My finances, my projects, my life
March 29, 2024

Artificial intelligence: a brave new dawn for investment management

  Compiled by myLIFE team myINVEST September 26, 2019 2716

Artificial intelligence can now write your essays or press releases, diagnose your illnesses, order your shopping or even drive your car. And increasingly it is also being employed in the management of your investments. This may involve selecting the right companies in which to invest, or gathering data on companies to inform investment decision-making. It may also help in managing risk – meaning investment managers relying solely on human intelligence are under threat of being left behind.

As computing power increases, it enables the collection and interrogation of vastly greater amounts of data. A range of industries have found uses for the information processing and analysis generated by AI, and investment management is no exception.

The financial industry is already testing whether AI can be used in the selection of companies. An experiment in early 2023 conducted by financial comparison site finder.com found that a basket of stocks selected by OpenAI’s ChatGPT outperformed some of the UK’s leading equity funds. Between March 6 and April 28, a dummy portfolio of 38 stocks gained 4.9%, while the 10 most popular UK funds on trading platform Interactive Investor recorded an average loss of 0.8%.

There were some obvious limitations to the research. It spanned a relatively short period of time; ChatGPT tended to gravitate to the largest stocks; and the experiment happened to coincide with a period of stronger relative performance for large-cap equities. There are also concerns over whether an AI-generated portfolio would offer sufficient diversification. However, it showed that with some finessing, AI might have a significant role in stock selection and portfolio construction.

Some specialist companies are already emerging, using proprietary technology and machine learning techniques to analyse thousands of data points and create portfolios based on an investor’s investment objectives and risk tolerance.

Investment managers are increasingly using AI applications to help them make better decisions and analyse companies in greater depth, whether by using analysis of big data to obtain better understanding of real-time sales patterns or examining social media to assess key risks.

Understanding companies

Investment managers recognise the threat, and are increasingly using AI applications to help them make better decisions and analyse companies in greater depth, whether by using analysis of big data to obtain a better understanding of a company’s real-time sales patterns, or examining social media to assess key risks.

The world’s biggest asset manager established the BlackRock Lab for Artificial Intelligence at Palo Alto in California’s Silicon Valley in 2018, looking at ways to gather so-called big data and use it to develop effective insights capable of delivering higher returns for their clients. Schroders, Goldman Sachs and BNY Mellon also have major technology programmes examining use of the technology.

Schroders says its goal in using AI is not to build models and algorithms to trade: “We use our datasets and AI techniques to enhance our investors’ views so that they can make better investment decisions. Rather than using AI to replace people, we use it to provide an information edge in investment decisions.”

“(…) Rather than using AI to replace people, we use it to provide an information edge in investment decisions” (Schroeders)

This has become increasingly useful for investment managers as the capability of AI has increased. Previously, computers could only analyse structured data presented in a specific way. AI can now analyse unstructured data such as e-mails, text messages or voice recordings, enabling the interpretation of information from far wider range of sources, including images and speech.

Goldman Sachs says: “Access to new types of data, along with the ability to capture and process that data quickly, has given us new ways to capture investment themes such as momentum, value, profitability and sentiment.”

Does AI confer an advantage?

Should investors be looking for companies that boast AI functionality when selecting investment managers? There is relatively limited data on whether AI helps fund managers perform better, especially because the technology is progressing so rapidly.

A report from Cerulli Associates in 2020 found that hedge fund managers that leveraged AI capabilities enjoyed a competitive edge over their rivals – returning 34% over the three years to May 2020, compared with 12% for the wider global hedge fund industry. However, subsequent data has offered much less conclusive results, with many AI-driven strategies struggling in the turbulent equity and bond markets between 2020 and 2023.

As with all analysis, the insights to be gleaned are only as good as the data that goes into the model. Data that contain significant gaps are likely to be misleading and could impair investment decision-making rather than enhance it. The data need to be accurate and consistent for the analysis to be robust.

However, it is also true that investment managers that fail to exploit these techniques could be left behind. The volume of data continues to explode; Statistica predicts that it will double between 2022 and 2025 to 181 zettabytes (a zettabyte is equal to a trillion gigabytes). For context, in 2010 only two zettabytes worth of data existed.

As data insights become more widely available, investment management groups that rely on human expertise alone may struggle. Luke Ellis, CEO of UK hedge fund manager Man Group, believes that failure to adopt quantitative approaches using big data could prove fatal for fund groups: “If you don’t understand how to treat data with respect, you will get eaten alive.”

Co-operation with data scientists

The solution lies in ensuring that investment managers and data scientists work together closely and harmoniously. Investment teams need to identify where data scientists should focus their research. Schroders argues that if each group is left to its own devices, data scientists won’t necessarily ask the right questions, while the investment team will fail to understand how the technology works.

Effective data science is set to unearth insight that is unlikely to have been captured previously by others.

Effective data science is set to unearth insight that is unlikely to have been captured previously by others. The greater the quantity of data that may be relevant to understanding a company or an investment, the more combinations and permutations of analysis it becomes possible to conduct.

Does artificial intelligence represent a brave new world for investment management? It is certainly an important development. Groups that can strike the right synthesis between data science and established portfolio management skills could have a significant advantage. However, in a worst-case scenario, management could become a slave to historical data and trends rather than analysing what businesses are doing now. Humans and robots need each other more than each might think.