Skip to content

How AI and big data are reshaping asset management

How is the integration into the investment process of artificial intelligence (AI) and big data transforming asset management?

  1. AI now forms the bedrock for analysis in many areas in society, such as healthcare and weather, and now that technology is being employed in asset management.
  2. The technology could be used to modernise areas such as portfolio management, risk management and trading, enabling asset managers to analyse broader data sets more deeply.
  3. Evidence shows that AI is a very useful tool in portfolio management, but one that requires close human supervision.

For more data-driven insights in your Inbox, subscribe to the Refinitiv Perspectives weekly newsletter.

Analytical techniques for everything from breast cancer screens to weather forecasts have been revolutionised by artificial intelligence and big data.

Now the backers of some AI businesses believe that the time has come for the technology to modernise asset management, often by analysing the oceans of new data being generated by smartphones, the internet, satellites and other innovations.

Project SentiMine: surfacing equity performance themes in unstructured content

Applying AI and big data to asset management

In a dystopian vision of asset management’s future, AI robots take over almost everything, from portfolio management to back office administration. Certainly, some hedge funds and ETFs have built AI statistical models for trading. Their managers believe that portfolio management is ripe to be modernised by applying techniques that analyse big data in a more dynamic way, while removing emotion from the investment process.

There is no doubt that AI has growing applications for portfolio management, risk management and trading – not to speak of back office administration.

A sizeable number of asset management companies are now using AI and statistical models to run trading and investment platforms, according to a CFA study. Yet it says that the models are generally controlled by some kind of human supervision.

Mixed investment performance

According to a study by the UK’s Alan Turing Institute, approximately 9 percent of hedge funds use AI and machine learning to build statistical models for stock analysis. It is used in a variety of ways to develop novel investment strategies, including through analysis of broader and deeper sets of data, analysis of social media and crowdsourced forecasts.

It may also help to improve the shortcomings of classical portfolio construction techniques.

However, there is mixed evidence that AI has helped investment performance. In a report published in 2020, the consulting and research firm Cerulli Associates reportedly found that hedge funds with AI capabilities had a big competitive edge over others.

AI-led hedge funds reportedly generated average returns of 34 percent in the three years to May 2020, according to the report, compared with a 12 percent gain for the overall global hedge fund industry in the same period.

Yet at the same time, there were news reports of some of the biggest and previously best-performing AI hedge funds struggling in 2020’s choppy markets, resulting in large losses.

The EurekaHedge AI Hedge Fund Index returned 11.24 percent, compared with 12.68 percent for its index covering the entire hedge fund universe.

Read the report: The rise of the data scientist

From risk management to the back office

Apart from portfolio analysis, trading is a particularly productive area for AI. It can be used to automatically identify the best times, size and venues for placing trades.

Similarly, it is being used to counter some of the shortcomings in traditional risk management, highlighted by the precipitous fall in equity prices in March 2020. The month saw one of the most dramatic crashes in history, as the U.S. S&P 500 stock index fell by more than 30 percent and other asset classes also fell.

AI approaches are helping to refine risk management by validating and back testing risk models. They can also extract data efficiently to generate accurate forecasts of bankruptcy and credit risk, market volatility, macroeconomic trends and financial crises.

In the back office, there is increasing potential for using AI and automation, as downward pressure on fees and the growing complexity of products, legal entities, vehicles and markets are increasing the need for economies of scale.

There is vast scope to automate – through both AI and robotic process automation. Both asset managers themselves and their administrators are investing in automation, seeking to cut costs, reduce errors and build more capacity so that they can handle peaks in workloads.

Watch: COVID-19 accelerates AI and Machine Learning in finance – A Refinitiv Data Moment

Why AI technology is a powerful tool

While the back office is one of the areas where the benefits of AI and automation are most certain, in portfolio management it seems that AI technology is only as good as the hands it is in. There is no doubting its usefulness in data analysis, risk management and trading, arising from efficiency, objectivity and the ability to self-improve that is key to machine learning.

Yet it’s said that AI will always generate a result, even when one should not exist. This tendency can cause problems, meaning that it is best seen as a tool for now – a powerful tool, but one that needs close supervision.

Project SentiMine: surfacing equity performance themes in unstructured content