Skip to main content

You are here

Blog Listing Page

Taming the factor zoo

By: Tom Goodwin, senior research director

Investors’ interest in smart beta has increased significantly in recent years, with a specific focus on factor investing.[1] That interest has helped drive an explosion of research on factors. But there can be too much of a good thing: “We now have a zoo of new factors,” declared Professor John Cochrane from the University of Chicago in 2011.[2] Since this statement, the proliferation of factors purporting to deliver excess returns has only accelerated. In just the years between 2010 and 2012, 59 new factors were “discovered.”[3] By some counts, there are now over 300 factors in academic and practitioner literature. But in a zoo of 300, how can anyone sort through the herd and find a group of factors that will generate reliable, sustainable long run premiums? This post outlines the process at FTSE Russell.  

Factors are broad and historically persistent drivers of returns that are present in all stocks. Properly specified, they account for all risk except idiosyncratic risk. It is important to remember that all stocks, indexes and portfolios have embedded factor exposures – whether they’ve been formally identified or not. Researchers at FTSE Russell were determined to filter through all the factors found in academic literature to arrive at a group of only the most reliable factors that can reasonably be expected to persist well into the future. They used three criteria to conduct their analysis, as seen below.

Factors should be grounded in long and deep academic literature that has proven to be robust over a variety of definitions. One or two academic papers do not make the cut. There is a long-standing bias against publishing negative results in journals, so the absence of follow-on research often means the factor has not survived further testing.

We also determined that there are three types of intuitive economic rationale behind a successful factor. The first is rewarded risk. Certain factors have earned higher long-term returns as a reward for bearing greater risk. A second rationale encompasses behavioral biases. Not all investors are perfectly rational all the time, generating opportunities for those who can take a contrarian view. A third rationale embraces structural impediments such as market rules or restrictions that can make some investments off limits to certain investors, creating opportunities for others who can invest.

As I mentioned earlier, academic research is chock-full of the discovery of new factors, most of which are based on ephemeral anomalies that do not survive further scrutiny. Another track is our own empirical research. At FTSE Russell we insist that a proposed factor must be robust across all the major regions.

Additionally, we maintain that the surviving factors should be distinct from one another. This has led us to actually reduce the number of factors from eight to six over the past couple of years by observing that the live return patterns of two of the factors were very similar to two others.[4] 

At the end of this rigorous process we have six surviving factors. Our methodology “tilts” a starting set of weights, usually market capitalization or equal weights, toward the characteristics of that factor.

Attempting to data mine 300 potential factors that supposedly affect stock returns would only produce a strange and possibly random assortment of beasts in the zoo. At FTSE Russell, we believe that by applying the three criteria discussed above we have arrived at a manageable grouping of factors that display compelling results consistently over time.

For more information on our approach to factor exposures, read our research, Multi-factor indexes: The power of tilting.



[1] FTSE Russell 2017 Smart Beta Survey

[2] John Cochrane, “Presidential Address to the American Finance Association,” 2011.

[3] Campbell Harvey, Yan Liu, and Hequing Zhu, “…and the Cross Section of Expected Returns,” Duke University, 2015.

[4] The two removed are “Illiquidity,” which turned out to be highly correlated with the Size factor and “Residual Momentum,” which turned out to be highly correlated with the Momentum factor.


© 2018 London Stock Exchange Group plc and its applicable group undertakings (the “LSE Group”). The LSE Group includes (1) FTSE International Limited (“FTSE”), (2) Frank Russell Company (“Russell”), (3) FTSE TMX Global Debt Capital Markets Inc. and FTSE TMX Global Debt Capital Markets Limited (together, “FTSE TMX”) and (4) MTSNext Limited (“MTSNext”). All rights reserved.

FTSE Russell® is a trading name of FTSE, Russell, FTSE TMX and MTS Next Limited. “FTSE®”, “Russell®”, “FTSE Russell®” “MTS®”, “FTSE TMX®”, “FTSE4Good®” and “ICB®” and all other trademarks and service marks used herein (whether registered or unregistered) are trade marks and/or service marks owned or licensed by the applicable member of the LSE Group or their respective licensors and are owned, or used under licence, by FTSE, Russell, MTSNext, or FTSE TMX.

All information is provided for information purposes only. Every effort is made to ensure that all information given in this publication is accurate, but no responsibility or liability can be accepted by any member of the LSE Group nor their respective directors, officers, employees, partners or licensors for any errors or for any loss from use of this publication or any of the information or data contained herein.

No member of the LSE Group nor their respective directors, officers, employees, partners or licensors make any claim, prediction, warranty or representation whatsoever, expressly or impliedly, either as to the results to be obtained from the use of the FTSE Russell indexes or the fitness or suitability of the indexes for any particular purpose to which they might be put.

No member of the LSE Group nor their respective directors, officers, employees, partners or licensors provide investment advice and nothing in this communication should be taken as constituting financial or investment advice. No member of the LSE Group nor their respective directors, officers, employees, partners or licensors make any representation regarding the advisability of investing in any asset. A decision to invest in any such asset should not be made in reliance on any information herein. Indexes cannot be invested in directly. Inclusion of an asset in an index is not a recommendation to buy, sell or hold that asset. The general information contained in this publication should not be acted upon without obtaining specific legal, tax, and investment advice from a licensed professional. 

No part of this information may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the applicable member of the LSE Group. Use and distribution of the LSE Group index data and the use of their data to create financial products require a licence from FTSE, Russell, FTSE TMX, MTSNext and/or their respective licensors.

Past performance is no guarantee of future results. Charts and graphs are provided for illustrative purposes only. Index returns shown may not represent the results of the actual trading of investable assets. Certain returns shown may reflect back-tested performance. All performance presented prior to the index inception date is back-tested performance. Back-tested performance is not actual performance, but is hypothetical. The back-test calculations are based on the same methodology that was in effect when the index was officially launched. However, back- tested data may reflect the application of the index methodology with the benefit of hindsight, and the historic calculations of an index may change from month to month based on revisions to the underlying economic data used in the calculation of the index. 

This publication may contain forward-looking statements. These are based upon a number of assumptions concerning future conditions that ultimately may prove to be inaccurate. Such forward-looking statements are subject to risks and uncertainties and may be affected by various factors that may cause actual results to differ materially from those in the forward-looking statements. Any forward-looking statements speak only as of the date they are made and no member of the LSE Group nor their licensors assume any duty to and do not undertake to update forward-looking statements. 

Blog Listing Page