Investors should always be skeptical about new strategies that promise abnormal returns. As I explained in last week’s post, some of those strategies are dubious to begin with, and even legitimate ones don’t always last.
But that’s not to say academic research can’t shed light on how markets work. Indeed, the Fama-French three factor model, described in a landmark 1992 paper, reveals how small-cap and value stocks offer additional risk premiums that have become the specialty of Dimensional Fund Advisors.
Now there’s new research that may prove equally influential. In The Other Side of Value: The Gross Profitability Premium, Robert Novy-Marx of the University of Rochester argues that “profitability, measured by gross profits-to-assets, has roughly the same power as book-to-market predicting the cross-section of average returns.”
As you can tell, the paper is rather technical, but here’s the basic idea. It’s rational to expect higher returns from cheap value stocks, since they are inherently riskier. It is harder to explain why profitable companies would reward investors when they trade at higher multiples and are less prone to distress. Yet Novy-Marx found that profitable companies actually performed about as well as value companies. (For more, see Larry Swedroe’s summary of the paper’s findings.)
Playing for all the marbles
I first heard about this idea from Art Johnson, a portfolio manager with Scotia McLeod in Calgary and one the early pioneers who adopted DFA in Canada almost a decade ago. Johnson feels that a profitability screen will take the Fama-French research to the next level. Here’s how he describes the significance:
“The value factor measures the number of assets you receive relative to the price of the stock. So think of a game of marbles: tilting to value stocks is really saying if I have more marbles (assets) than you, I believe I will win in the end. The way we identify the number of marbles we’re buying is by using the book-to-market ratio.
“The problem with this strategy is it’s risky. In order to get more marbles I have to trade with someone in trouble who is willing to surrender them. And I don’t know if these are good marbles, I just know I have more of them. We can partly overcome this through mass diversification: companies like DFA buy thousands of marbles (value stocks). But the exciting thing is that academics now seem to be making headway into measuring the quality of those marbles.
“Now we can think about a method whereby we’ll not only have more marbles than the other guy, but we’ll also be able to sort them based on how good they are. This is a big deal: before this work, you could argue we only had half the puzzle solved. We knew where we could take more risk, but now this allows us to sort through companies and only retain the most profitable ones.”
Low correlation with value
The other promise that comes out of Novy-Marx’s paper has to do with the relationship between value and profitability premiums. As the author writes: “Strategies based on gross profitability generate value-like average excess returns, even though they are growth strategies that provide an excellent hedge for value.”
In other words, the two strategies not only deliver similar returns, each tends to zig when the other zags—and negative correlation is the Holy Grail of diversification. “This means you can run them in tandem,” says Johnson, “and since they are separate risks, you actually end up lowering volatility and enhancing returns.”
What is the main risk of the profitability factor? Imagine buying highly profitable companies and shorting those with low profitability, a hypothetical strategy reviewed in the paper. “Think of how that would have worked in the tech bubble,” Johnson says. “You would have been shorting very unprofitable tech stocks that kept going up, while buying value and profitable companies [that lagged significantly]. Ouch. So there is no free lunch.
“Our knowledge of how you can be paid for being different from the total stock market has just undergone a massive expansion because of this work,” Johnson says. “Just be cautious, and be sure you know what the risks are before you naively see only the reward.”
Super interesting, but I’m not entirely sure that I understand the risk that this particular risk-premium is coming from. Intuitively, is this just a bubble proof way of going after growth stocks?
Also intuitively, shouldn’t this be closely related to dividend stock strategies? Aren’t dividend stocks just profitable growth stocks who choose to distribute those profits in dividends?
Book-to-Market strategy can also be combined with momentum as explained in this must-read paper Value and Momentum Everywhere, C. S. Asness, T. J. Moskowitz, L. H. Pedersen, 2009
Kiyo, that is a great question. Where is the risk-premium coming from? The risk is that the growth firms will not be able to achieve the high asset turn. (In other words they used their assets better than the other growth stocks, ergo the premium. Without the higher asset turn they have average growth returns.)
This took me a long time to get my head around since I am a true markets are efficient, it has to be a risk factor, there is no free lunch kind of guy. Just as value stocks have to surrender more assets, growth stocks have to achieve higher asset turn. The risk is that this may not happen so tilting may or may not pay-off.
As I said to Dan, there is no magic in the value factor: all it measures is the number of assets you receive relative to the price of the stock. The argument for “tilting” your portfolio, being different from the total stock market (TSM), in this manner is that companies that have low prices will give you more of their assets as a fair trade-off for their risk. The inverse of this is the growth stocks have to achieve higher asset turn.
A TSM investor is not willing to take that risk but I am. Now that I have two different risks with a negative correlation I am a much happier risk taker.
Dividends are just from the growth in profits. A dividend is not free money: it has to come from somewhere. When companies are sorted on dividends you really are capturing the value premium rather than some magic in dividends. But that is a whole other topic that created a firestorm on Dan’s blog earlier. Canadians, me included, love dividends but as an investor you should step back and realize you don’t have a dividend problem you have a cash-flow problem.
Art
I don’t understand the second part of this sentence:
It is harder to explain why profitable companies would reward investors when they trade at higher multiples and are less prone to distress.
Shouldn’t they be more prone to distress? At least the stock should if profits fall as the multiple is higher. Can you elaborate a bit on this sentence Dan?
@Noel: You’re thinking like a true-blue value investor. :) But if we forget about stock price for a moment, a company that is highly profitable is clearly less prone to distress than a company that is not earning profits. A rational investor might conclude that profitable companies are less risky, and therefore should have lower expected returns. But according to the paper, profitable companies actually deliver higher returns “despite having, on average, lower book-to-markets and higher market capitalizations.”
CCP,
One of the most valuable features of this blog is the willingness to cover investor issues that occur on the fringes of the couch potato strategy, so kudos for that.
It’s also a great example, for me personally, of not wanting to complicate an already working CP portfolio. I have to admit, I don’t understand the theory above, so wouldn’t try and buy into it.
I was looking at returns as presented at Scott Burns private company as reported for a number of portfolios, and the simplest of them all (two funds) seemed to be doing the best.
Do you have any thoughts on when an investor might really need to consider adjusting a portfolio for a more complicated approach? (Disclosure: I have a hybrid of a complete couch potato portfolio, adjusted for more dividend income and matched to an Investment Policy Statement’s weighting; I think I may be move than diversifed enough).
@JAH: Your comment is very insightful.
I like to broaden the discussion occasionally to these advanced strategies, because I find them interesting and I think many of my readers do, too. A blog that only dealt with plain vanilla ETFs and “stay the course” would be valuable, but a bit boring.
But to answer your question about when an investor might need to consider a more complicated approach, the answer is clear: never. Absolutely no one “needs” to make their investment strategy complicated. As you say, the simple strategies often do best, if for no other reason than they are cheaper and easier to adhere to over the long term.
I present this stuff as bit of advanced education for the keeners. But if it doesn’t interest you, or if you don’t understand it, you should confidently ignore it and stick to what makes you comfortable. You’re not “settling” by using a simple strategy.
FYI, here is the DFA International Core Equity Fund CAGR (MER 0.60%) compared to the S&P500:
“A person who invested $1,000 in this class of fund on June 6, 2005 would have $958.74 as of May 31, 2012. This works out to an annual compound return of -0.60%.”, vs. +1.31% for SPY (dividends NOT included)
@SPY: You’re comparing international stocks to US stocks, which is meaningless. Excluding dividends also renders the comparison useless.
If you want a US comparison, since its inception in July 2005, the DFA US Core Equity Fund (DFA293) has delivered 0.86% annualized, compared with 1.83% for the Russell 3000, a proxy for the broad US market. That’s a lag of -0.97%, but of course, the index has no associated fees. Value stocks have unperformed since 2005, so the result is pretty much exactly what you would expect.
@CCP:
I get your point. VEU is a better candidate for the comparison.
You write “excluding dividends renders the comparison useless”. I call it conservative.
But why the aggressivity?
Russell 3000 ETF is IWV or VTHR with a 0.16% annual fee.
Hope your readers have a better idea about these DFA funds, whichever benchmarks are used to compare them to.
@The SPY Surfer
Investors will not have much luck comparing the U.S. dollar returns of VEU to the Canadian dollar returns of the DFA International Core Equity Fund (DFA295).
Same thing goes for comparing the returns of any Russell 3000 U.S.-listed ETF to the Canadian dollar returns of the DFA US Core Equity Fund (DFA293).
The return Dan quoted for the Russell 3000 Index was in CAD – a reasonable comparison.
@SPY: I didn’t mean to sound aggressive, and sorry it came across that way.
Re: dividends: the issue is that a value strategy is likely to include higher-yielding stocks, so a greater portion of the return will come from dividends compared with a broad-market approach. If you look only at price changes, the broad-market fund might outperform by 1%, but if the value fund yields 1.25% more its total return would be higher. (I just made those numbers up, but that’s the idea.)
By the way, thanks for mentioning the Asness paper in your earlier comment. I added a link so other readers can check it out. Cheers.
You’re welcome! ;-)
Dan being aggressive! I would like to defend Dan to his readers. Dan has the support and ear of some of the top portfolio managers practicing passive strategies in Canada. As a group we believe the quality of the work that comes out on his blog is far and above better and more balanced than the lion’s share of other financial blogs or financial media for that matter.
I personally believe Dan is the rare breed of financial writer that is able to appreciate the nuances of these strategies and has the unique ability to inform his readers about what is going on. That is why he has gained our trust and support.
I believe it is important, as one of Dan’s readers, to recognize that in the world of finance the standard idea of risk and return is evolving. That is the point of this article. It is not to point at one uber alles solution that trumps them all. I also believe if you are going to think about these things you have to think probabilistically rather than black or white.
In a black and white world you would say that the DFA US Core Equity Fund (DFA293) has delivered 0.86% annualized, compared with 1.83% for the Russell 3000, a proxy for the broad US market which has a lag of -0.97%, and higher fees has been a bust and that you should not do anything but buy an index.
Now shift your thinking to probabilities and value as an asset class has a 72.4% chance of outperforming the S&P over 25 years. Small value has an 88% chance.
The point of all of this as we expand our knowledge of how markets work we can better apply this knowledge to meet clients goals. It is also important to recognize none of this is a free lunch. What is fascinating about the recent work being done is that it looks like we can blend two negatively correlated asset classes and increase our probabilities thus enhancing out ability to meet our clients’ goals.
Dan is one of the thought leaders in this space and I thought it was important for him to pass this along. It is important when you are reading any of this to understand that a TSM position is a choice that will satisfy the majority of investors and the way you are doing is far superior to what most Canadians do. But also recognize that probabilistic thinking has a place in this discussion.
Art
Source: Truman Clark 250,000 Years of Stock Returns.
+1
Dan is probably one of the most articulate writers on index investing, and investing as a whole for that matter, and his blog is an incredible ongoing resource for me. And, he even responds to specific reader questions (and in a timely manner).
Reading through the last two blogs and all the comments and comparisons I wonder how anyone is supposed to decipher any strategies and make concrete conclusions. There is a study to prove and disprove anything and everything. Further to that we can choose to believe or disbelief any evidence that will help or hurt us. I thought to myself “why are these people debating with each other?” Don’t they have anything better to do? Then I thought…..this is what my wife must think when she here’s me talk about golf! LOL! Passion is the greatest gift! Keep up the great work!
@Art and Noel: Thanks for the kind words. I’m fortunate to be tapped into a network of very smart people who are willing to share what they know with the investing public.
@Al M: You’re totally right—a lot of this stuff is “inside baseball,” and as I wrote to JAH above, investors should feel free to ignore whatever they think is too esoteric. You certainly don’t have to absorb all of this to be a successful index investor, but as Art would say, sometimes a little “Geekapalooza” is fun.
The irony is that my next post will be part of the Blog for Financial Literacy campaign, in which bloggers are invited to share one simple tip to help the average Canadian. I promise not to suggest that investors learn how to do Fama-French three-factor regressions. :)
@ Noel November 13, 2012 at 5:48 pm
I second your comment.
Keep up the good work Dan :)
The main reason why DIY investors enjoy CCP is that not only did he teach us how to build low-cost portfolios with broad-based ETFs but also how to say bye-bye to financial advisors and their mutual funds. :-)
Excellent post and comments. Unlike some publications I read, these comments are well written and everyone’s point of view gets treated with openess and respect. It is good for us DIY enthusiasts to get exposure to new ideas and then discuss them.
My own groundbreaking and influential (to me at least) capital-limited efficient market hypothesis states that if they really have found a flaw in the market, then they have just eliminated it by telling the market about it.
@Patrick: hey, self, don’t you know Dan said exactly this just a few days ago? Sheesh.