Skip to main content

AI And Machine Learning Can Take Fund Managers To The Next Level

The COVID-19 pandemic made 2020 a year to forget in the financial markets, leaving fund managers struggling and disappointed investors in retreat. But in 2021, with economies with some sign of recovery and interest rates still very low, new investors with an appetite for risk and high expectations are pouring capital into funds.

The six months leading up to the end of April 2021 included four of the highest ever monthly total inflows to equity funds, including a new record of £2.98billion in April, and a record total inflow to all asset classes of £6.1billion the same month, according to data from Calastone.[1] The need to try to meet those investors’ expectations means that one long-term consequence of the pandemic may be a decisive shift towards use of artificial intelligence (AI) and machine learning technologies in fund management.

Widespread adoption of AI and machine learning is a new trend: 2019 research from the Chartered Financial Analyst (CFA) Institute showed that no more than 10% of portfolio managers were using AI technologies in investment processes at that time, with most still using spreadsheet-based tools and desktop market data feeds.[2]

Francesca Campanelli
Axyon AI's Francesca Campanelli

One reason that is changing today is a growing body of evidence suggesting that AI works. A 2020 study by Cerulli of hedge funds that used AI tools revealed cumulative returns for those funds were almost three times higher than those achieved in the overall hedge fund universe between 2016 and 2019: 33.9%, compared to 12.1%. The same study also showed that European AI-led active equity funds grew at a faster rate than other active equity funds in Q1 2020; and suffered a less severe fall than other funds when the COVID-19 crisis took hold in March 2020.[3]

Why might this be the case? AI and deep machine learning technologies can help fund managers in many different ways. Deep learning solutions, based on algorithms tailored to suit specific requirements, can sift through vast quantities of structured market data and unstructured data from a huge range of other sources, searching for information related to companies, countries, the weather, political or social trends, for example. This can then drive analytics and forecasting, helping managers to identify market trends, anomalies and opportunities; to model operational, liquidity, compliance and other risks; and therefore, to have better operational and strategic decisions.

Deep learning tools can offer portfolio managers insights into the future performance of specific asset classes and asset pools, enabling analysis of the relative strengths of different drivers, risk indicators and other metrics that could produce multiple possible outcomes. They can give hedge fund managers predictions of cross-sectional performance of asset pairs over multiple time horizons.

These technologies analyse data without being guided by theories or the linear patterns that govern many investment strategies. This means they can also predict unusual events, or spot new patterns and trends that more conventional analytics would miss; giving managers extra time to prepare for and respond to those events. As well as a predictive solution, fund managers can also use AI in managing information, documents and regulatory requirements. For example, in the ESG space, several ESG scores are available based on AI technology to read text and to assign values to specific words or content. In addition, machine learning AI can streamline back office data management processes; and can be used to improve the efficiency and accuracy of some customer-facing processes, such as customer on-boarding.

Some observers have wondered whether AI-based tools could eventually conduct more strategic tasks currently performed by expert fund managers. But this would be neither desirable nor the best way to use these technologies. Instead, by hugely increasing processing and analytical capabilities, they increase a manager’s potency and reach, giving them the best possible chance of meeting and exceeding customers’ expectations.

The fundamental challenges in fund management are always the same, through economic depression, wars and pandemics: to identify and act upon risks and opportunities. Deep learning tools offer a new and more effective way to do this, making them the perfect complement to the skilled manager during these strange and chaotic times – and their value will continue to increase as they evolve in future.

Francesca Campanelli is Chief Commercial Officer at Axyon AI

[1] Calastone data: www2.calastone.com/ffimay2021

[2] CFA Institute research: www.cfainstitute.org/-/media/documents/survey/AI-Pioneers-in-Investment-Management.ashx

[3] Cerulli data www.cerulli.com/edge/global-edge


© The Sortino Group Ltd

All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or scanning or otherwise, except under the terms of the Copyright, Designs and Patents Act 1988 or under the terms of a licence issued by the Copyright Licensing Agency or other Reprographic Rights Organisation, without the written permission of the publisher. For more information about reprints from AlphaWeek, click here.