Who Is Better at Investment Decisions: Man or Machine?

Abstract

We conducted a simulated “contest” of the returns of a noted human adviser’s constructed portfolio against the portfolio constructed by a top-rated robo-adviser over a 22-month period (a 17-month backtest period and a 6-month forward test period). Advisers were tasked to construct moderate risk portfolios for nine scenarios of varying ages (30, 50, and 70 years old) and investment amounts ($100k, $500k, and $1MM). Both the backtested returns and the recent returns indicated a solid win for the human adviser even with fees included. These unexpected results showed that at each investment level in our test range ($100k < x < $1MM), the human adviser outperformed the robo-adviser; it was also seen that the robo-adviser was not sensitive to investment or age level, only investor-declared risk tolerance. Therefore, within these customary parameters, investing with the human adviser yielded superior returns.

TOPICS: manager selection, portfolio construction, portfolio theory, wealth management

Key Findings

• • This simulated contest between a human financial adviser and a robo-adviser clearly provides the evidence to dispel potential automation bias in financial adviser selection.

• • This simulated contest clearly provides an observation of inflexibility in robo-advice algorithm application in the same risk profile.

• • This simulated contest clearly provides a demonstration of human financial advisers’ added value in recent market conditions.