RAFI Case

FIN 3300 - Winter 2020 Research Affiliates

Due Wednesday February 26 by start of class

Instructions

*Note: This case is in the HBSP coursepack listed in the syllabus.

In this case you will reconsider a historically important decision by a large consulting company called Towers Watson (www.towerswatson.com). Towers Watson decided to en- dorse the Research Affiliates fundamental index strategy (RAFI) to its institutional clients. The ETF based on the RAFI strategy was the first of a new category of ETFs called “smart beta” that has become very popular and is still rapidly growing. The Towers Watson en- dorsement was one of the key points in history that led to the growth of these smart beta products. Your job is to pretend you are Towers Watson at the time of this decision and decide whether of not to recommend the RAFI strategy to your institutional clients.

You may work alone or in groups of up to four. Turn in one writeup per group. You may assume that I know the content of the case when you answer questions; you do not need to summarize case background in your answers. In addition to your writeup, be prepared to discuss the story in the case during class on the due date.

Your case write-up should be double-spaced in 11 or 12pt font. You are to answer the following questions (number them in your writeup) in at most one paragraph each. If you do not meet this style criterion, you will be asked to revise and resubmit the writeup until it is meets the limit (possibly with a 10% markdown at my discretion). You should view this as an opportunity to exercise your business writing skills. In business writing, you put the conclusion first—as you will with question 1—and you get to the point as fast as possible. If you do not have to say something extra in business writing, DON’T! You give the readers the answers they are looking for and nothing more to get in the way. I am not looking for extensive details, just simple straightforward answers to the questions. Please turn in your case writeup and Excel spreadsheet via the Canvas link.

Hints: Read the case very carefully. Some people missed the point of questions, which hurt their grade, because they tried to answer the questions without thoroughly understanding the context.

In the past, some people automatically assumed that Towers Watson got the decision right, do not assume this. Make up your own decision. Also, please be advised that out of fairness, I can only ask clarifying questions; I can not answer any question that has the effect of giving you the right answer (for example, I can not look at your spreadsheet to help you look for errors).

1

Questions

1. (Executive summary) Based on your analysis in the following questions, would you endorse the RAFI strategy to any of your institutional clients? (Hint: Recall discussion from class; given a CAPM alpha on an asset, how much weight should you put on that asset in a portfolio?)

2. What is Research Affiliates view of market efficiency (they call this “price efficiency”, not to be confused with “mean-variance efficiency”, which they also discuss) and why does Research Affiliates think the Fundamental Index strategy is better than the cap- weighted approach? What does this implicitly mean Research Affiliates believes about the usefulness of accounting values of corporations’ stocks relative to their market prices?

3. Even if Research Affiliates views about fundamental weights being better than market cap weights were true, are there any extra costs for using fundamental weights in trading strategies that would not be present using market-cap-based weights?

4. Did the RAFI ETF earn CAPM alpha leading up to this decision? The data is on Canvas.

5. RAFI’s sales pitch was based on the analysis in Exhibit 4. Notice the time period of their analysis and read the notes at the bottom of the Exhibit. Do you have any reason to trust your analysis more than RAFI’s analysis in Exhibit 4?

6. (Optional) Did the RAFI ETF earn alpha with respect to the Cremer-Petajisto-Zitzewitz (CPZ) model? The description of the CPZ model is below.

Discussion questions

Do not put answers to these in your case writeup, but be prepared to discuss them for participation points.

1. Why do recommendations of large consulting firms like TW matter in investments?

2. How are fees likely to differ across market-cap weighted funds and fundamental index funds?

3. Based on Exhibits 12 and 14, what is the value or growth orientation of the Fundamen- tal Index Strategy? Are the related factor exposures from the CPZ model consistent with this?

Supplemental information

1. The Fama-French 3 factor model we discussed in class uses hypothetical size- and value- based portfolios SMB and HML that are not easy to invest in. Hence, a famous 2013 paper by Cremers, Petajisto, and Zitzewitz (CPZ) proposes the following factor model

2

that incorporates size and value-related factors, but is better for evaluating investment performance than the Fama-French 3 factor model because it uses only factors that people can invest in easily with mutual funds or ETFs:

rit − rft =αi + βmkt(rSP500,t − rft) + βMC(rRMidCap,t − rSP500,t) + βSC(rR2K,t − rRMidCap,t) + βV G(rR3KV,t − rR3KG,t) + �i,t

• rSP500,t −rft is the excess return on the S&P500, a proxy for the market portfolio, or at least for large cap stocks.

• rRMidCap,t − rSP500,t is the return on the Russell Midcap index in excess of the S&P500. This captures the common factors driving mid-cap relative to large-cap returns.

• rR2K,t − rRMidCap,t is the return on the Russell 2000 index in excess of the Russell Midcap index return. The Russell 2000 index is a small-cap index. The factor rR2K,t − rRMidCap,t captures the common factors driving the returns of small-cap stocks relative to mid-cap (or really just non-small-stock) returns.

• rR3KV,t − rR3KG,t is the return on the Russell 3000 Value index in excess of the Russell 3000 Growth index. The Russell 3000 index is a broad stock market index that consists of 3000 US stocks. The Russell 3000 Value and Growth indexes are the portfolios of value and growth stocks, respectively, that are in the Russell 3000 index. The factor rR3KV,t −rR3KG,t is similar to the Fama-French HML as it captures the common factors driving the returns of value stocks relative to growth stocks.

Data for this model are included with the RAFI data on Canvas. This model is be- coming more popular in practice for evaluating performance.

Grading priorities

Writers of an ‘A’ case writeup will be able to answer ‘yes’ to the following questions:

• Quality of Reasoning and Presentation: – Did you answer the question asked?

– Did you use correct reasoning based on principles from class?

– Is the writing clear?

– Do you include more than necessary to make the point (not good)?

– Is your format correct?

• Are your calculations correct?

3