The idea that the risk and return characteristics of global assets can be neatly decomposed into common driving factors is highly appealing to today’s demanding investors. Per Ross1,factor modeling is appealing precisely because “Identifying a small number of factors that influence or describe market returns formalizes our intuitive understanding of the market.”

StatPro is pleased to introduce our Global Equity Fundamental Factor Model, which offers clients the ability to decompose equity portfolios into core factor drivers across style, industry, country and currency.

Making a factor model

Fuss about factors - quoteEquity managers implement their investment theses by over/under weighting stocks across industries, countries, and investment styles - such as value, growth or dividend yield. A fundamental factor model allows us to decompose the actions of the manager and understand the risk and return characteristics of isolated bets on each of these industries, countries and styles. These isolated bets are described by the factors in our model and each has a distinct risk and return profile. The remaining portfolio risk and return not described by the common factors is non-factor, or stock specific risk.

For example, investing in a portfolio of popular internet technology stocks will bring not only exposure to the technology industry factor as expected, but also to the US country factor due to the concentration of IT companies there. An IT stock portfolio in the current market cycle will also bring exposure to the sales growth and momentum style factors.

In order to determine the risk and return profile of each factor we create a long/short portfolio of stocks for each, using a specific set of weights that results in a pure exposure to only the single factor of interest. For industry or country factors the weights will be based on industry and country group membership. For style factors the weights will be based on one or more fundamental descriptors of the stocks, such as price to earnings or market capitalization. A cross-sectional regression is then carried out in order to produce the factor portfolios, each of which represents pure exposure to the desired factor.

The task of the factor analytics provider is to create these pure factors for use in portfolio risk and return analysis. The task of the equity manager or factor product provider is to create investable portfolios which implement the desired factor exposures or tilts, while also maintaining investability - for example, in managing trading and short carry costs, as well as holding to investment constraints.

Understanding style factors

At the end of the day what defines a factor is evidence of a historically distinct risk profile, which indicates the existence of a common driver shared by a group of stocks. It’s intuitive that industries and countries share common drivers of risk, but it’s not always as clear cut for style factors.

Take for example the value style, which in addition to size represent the two style factors originally identified by Fama and French in their study of cross-sectional stock returns. Value in the in Fama-French three factor model is represented by the risk and excess return due to a long position in stocks with favorable book to market ratios and a short position in those with unfavorable ones. Value investing is well-known and there is plenty of appealing economic rationale for why the value factor exists and will continue to exist. Investors are much more likely to be found debating when the value factor excess returns will make a comeback, a topic that touches on the cyclical nature of style factors, rather than debating the existence of the value factor itself.

Other style factors have economic rationale that is more difficult to come by, notably the low volatility and high momentum factors.

The low volatility factor is the risk and excess return due to a long position in stocks with low volatility or low beta and a short position in those with high volatility or high beta. The observation that less risky stocks outperform more risky ones is counter to the CAPM idea of expected return per unit of risk taken. It suggests that there is a persistent anomaly in the pricing of risk. One proposed rationale for its existence is that investors overpay for ‘lottery ticket’ like high volatility stocks in the hopes that one will have an outsized payoff.

The momentum factor, represented by a long position in stocks with larger recent cumulative returns and a short position in those with lower recent cumulative returns, is similarly challenging to explain in that it does not make use of fundamental descriptors in its definition. Factors like momentum and volatility likely exist due to behavioral as well as economic reasons.

Academics and practitioners alike continue to refine style factor models as well as continue to pursue satisfying rationale for their existence - the latter of which is desired to give us comfort that a factor will continue to exist in the future. For instance Fama and French in 2015 updated their three-factor model to a five-factor one, adding two factors measuring company quality; these are operating profitability and investment (companies with lower investment activity provide excess returns).

The goal of the StatPro Global Equity Fundamental Factor Model is to offer a robust and transparent factor model implementation easily accessed via the StatPro Revolution platform. We hope it will serve as a base from which clients can launch an investigation into the myriad of elements in the equity factor universe. We would suggest running the model against a well-known portfolio where there is an intuition that factor exposures exist, and then comparing this intuition vs the StatPro model results. 

1Ross, S.A. (2017). “Factors—Theory, Statistics, and Practice”, The Journal of Portfolio Management

New call-to-action

Paul Rodriguez

Paul Rodriguez

Head of Client Services, New York

Recent Posts