“There is so much that’s false and nutty in modern investing practice and modern investment banking, that if you just reduced the nonsense, that’s a goal you should reasonably hope for.”
We’ve heard that before. Modern portfolio theory is Buffett and Munger’s favorite martyr child. Having been introduced to the world of investing through Graham and Buffett’s writing, it is nothing less than a mind-opening exercise for me to attend business school and study modern finance. While I try to pound all those complex greek-salad-formulas into my head I often hear this little voice in the back of my mind reminding me that it might all just be a huge waste of time. I feel like I’m in the position of a judge deciding the outcome of a boxing match between Buffett and my university teacher with net worth as the only judging criterion.
My goal here is not to support or discredit Buffett’s notorious MPT bashing (although I personally think he’s right for the most part). What I want to explore is if MPT can be used in any useful way by value investors and if so, how. I’ve been thinking a lot about how to reconcile the Graham school of investing with the modern one. Since I’ve made the choice of going through b-school, I might as well try to make something useful out of it. Truth is if Buffett is right (which I highly suspect), I have little interest in Markowitz’s theories. I’m not here to complete a PhD, write a ground-breaking thesis or even achieve a 4.0 GPA. I’m here to learn how to invest and make money.
So let’s explore how, as value investors, we can use what they teach in modern finance courses to our advantage.
First, let’s start with what I think should be discarded (hint: pretty much everything). The other day I stumbled on a study published in 1990 by Richard Baillie and Ramon DeGennaro¹. These academics had the brilliant idea of challenging the prevalent thinking that there is a positive relationship between returns and risk measured as variance or standard deviation. Here’s what they conclude (emphasis added):
“Controlling for excess kurtosis by use of the student t density is found to be important, and the estimated models show very little evidence for a statistically significant relationship between a stock portfolio’s return and its own volatility. Our results suggest that traditional two-parameter models relating portfolio means to variances are inappropriate and indicate the need for research into other measures of risk.”
Now think about that for a second. The entire body of finance theory taught in universities around the world is based on the premise that in order to achieve higher returns one needs to take on more risk in the form of volatility. Basically higher returns goes in pair with higher variance. Meanwhile these two guys come out with a study showing there’s no link between volatility and returns as measured by MPT. Could it be that most of what we learn in business school is plainly wrong?
Even by my own irreverent standards, this would be a bold statement. Of course if you conduct hundreds of similar studies testing a link between volatility and returns, you are bound to end up with a small number of them telling you there’s no significant correlation. That’s how statistics work. But the fact is that the body of such studies is small and Baillie and DeGennaro are far from being alone in their conclusion. If you Google “studies volatility return link”, you find yourself with a goldmine of papers pointing towards the same thing. Some of them even find a negative correlation between variance and mean returns (in MPT parlance this would imply high risk, low returns). Here a short list of what you can find out there in addition of Baillie and DeGennaro’s research²:
- Dimitriou and Simos from the University of Ioannina in Greece measured the relationship between returns and volatility in more than a dozen markets around the world. The conclude that the link is weak and even find a negative relationship in some cases.
- Van Vliet, Blitz and Van Der Grient from Robeco Asset Management show how the results from different studies looking at the correlation between risk and volatility can be contradictory depending on what methodological methods are used.
- Qi Li, Jian Yang, Cheng Hsiao and Young-Jae Chang respectively from Texas A&M University, Prairie View A&M University, University of Southern California and Inje University look at the 12 largest international markets and find a significant negative relationship between volatility and return in six of them.
I could go on but you get the point.
Assuming the relationship between volatility and returns is really weaker than previously thought, that means Markowitz, CAPM and APT don’t hold in the real world of investing. Same goes for Beta and SML. Grossly, it means investing and finance has (almost) nothing to do with volatility. How crazy is that? It’s the equivalent of saying gravity isn’t caused by the curvature of spacetime caused by the uneven distribution of mass and energy. (If you don’t understand this last sentence, fear not being alone. That’s exactly how I feel in Portfolio Management 3102.)
Now, if you’re going through the same thought process as me, you’re probably asking yourself: “Should volatility be completely ignored?”
My answer is yes, if we’re all emotionless robots. But we’re not, and human nature being what it is, we instinctively react to volatility. We know that thanks to Daniel Kahneman and all the other courageous researchers who are currently working towards unbuilding the walls that separate finance and psychology. The fact is we’re all frightened by the idea of seeing our net worth drop by 20% on any given week or month. Money managers need to take that into account when building portfolios.
It’s important to keep volatility in check. But it should never be central to any investing strategy. Volatility is just noise. The fact that share prices change on any given day has nothing to do with the value of the underlying businesses, which should be the only proxy used to evaluate an investment opportunity.
So when managing portfolios of stock and bonds, volatility should be reduced enough so that you avoid the pain of waking up to a 40% paper loss in your portfolio (Valeant, anyone?). This should be enough to ensure that everyone involved keep a certain level of day-to-day sanity. Usually this is achievable by constructing a portfolio of 12-20 great businesses competing in different industries. Past that point, worrying about volatility becomes useless. Charlie Munger has pushed this to the extreme. He’s well known for holding stocks in only 3 or 4 companies at the time. He never cared about day-to-day changes in stock prices. None of this precluded him from becoming a billionaire.
There’s no reliable way to quantify risk in the stock market and my guess is that there will never be. True risk is a qualitative variable. There are too many forces at play in the business environment for us to be able to reduce their effect to a couple of variables.
The lesson we should learn from this is that investors will do much better if they allocate more of their time assessing business fundamentals instead of relying on past price movements to try to predict anything. A failure to do so basically led to the financial crisis of 2007-08.
Let’s stop kidding ourselves. Sharpe ratios and Treynor measures don’t mean anything. We have scientific evidence of that.
P.S. If you think I’m crazy, I highly encourage you to read this essay from Warren Buffett about how mastery of modern finance is unnecessary to achieve above-average investment returns : http://www8.gsb.columbia.edu/rtfiles/cbs/hermes/Buffett1984.pdf
¹Richard Baillie is an A.J. Pasant Professor of Economics and Finance at Michigan State University. Ramon DeGennaro is a Professor in Banking and Finance at the University of Tennessee, Knoxville. Here’s the link to their paper: https://www.msu.edu/user/baillie/JFQA.1990.pdf
²Links to papers listed:
Dimitriou and Simos: http://www.scirp.org/journal/PaperDownload.aspx?paperID=3893
Van Vliet, Blitz and Van Der Grient: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1881503
Qi Li, Jian Yang, Cheng Hsiao and Young-Jae Chang: http://www.geocities.ws/jian1_yang/published/JEF05.pdf