Practitioner's Perspective on New Factor Models
Discussing how practitioners view and analyze new factor models in asset management, with insights that go beyond academia. The discussion includes comparisons, biases, and insights gained from conversations with renowned figures in the field. Key points cover factors like the Q-factor model and the importance of investing based on empirical and theoretical considerations.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
You are allowed to download the files provided on this website for personal or commercial use, subject to the condition that they are used lawfully. All files are the property of their respective owners.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.
E N D
Presentation Transcript
411 BOREL AVENUE, SUITE 300 SAN MATEO, CA 94402 650. 931.1500 WWW.FULLERTHALER.COM info@fullerthaler.com Discussion of: A Comparison of New Factor Models by Kewei Hou, Chen Xue, & Lu Zhang The Practioner sPerspective : Raife Giovinazzo, Fuller & Thaler Asset Management Disclaimer: These are my views, and not the views of Fuller or Thaler or Fuller & Thaler March 20 2015
My discussion (hopefully) 1. Tell how practitioners view these factor models 2. Say some things you can t say in academia 1 For Discussion Only
My Practitioner Background (& Biases) Partner, Director of Research at Fuller & Thaler Asset Management Over $3 Billion in assets under management All strategies based on behavioral finance 5 years in strategy consulting Many discounted cash flow models to make investment decisions Undergraduate at Princeton, 1995 Daniel Kahneman was thesis advisor PhD in Finance from University of Chicago Richard Thaler & Eugene Fama on dissertation committee 2 For Discussion Only
My Conversation with Kahneman 20+ years ago Raife: Many of the seminal papers we cite [in psychology] had multiple preceding papers that said almost the same thing. How do we pick who to cite? Kahneman We cite the person who convinces the world, not the person who thought of it first. And that is a good set of incentives 3 For Discussion Only
Overall Thoughts I like the investment-based asset-pricing research agenda The key differences 1. Annual ROE vs Quarterly ROE 2. No Value The key result Q-factor better empirically Challenge: What is the purpose of a limited factor model? Describe [empirical] Then why omit value, esp. up-to-date value? Explain [theoretical / causal] Then why quarterly ROE What is cool about this model: Quarterly ROE Maybe managers are more rational than investors 4 For Discussion Only
Motivation I like the investment-based asset-pricing research agenda In a world with rational participants Executives Investors Average Realized Rate of Return Required Rate of Return Expected Rate of Return Marginal Return On Investment = = = 5-Factor Model (Agnostic, IRR) q-factor Model (Investment) vs. Fama & French 2014 Hou, Xue & Zhang 2014 5 5 For Investment Professional Use Only Not to be Shown or Distributed to the Public
The Factors q-factor Model (4-factors) vs. 5-Factor Model Hou, Xue & Zhang 2014 Fama & French 2014 1. Market 1. Market 2. Size 2. Size (SMB) 3. I/A: Change in total assets / one-year lagged total assets 3. Investment (CMA) 4. ROE: Quarterly IBQ / Book Equity 4. ROE: Annual ROE (RMW) Monthly sorts o Annual sorts o 5. Value (HML) Note: Some differences in orthogonalization (2x3 vs. 2x2x2x3) For Investment Professional Use Only Not to be Shown or Distributed to the Public 6 6 6 6
Practitioner Perspective: Five Factor Model Is An Endorsement of Q-Model even the four-factor model [Fama-French-Carhart] had problems there were many anomalies that it couldn t explain. Kewei Hou, Chen Xue and Lu Zhang, authors of the September 2012 study, Digesting Anomalies: An Investment Approach, proposed a new four-factor model Professors Fama and French, in a June 2013 paper, A Five-Factor Asset Pricing Model, took a close look at the new model, to see if these new factors investment and profitability added explanatory power. Only time will tell if the new model replaces the Fama-French one. But with their endorsement, it does seem like a good possibility. 7 For Discussion Only
Practitioner Perspective: Dont Believe Q-Theory Investment Low Returns is not about low cost of capital This new factor is called investment. firms that significantly increase capital investment tend to achieve sub-par subsequent returns. In other words, given sufficient opportunity and proclivity, most managers become capital destroyers. 8 For Discussion Only
Theoretical Critiques Dont Make Sense To Me FF internal rate of return = long-term average expected return, maybe not one-period- ahead expected return Huh? This applies to both Five-Factor & q-model (managers don t make 1-month investment decisions) 1. 1. q-theory says value should not matter, and it doesn t matter in the data Two issues 1. Emphasis: Blockbuster finding! 2. It matters if you have up-to-date value ? 2. 2. Uncertain relationship between investment & returns 3. 3. Huh? This applies to both Five-Factor & q-model Past investment does not predict future investment (R- squared less than 5%) 4. 4. 9 For Discussion Only
Empirical Findings: q-factor is better at pricing Q-factor model prices more anomolies than Five-Factor Model 36 high-minus-low decile anomalies using NYSE breakpoints and value- weighted returns Q-factor model: 7 out of 36 anomalies have alphas Fama-French model: 19 out of 36 anomalies have alphas Q-factor model can price the 5-factors, but 5-factors cannot price the q-model (still some alpha) 1. 2. 10 For Discussion Only
Value matters to investors today Talk to investors: they care about Size and Value Morningstar style box classification Names of funds often include size or value in the name E.g., Undiscovered Managers Behavioral Value Fund Fama-French or Carhart alphas, that control for value Including Value (even if zero alpha) is a feature, not a bug, of the Five-Factor Model 11 For Discussion Only
And maybe Value is not redundant AQR: Our Model Goes To Six And Saves Value From Redundancy Along the Way Uses HML-DEV, which uses latest month prices for book/market Strongly significant once UMD added Challenge: If you think using up-to-date ROE is important, why not up- to-date value? 12 For Discussion Only
Whats Weird & Cool About Quarterly ROE Weird Real investment decisions about plant, property, equipment are made by executives (i.e., using discounted cash flows) are made using long-term forecasts, not quarterly realizations Quarterly ROE has a problem of seasonality E.g., Chang, Hartzmark, Solomon & Soltes 2014 Does Quarterly ROE pass the analyst-forecast-IRR tests Finding in paper: Robust-Minus-Weak does not Side Note: Does investment? What about international data? Especially countries that only report data annually Cool It makes sense to include the most up-to-date information Empirically it works better (e.g., prices momentum) 13 For Discussion Only
Four Possible Directions for the Paper Value Is Dead Title, emphasis matters! Because if its about factor models, practitioners want value The Devil in the ROE Details Why Quarterly, Monthly sorts on ROE make sense Executives are Rational (even if Investors are Not) My dream: A model that separates the rational from the behavioral, instead of implying everything is rational or everything is behavioral The Six-Factor Model With up-to-date ROE and up-to-date Value and UMD 1. 2. 3. 4. 14 For Discussion Only
Summary Challenge: Write a factor paper that practitioners will use You ve persuaded the world to include ROE & investment as factors To become the most cited factor paper . Titles, variable names Emphasis matters How it fits with current practice matters (practitioners want value!) Q-factor model is empirically better at pricing anomalies than the Five Factor Model Better empirically because it uses quarterly ROE. Weird and intriguing. Empirically: Practitioners want a model that includes value Even if irrelevant But probably not if you use up-to-date value Theoretically: Why quarterly ROE Quarterly ROE is weird if about real investment decisions And by the way, people doubt that investment = low cost of capital 15 For Discussion Only
