Coleman Capital, a private investment boutique offering hedge funds and real estate investments for institutional, family office and other professional investors, has engaged Swedish Artificial Intelligence (AI) provider Batonics to research and potentially develop an AI engine to assist with risk management within its hedge fund business.
Initial discussions have focused on Coleman’s event-driven and credit long/short strategies. Within event-driven, hedge funds typically aim to capitalise on temporary stock mispricing that may occur before or after a corporate event. Merger arbitrage, for example, seeks to profit from price movements caused by announced mergers.
“We want AI to help with early signals for when M&A deals are about to break in the Asian market”, said Ernesto Becker, hedge fund COO, who built out the firm’s trading and risk management infrastructure, noting that algorithms are already in use, but the predictive power of AI is still untested.
Edwin Garcia, portfolio strategist, said: “We have no pre-set conclusions on what AI can do but want to investigate if and the extent to which AI can offer predictive power into future price movements within certain strategies”.
Gholam Bakhtiar, managing partner at Coleman’s real estate business, said the journey with Batonics started at the end of 2023 to see what they could offer and what they could do on the real estate side as that is where they had the longest track record. “We are looking at how their data points can enhance risk management within our portfolios. Then we have to illustrate to our portfolio managers how AI predictions can add value,” he added.
A recent report by the Boston Consulting Group (BCG), the strategic consulting firm, shows that while the asset management industry is enthusiastic about the potential of AI, few have implemented it at a company-wide level at all business levels. Bakhtiar agrees that at this point AI in asset management is about data optimisation rather than portfolio construction.
Becker said the ideal outcome of AI in investments is its ability to create a level playing field where everyone has the same information. “But if you can get information that nobody else has, or get it first, you will have an advantage,” adding that the biggest players may already be far ahead in their use of AI but keep quiet to hold onto their advantage.
When it comes to cybersecurity and AI, Becker is more concerned with risks at the underlying data providers than AI companies themselves. “If this data is hacked and manipulated it is of concern,” he said. He noted that a downside with cybersecurity concerns regarding AI is that “people want to control AI so much that it is no longer intelligent.” He questioned whether these concerns are based on the fear of being replaced or simply not trusting the data enough to use AI.
He added that Coleman would never use AI to execute trades, a portfolio manager’s role, even in the future. Becker said that at the moment AI is about factor analysis and “one of the factors is the portfolio manager.”
Becker explained that AI can be used to analyse the portfolio managers performance and not just the performance of their strategy. “If we see patterns where a portfolio manager underperforms during a specific time of the day, perhaps they should not be at their desks during that time and rather go to the gym instead, which I am sure they would not object to.”
Bakhtiar said the plan is to expand the use of AI tools in Coleman’s other portfolios beyond Asian M&A and credit long/short, but we need to see what they can do first.
Coleman’s hedge fund business includes, apart from event-driven strategies, credit long/short and equity long/short strategies. Credit long/short strategies aim to profit from taking long and short positions in credit securities such as bonds, loans or credit derivatives. Long positions are used when a specific instrument is expected to increase in value. Short positions are when a fund borrows credit instruments it expects to decrease in value, sell them, and then buy them back at a lower price to return to the lender and pocketing the difference. Similarly, equity long/short hedge funds take both long and short positions in stocks to capitalise on stocks that are undervalued (long) and overvalued (short).
Becker explained that just like their peers, AI tools will be run in a sandbox with shadow portfolios before any of them go live.
He does not see the future of AI as increasing the herd-mentality of investment management because the output into AI depends on the input and underlying data which can differ vastly. “Beyond the current uses of data optimisation, factor analysis and enhancing human capabilities by increasing efficiency, a question mark remains around whether or not AI in itself is capable of generating alpha”.
“I am pretty sure that quite a few systematic investors, often referred to as quant managers, have already integrated AI into their investment processes, where AI is used to understand market trends and optimise portfolio allocation, and a majority in the industry expect AI to match or exceed traditional investment analysis within a decade”, Becker said.
He noted that the AI optimists see AI as a tool to move us to a new era of prosperity quicker whereas the AI doomsayers say it is moving too quickly. “I see it firstly, as a tool for gaining insights, secondly, potentially revolutionising investments by evaluating historical trends and predicting outcomes before you buy or sell and thirdly, enhancing risk calculations,” he said, adding that in the future the intersection between human talent and technology will remain even when AI improves and learns on its own without much human involvement.
Coleman Capital was founded in 2014 by Matt McClean after senior appointments at Morgan Stanley & KCG Europe. He wanted to create an alternative investment boutique with top talent and Ernesto Becker was one of the first hires and the infrastructure he built is still in use today.
Garcia said currently there are 12 professionals employed at Coleman with the majority of clients coming from the UK, US and South Africa because of the contact network of the senior executives. “We expect to expand to Continental Europe and Asia,” he concluded.
By Pirkko Juntunen
Twingly offers News Data APIs for the financial industry, providing over 3 million daily news articles from 170,000 active global sources. It gives you access to noise-free news data for timely updates. Additionally, we provide data from financial forums through our Forum API, featuring over 10 million forum posts per day.
