Review of new trends in the literature on factor models and mutual fund performance

Abstract

In this paper we provide critical review of recent developments in the mutual fund performance evaluation literature. The new literature centres around two main themes: enhancing explanatory power of the standard Fama-French-Carhart factor models by augmenting them with different factors and altering standard models to account for presence of non-zero alphas in passive indices used as fund benchmarks.The latter includes the literature providing solutions for scenarios in which those benchmarks do not match fund objectives. We find that in the plethora of suggested 'missing' factors, not one can be universally used to explain all anomalies or price all stocks. We also find that new models that adjust a fund's standard Carhart alpha for alpha of its benchmark or for commonalities in its peer-group, provide additional information on fund performance to that given by the standard models. Specifically, these models give account of fund's relative performance -to the benchmark or the peer-group, which is of use to investors.

Mutual fund performance: Some recent evidence from European equity funds

“…Our findings for U.S.-based European equity funds are consistent with previous studies on the performance of mutual funds that invest in the United States (see, for instance, Mateus et al 2019 for an overview). As we pointed out in Section 1, not many studies focus on the performance of European equity funds.…”

“…Hou et al (2017) were able to identify as many as 447 different average-return anomalies. Mateus et al (2019) provide a thorough overview of the known anomalies in the context of fund performance and persistence measurement. Titman et al (2004), Novy-Marx (2013), and many other authors have since pointed out that a possible reason why the three-factor model is incomplete is the lack of variation in average returns that originate from company profitability and investments.…”

“…Historically, only a fraction of funds has been able to generate abnormal returns when trading friction is taken into account (Wermers, 2000;Kosowski et al, 2006). There is a compelling body of empirical evidence that 'top' mutual funds (irrespective of the ranking criteria) are unable to generate persistent returns and that their performance is mainly ephemeral (Mateus et al, 2019). Only the 'losers' tend to exhibit persistent losses (Carhart, 1997).…”

PackageCovRegpy: Regularized covariance regression and forecasting inPython

“…Less developed families of multivariate dynamic volatility models for covariance forecasting are the families of covariance regression factor models. These are multivariate generalizations of classical factor models such as capital asset pricing model (CAPM), arbitrage free pricing theory (APT) models, and extended Fama-French 3-and 5-factor models, see a review in Mateus et al (2019). Importantly from the perspective of the software package developed in CovRegpy, many of the new classes of dynamic factor models that can be used for covariance forecasting in multivariate asset return time series are not widely available in packages in R or Matlab.…”

A Comprehensive Method for Ranking Mutual Fund Performance

“…However, investors are faced with difficulties in choosing a mutual fund among hundreds and thousands of equity funds or bond funds. The guiding need in making this important decision has led to the development of an abundant work on how to measure and rank mutual fund performance (Durán Santomil et al, 2022; Grau-Carles et al, 2019; Mateus et al, 2019; Parida & Teo, 2018; Venkataraman & Rao, 2021; Yu et al, 2022).…”