6 December 2019: Sheri Markose gives talk at the Innovation Lab of the Reserve Bank of Australia on AI, FinTech, Robot Advisors and Cryptocurrencies
Professor Sheri Markose, University of Essex, UK
The application of AI is proliferating in every aspect of financial and monetary technology. The talk will focus on two of the latest such applications: Robot Advisors and Cryptocurrencies.
An extensive back testing of leading Robot Advisor investment strategies finds that the fears about a black box style algorithm leading to not so easy to explain outcomes are not warranted. The Robot Advisor investment strategies conform to well-known Markovitz style portfolios and their performance is robust by standard back testing rules. However, the risk to investors and also systemic risk come from the asset of choice of Robot Advisors being exchange traded funds (ETFs). Additional back testing for the sudden loss of liquidity is proposed for the Robot Advisor ETF portfolios. In part II of the talk, issues regarding the competition from cryptocurrencies for state based fiat digital currencies is critically examined from the perspective of double spend and stability problems. A simple agent based model is used to show the potentially socially wasteful side to the boom bust dynamics of competing cryptocurrencies that are investment vehicles and means of payment. The inherent boom bust cycle in the price of cryptocurrencies arises from the inherent tendency for agents to use the cheapest currency as means of payment (as in Gresham’s Law) and the expensive ones for mining and as an investment asset.