Mirae Asset Management has launched a pair of artificial intelligence (AI)-driven funds in South Korea that will invest in various smart beta index exchange traded funds (ETFs), in a first for the fund manager firm.

The Mirae Asset AI Smart Beta Fund and Mirae Asset AI Smart Beta Market Hedge Fund will invest primarily in KRX-listed ETFs, and the respective portfolios are formed entirely by the AI with a final call from the fund's manager, a Mirae spokeswoman explained.

"The AI takes in various factors that are known to affect stock prices and calculates the allocation weights among 'Smart Beta ETFs' (such as momentum, low-vol, contrarian and high-dividend), some market index ETFs and a few selected large-cap stocks," said the Mirae official. "Once the AI comes with a portfolio, the manager decides the final call on execution. The AI then further adjusts the portfolio via deep learning process."

The AI algorithms are developed and managed by the Mirae Asset AI Financial Research Center, which was established jointly by Mire Asset Global Investments, QRAFT Technologies, and Complex Data Analytics Lab at Korea University.

Mirae plans to "keep a close eye" on the AI and fund performance for the time being until the funds acquire a satisfactory track record before further initiatives are developed with the proprietary algorithms. "Since Mirae Asset AI Smart Beta fund series is our first AI-backed fund, our priority remains on keeping good and stable track records. Once we achieve this goal, we would then likely expand to other products and services," said the Mirae official.

The funds are developed primarily for the South Korean market, with all support including factsheets, customer service will be provided in Korean, but foreign buyers who still wish to purchase the fund may do so via official sellers.

In a separate development, Nomura Securities announced that it will begin publishing a new AI-related equity index.

The Nomura AI Companies 70 is an equal-weight index composing 70 stocks of companies covered by the media in connection with business related to artificial intelligence. The component stocks are selected from the top 98% free float-adjusted market cap from all common stocks listed on Japanese stock exchanges, with quantitative valuation indicators based on AI-related keyword searches for articles and reports from TV news, magazines, newspapers and other media.

Likely AI-related keywords include deep learning, machine learning, big data, Internet of Things (IoT), robotics, data mining, sensing, cyber security, robot, and autonomous driving, the company noted. The indices have been constructed by Nomura Securities' Quantitative Research Department and calculated by the Index Operating Department.

Patrick Armstrong, senior risk analysis officer, innovation and products team at the European Securities and Markets Authority (Esma), said in a speech last week that AI could impact 'all kinds of financial activities and products'.

'What is changing the sector today are three factors: first, the volume of data available to support decision making; second, the fall in the cost of computing power and storage; and third, algorithms have been developed to leverage the data,' said Armstrong. 'The use of AI in quant investing is different from the earlier generation of factor investing insofar as today's AI program seeks to use that data to improve its decision-making skills over time.'

Relates stories:
AI could impact all kinds of financial activities and products, Esma

Index round-up: FTSE Russell broadens China portfolio with two new indexes