Equity Factor Styles and the Challenge of Factor Discovery

# Have any other factor "styles" which explain equity returns been uncovered?


# Have any other factor "styles" which explain equity returns been uncovered?












I know this is an inherently broad question, so I will attempt to clarify what I mean by factor "styles". I am not looking for a compendium of "anomalies", per se, but rather for categorical themes that are institutionally recognized as plausible explanations for equity market returns.

Non-trivial distinctions between "styles", "factors", and "anomalies":

There are many anomalies that have been identified with regards to equity price returns. However, it has been suggested that most such relationships are spurious. Anomalies which stand the stress of "p-value hacking" and academic skepticism may one day get to stand on their own as factors. The difference individual factors and styles therefore may seem somewhat semantic, which is why I defining factor style as broadly-based categories which can arbitrarily contain any number of individual factors and which may be arbitrarily approximated through any number of metrics. Also, I will distinguish that styles have a plausible explanation in financial theory and/or behavioral economics, are persistent across time and markets, and are robust to fitting bias.

Starting with the 800 lb gorilla:

In the April 2015 paper, "A five-factor asset pricing model", Fama-French lay out five factors which explain equity returns. The components of the multifactor model are summarized as such:

- Market Beta (Rm-Rf): the beta coefficient to the market returns

- Size (SMB - small minus big): sorts by size (market cap)

- Value (HML - high minus low): sorts by book-to-market

- Profitability (RMW - robust minus weak): sorts by operating return to book value of equity

- Investment style (CMA - conservative minus aggressive): sorts by asset growth

According to FF (2014), "the five-factor model leaves 42% to 54% of the dispersion of average excess returns unexplained."

In addition to the FF factors, it is generally accepted that the following factors help to explain the variation of returns:

- Momentum (vis-a-vis Jegadeesh and Titman (1993) and Carhart (1997))

- Short term mean reversion

- Low beta/volatility (i.e., the inverse result from which CAPM anticipates)

- Changes in sentiment (e.g. the divergence between analyst expectations and reported results as measured by: post-earnings announcement drift; standardized unexpected earnings; earnings and price target revisions; etc...)

- Informational asymmetry (insider trading signals)

As I mentioned there are plenty of factors which deserve mention, but which I believe fall under one of the factor styles previously listed. For example, earnings yield and net shareholder payout yield can be seen as proxies for value. Balance sheet and cash flow accruals are a conflated means by which to gauge under-regarded profitability and/or investment style. Short interest as well as 13-F alpha cloning can also be viewed proxies for measuring informational asymmetry.

Again, the list of factor-based anomalies could go on ad nauseam, but that's not the intent of my question. The intent is to identify stylistic themes which are different that what has been previously discussed. For example, ESG (environmental, social, governance) factors, and investment flow of funds and/or margin debt might qualify.

Detailed explanations with good references will get upvoted.

## Answer by Helin (score 5)

https://quant.stackexchange.com/a/38030

A wonderful recent paper that might be of interest is Feng, Giglio, and Xiu's "Taming the Factor Zoo."

First, the paper lists nearly 100 "factors" that have been proposed from 1965 through 2016. The corresponding papers that first discuss these factors are also tabulated.

Second, the authors analyzed quite thoroughly whether the newly discovered factors, such BAB (betting against beta) and HML Devil, are truly distinct from existing, well established factors.

## Answer by Frank Fingerman (score 1)

https://quant.stackexchange.com/a/38024

I like your classification (though I would possibly combine sentiment + momentum and profitability + low vol), and it seems as reasonable as any. Most classifications I've seen don't have more than 6-7 categories.

For a completely different approach, you can look at the classification of Harvey, Liu, and Zhu.

If you want to extend beyond "factors" into "strategies", you can consider effects such as Merger Arb, Closed-End Fund Arb, or Index Arb. Those all can be generalized to "classes" of strategies that are distinct from the styles you've mentioned above.