Bloomberg: artificial stupidity – the new AI and future of fintech

The future of AI as applied to finance is still very much evolving, and at a recent Bloomberg Quant Seminar, Andrew Lo of MIT offered his thoughts on the economic theory and psychological frameworks that underpin modern finance, showing how the human element may be augmented by intelligent machines.

Drawing on his more recent work, Lo has advanced a new conceptual framework that extends the past work by Markowitz and others. His so-called “Adaptive Markets Hypothesis” does not maintain that the theory of efficient markets is wrong, but rather, it is incomplete. If we take a closer look at the gaps and weaknesses in the EMH (efficient market hypothesis), it will change the way we think about investing. Part of the process of exploration includes active research in several areas, including the survey data drawn from a pool of institutional investors, machine learning models of investors, and biometric research on the psychophysiology of traders.

In portfolio management, one of the potential offshoots of current work combining AI and factor analysis could be the development of “precision indexes.” This is reminiscent of the robo-advisor concept, but extends it to create an intelligent trading system that could take not only market data into consideration, but also tailor indexes to an individual according to income, expenses, age, state of health, tax bracket, and risk preferences, for example. While the robo-advisor in its current form is perhaps best used to supplement a human advisor’s knowledge and understanding, a precision index could be automated and adjusted only when needed.

Lo said: “This is just one great example of how the technology has transformed the investment world. The notion of smart beta and more sophisticated passive investment vehicles came about because of the technological advances that enabled us to take these ideas and implement them in an automated fashion, also allowing investors to pay much less for trades and management services.” Clearly, the fusion of human input and machine learning can produce new products and services for an industry always in search of an edge.

Read the full blog

Related Posts

Previous Post
GFMA and PwC: global tech and innovation trends and challenges for investment banks
Next Post
Banking industry prevented fraud of $2bn+ in 2018

Fill out this field
Fill out this field
Please enter a valid email address.

X

Reset password

Create an account