Wealth through Investing

Big Data and Artificial Intelligence in Investment Management: FAQs and Answers


A technological advance can impact — even transform — the global economy let alone a particular sector.

The steam engine in the first industrial revolution, electricity in the second, and internet technology in the third fundamentally changed human history.  Today, big data and artificial intelligence (AI) have the same transformative potential.

When it comes to big data and AI in investment management, some people anticipate a rosy future with various new sources of alpha. Others worry about the jobs that might be lost to machines through process automation.

Regarding the new sources of alpha, developments in this fourth industrial revolution raise a series of critical questions. How investment professionals address them will go a long way in determining who will successfully adapt and who might be rendered obsolete.

Frequently Asked Questions (FAQs)

1. By using big data and AI, can investment managers achieve competitive advantages in securities selection and asset allocation?

It depends on the investment time horizon. With short-term high-frequency trading (HFT), for instance, investment opportunities — the short-term noise rather than economic added value or mis-pricing due to structured behavioral tendencies — could be arbitraged away. On the contrary, long-term investments, such as environmental, social, and governance (ESG) factors, engagement funds, and private equity funds will still require interactions with company management. Additionally, each potential investment that is not backed by big data that AI can sift through will require investors to conduct a more granular and hands-on analysis. In this sense, there could be opportunities in public markets over a mid-term investment horizon.*

2. If AI is the best tool to scour big data for sources of competitive advantage, will these advantages be limited in scale and sustainability? Or will they be larger and more enduring?

Mid-term investments in public markets are crowded with investment managers. The steep cost of big data acquisition and AI implementation may mean only larger managers and niche players will be able to capitalize on the opportunities these technologies present. If only a few large managers dominate the space, their advantage could endure over the long term. But the big data will have to be high quality if it is to yield long-lasting insights: No matter how sophisticated a firm’s AI techniques, they cannot extract actionable investment opportunities from big garbage. Moreover, looking back through history, many market crashes have resulted from overcrowding and glitches in mechanical investment approaches — Black Monday in 1987 and the quant liquidity crunch in August 2007, for example. Such outcomes may be inevitable, even for big data and AI.

3. Are big data and AI incremental steps forward or quantum leaps?

Both asset managers and asset owners believe in big data and AI. But machines cannot persuade humans. Nor do they owe them a fiduciary duty. Moreover our investments have to be accounted for and intuitively understandable. Machines cannot communicate the economic and behavioral rationale of a particular investment strategy. That will take time, perhaps a generation or more. Thus, big data and AI are likely an incremental step forward rather than a quantum leap.

It is critical to avoid excessive expectations and rampant speculation and focus instead on how these tools can be applied appropriately. It is not the best-performing investment manager that survives or the most knowledgeable. It is the one who is most adaptable to change.

If you liked this post, don’t forget to subscribe to the Enterprising Investor.

* Investment professionals should keep the following points in mind:

1. Financial data distribution exhibit non-stationarity and fat-tails both in time series and cross-sectionally.

2. Some phenomena don’t repeat — certain events are one-offs. For example, the Swiss franc soared by more than 20% in value against the euro on 15 January 2015. Will this repeat in the foreseeable future? Probably not.

3. Unlike self-driving cars, for example, we cannot test various economic and political scenarios in the real world. Macro political and economic instabilities are also challenging. Neither the future nor the past is the weighted average of various estimations — there is only one path. Big data and AI might give us a probability-weighted result and every single scenario, but humans have cognitive and decision-making biases that make results subjective.

All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.

Image credit: ©Getty Images/ FrankRamspott

Yoshimasa Satoh, CFA

Yoshimasa Satoh, CFA, is vice president and product strategist/solutions specialist, APAC, at eVestment. He has been in charge of portfolio management, multi-asset investment strategy, and asset allocation model research and development throughout his career. Previously, he served as a portfolio manager of quantitative investment strategies at Goldman Sachs Asset Management and other companies. He started his career at Nomura Research Institute, where he led Nomura Securities’ equity trading technology team. Yoshimasa is a board director of CFA Society Japan. He holds a bachelor’s and master’s degree of engineering from the University of Tsukuba.


Source link

Leave a Reply

Your email address will not be published. Required fields are marked *