is a strong advocate of the company’s huge investment in and application of technology, as well as the ability to plot a firm course ahead using cutting-edge techniques.

‘The great thing about computers is that if there is a Federal Reserve meeting this afternoon, let’s say, I can tell you with 100% certainty what all our computers will do in terms of the next decision.

‘Now, this is probably going to get me in trouble, but my wife is currently visiting Japan. I will call her at the end of her day there and I have no idea what reception I am going to get – with a human it all depends what has happened during the day, or what their mood is.

‘So humans, even the ones you know best, can be unpredictable. Of course, in the middle of a conversation you get a sense of how a human is acting or going to react, but that’s not concrete. I think there is a feeling that some people are scared of computers and, in fact, that is probably why we can make good returns. Because if everyone was comfortable with them, we’d have more competition.’

Ellis likened the situation to the growth of driverless cars, which have had similar pushback from those who think human drivers are more reliable. ‘You have over a million people die on the world’s roads every year, but a very small fraction related to driverless car accidents and yet that’s often the focus.

‘It’s strange that people think you should let the computer drive the car until there is an accident or issue, at which point the human should leap in and intervene. That’s the exact point where you don’t want that.

‘We have been biologically trained over thousands of years of evolution to have a fight-or-flight reaction – a panic – in these situations, when emotion actually needs to be taken out.’

The shorter-term focus of investing is also a challenge, Ellis said. With fund managers able to gauge their performance in real time, they can be more prone to make emotionally-driven decisions when it would be better to simply leave the position to play out.

‘What I would say is this – computers don’t go to drinks parties; meaning there is there is no innate need to show off, or have something interesting to talk about. So many people lost money in crypto because they weren’t invested early enough and felt they were losing out. They went to social events and heard all this talk and then went and bought some. Now they are all in trouble.’

Information in focus

Ellis’s affinity for computers is perhaps reflected in his straight-talking approach. During Citywire’s CEO Summit 2022, he pressed his peers from rival asset managers to give straight answers, he also praised his colleague Russell Korgaonkar’s recent assertion that more investors should admit when they don’t know something – a move he prefers to ‘making up a bullshit answer’.

This aligns with Ellis’s belief that information is paramount, but not all information is useful. He highlighted the meme-stock rally as something that was happening in plain sight but perhaps wasn’t fully understood, despite wide reporting and obvious investor interest.

‘There are always new developments you should be looking to take advantage of, some of which will end up in permanent change,’ he said. ‘We have been trading momentum for 35 years and the way we do it today is more sophisticated but the fundamental set-up is the same as it was when we started. There is this belief that as soon as a piece of information is out there, then it is discounted by the market.

‘Information today can be perfectly disseminated but the speed of decision making is very varied, particularly among institutional investors. Even if you knew with 100% certainty the next two years would be a long and grim recession, for example, it could take the average pension fund a long time to change course.’

This is where the idea of ​​concentrating on ‘useful’ information is important. Reverting to the meme-stock rally, Ellis highlighted how effectively six or seven stocks in a 10,000-stock universe were having differently to how they were expected to.

‘An investment model would single those out – and potentially even seek to trade them – but the key questions are: where are they and where are they going to be? This is where Man Group can tap into its technology and seek to uncover something that others might not be seeing.’

word searches

For instance, it filtered thousands of comments on Reddit through natural language processing to spot common trends and sentiment. This can be extrapolated to other tasks, such as deciphering transcripts of earnings calls and highlights Man Group’s quant emphasis.

‘The balance we have as a company is very much tilted towards quant compared to the average. That’s a conscious bet. I don’t think there is any industry in the world where technology is not having an increasing impact, so why would it not be true in investing? There are, of course, things that humans do that computers cannot, but the job is to ensure the humans are focused on those bits, then they can add value.

‘It’s a different kind of skill that humans have; that’s why we’ve got a discretionary business too. There’s clearly value in harnessing the two elements, we’re seeing the quants and technologists learn from the discretionary PMs, and vice versa.

‘On the human skills side, our discretionary PMs can go and interact with management teams and get a sense, for example, of whether a CEO really understands if they have a killer product with the power to revolutionize their market segment. A computer wouldn’t naturally identify that.

‘We see this human element as key to our credit business too; we’ve recruited some credit PMs into GLG over the last few years and we’re keen to develop this. That’s down to the sense that we’re in an environment where selectivity and name selection are key, and not just an environment where you want to own the entire market.’

high ambitions

Alternatives is a recurrent theme of CEO ambitions at the moment, as firms seek to diversify their business model while unlocking new sources of alpha. Man Group’s Q1 2022 results showed a huge driver of the $3.1bn of net inflows was drawn from its alternatives arm, while its long-only franchise ended the three-month period in negative territory.

‘I think to a degree that is driven by the market environment. We have had a period of unbelievably low economic volatility, both within a particular economy and between economies. Inflation has been low but is now climbing, fiscal and monetary policy was basically the same everywhere until now. Now we are entering the sixth month of what is arguably extreme economic volatility.

There are large parts of the asset management industry – the active asset management industry – that don’t generate alpha

‘It’s extreme compared to the last 20 years, because – if you take out 2008 – everything was more aligned between economies. In my own view, I think we are facing an extended period of years, not weeks, of that scenario. It doesn’t mean markets will go down for years, but it does mean that taking a simple portfolio and just sitting on it might not make as much sense.’

Ellis said the demand for alternatives has grown as a result, as investors are worried about ‘flat-lining’ equities, but want to remain active. However, he said this would prompt some unnecessary rush to release more products via the AHL division or even in an alternative format through GLG.

‘I suspect there will be more demand for alternatives, yes, because they have done well. They preserved client capital and they made some money but as a client – ​​and as a provider – you need to think long term. Everyone buys insurance after they’ve been burgled. So demand could stay but we will stick with what you have generated alpha, and generated it at scale.’

The industry

This question rolls into another common theme dogging the industry at the moment: consolidation. With many listed asset manages trading at low levels while posting negative performance, there is a growing sense that activists or merger specialists could be circulating.

‘There are large parts of the asset management industry – the active asset management industry – that don’t generate alpha on average over time, so you have to question if they are serving a useful purpose,’ he said.

‘The industry has a massive number of players and is very fragmented, with little-to-no barriers to entry. If you read any business school book about market segments with big fragmentation and no barriers to entry, that industry would typically have low margins. Yet, asset management in good times – which is an important caveat – is a profitable industry, with 30% margins.

‘Now, with the market environment we have seen – markets being down 20% and redemptions growing – that margin is under pressure. That margin will disappear in a hurry, and the truth of asset management is: there are too many asset managers that don’t generate value for clients. It is absolutely right that those players get consolidated out of existence.’

So, where does this leave Man Group? Ellis said a lot of the merger and bolt-on discussions are being done with an eye on becoming a ‘mega player’ where size dominates all and investors can access a true one-stop shop.

Leave a Reply

Your email address will not be published.