Rating a Robo-Rater
Posted: 20 Apr 2020 Last revised: 5 Nov 2020
Date Written: July 6, 2020
Since 2011, Morningstar has issued Morningstar Analyst Ratings on many of the largest mutual funds in the United States. In June 2017, Morningstar launched the Morningstar Quantitative Rating™ to provide a forward‐looking rating on all mutual funds. Morningstar uses a “robo‐rater” machine‐learning model to assign Morningstar Quantitative Ratings. However, the “robo‐rater” cannot utilize the complete set of information available to Morningstar's analyst as it cannot process “soft information.” The purpose of this study is to evaluate if and how this “robo‐rater” is conducive to mutual fund selection. I find no evidence that the “robo‐rater” offers value to investors beyond its assessment of mutual fund expenses and I find that its inability to process “soft information” makes the Morningstar Quantitative Rating™ much less useful than the Morningstar Analyst Rating™. I also examine the relationship between Morningstar Quantitative Rating™ and mutual fund flows and find that the “robo‐rater” has little to no influence on investors' choice of mutual funds.
Paper is available through the Wiley Online Library at https://onlinelibrary.wiley.com/doi/full/10.1002/cfp2.1090
Keywords: FinTech, mutual fund fees, mutual fund flows, mutual fund performance, mutual fund ratings, robo-advising, soft information
JEL Classification: G11, G23, G24
Suggested Citation: Suggested Citation