Systematic Relative Strength

Tail risk, or sometimes hedging tail risk, has been a hot topic lately. Tail risk is essentially the song of the black swan—it’s what happens when a negative event with a 1% probability happens. Ever since 2008, concern of being caught in another large market decline has caused investors to be very aware of tail risk. Tail risk has also engendered some very interesting exercises in portfolio construction, with all manner of alternative assets.

Consider, however, this from Javier Estrada’s recent paper, Rethinking Risk:

The evidence discussed here, based on a comprehensive sample of 19 countries over 110 years, suggests that investors that focus on uncertainty are likely to view stocks as riskier than bonds, and those that focus on long‐term terminal wealth are likely to view stocks as less risky than bonds even if they are concerned with tail risks. This is the case because, even when tail risks do materialize, investors are more likely to have a higher terminal wealth (that is, more capital at the end of the holding period) by investing in stocks than by investing in bonds.

In other words, a lot of your definition of risk depends on whether you view risk as uncertainty (volatility, standard deviation) or you are focused on terminal wealth—how much money you have at the end of the day.

Here’s another important excerpt:

Obviously, it is for nobody but the investor himself to say what lets him sleep at night. That being said, Charlie Munger, Warren Buffett’s longtime partner at Berkshire, has some advice for investors in the setting described. In fact, Munger (1994) argues that if “you’re investing for 40 years in some pension fund, what difference does it make if the path from start to finish is a little more bumpy or a little different than everybody else’s so long as it’s all going to work out well in the end? So what if there’s a little extra volatility.”

To be sure, there may be long‐term investors that simply cannot help being concerned with, and react to, the short‐term fluctuations in the value of their portfolios. And if that is the case, short‐term volatility is how they assess risk and little of what is discussed in this article may be relevant to them. That being said, would it not make sense to at least also worry about whether a conservative strategy will, by the end of the holding period, enhance purchasing power or underperform an aggressive strategy by a wide margin?

Granted, an investor may be fully aware that a conservative strategy is likely to underperform an aggressive one and still be happy with choosing the former if he is concerned with tails risks, such as a big loss close to the end of the holding period, or a holding period of very low stock returns, however unlikely they may be. And yet, should not this investor also consider whether he will be better off (that is, end with a higher terminal wealth) by pursuing an aggressive strategy even in the case that tail risks do materialize?

That last part is critical. It is clear that the longer is the holding period the more likely is a riskier strategy to outperform a less risky one; that is, in fact, what theory suggests and what the evidence shows. And yet some investors may stay away from an aggressive strategy simply out of fear of tail risks without grasping that, even if these risks do materialize, their terminal wealth is likely to be higher than with a conservative strategy.

Pretty interesting stuff. The author notes that there are times and conditions when concern with volatility could dominate, but if you are talking about a long investment period, the data shows that being overly conservative can impact terminal wealth more negatively than tail risk. (I added the bold.)

This is just another way to point out that the costs you can measure (volatility, drawdown) are often swamped by the cost you cannot measure directly—opportunity cost.

Despite being embedded into Modern Portfolio Theory, volatility might not be the only kind of risk that matters. (In fact, I’ve pointed out that one handy use of volatility is to add on dips. Volatility can be harnessed productively in certain situations.) I am not suggesting that risk management be tossed aside, but if your primary concern is terminal wealth, you need to think about portfolio construction in a nuanced way.

One of the reasons that we like to use relative strength as a return factor is its precision, its lack of assumptions, and its universality–basically its ability to rank even very disparate assets.

Although value managers often discuss assets or stocks that are undervalued, there is little agreement on how value works. Take, for example, a Bloomberg story from today, entitled U.S. Stocks Cheapest Since 1990 on Analyst Estimates. With a title like that, you would expect an article that discusses how cheap stocks are relative to earnings. And, indeed, the article cites such a source:

“The stock market is incredibly inexpensive,” said Kevin Rendino, who manages $11 billion in Plainsboro, New Jersey, for BlackRock, the world’s largest asset manager. “I don’t know how the bears can argue against how well corporations are doing.”

Not only are stocks cheap, but analysts seem reluctant to buy in, something that strikes me, from a contrary opinion point of view, as unabashedly positive.

While analysts are raising estimates, they’re not boosting investment ratings. Companies ranked “buy” make up 30 percent of all U.S. equities, the data show. That compares with 45 percent in September 2007, a month before the S&P; 500 reached its record high of 1,565.15 and began a 17-month plunge that erased $11 trillion from the value U.S. shares.

But hold on a mintue– here is the part I find most fascinating about the article: no one can agree on the valuation. (I guess the tendency to cover both sides of the story in the same article comes from political opinion coverage, where the media thinks they have to give equal time to both sides, whether rational or not.)

David Rosenberg, chief economist of Gluskin Sheff & Associates Inc., says U.S. stocks are poised for losses because they’ve become too expensive. The S&P; 500 is valued at 22.1 times annual earnings from the past 10 years, according to inflation-adjusted data since 1871 tracked by Yale University Professor Robert Shiller.

So while you have reputable analysts arguing that the market is cheap, you have equally reputable analysts arguing that the market is expensive! How does that happen? Well, it happens because everyone in the value camp operates from a (sometimes radically) different set of assumptions. Are you using trailing estimates, forward estimates, or 10-year normalized estimates? And, by the way, they are estimates.

Source: DenverPost.com

Relative strength is nice because it is precise. No one argues about what the trailing 6-month or 12-month return was. It’s cut and dried. Although different firms use slightly different relative strength calculations–Dorsey, Wright Money Management has a proprietary method–it is hard to imagine a scenario where there would be much disagreement about the strength or weakness of the same market.

Which is stronger, U.S. stocks or U.S. long-term bonds? Just looking at a comparative chart of the two assets makes it obvious, even to someone not sophisticated in quantitative analysis (SPY vs. TLT).

Click to enlarge. Source: Yahoo! Finance

On the other hand, determining which asset is a better value is quite indeterminate. It all depends on whether you take as your starting point the view that the U.S. stock market is cheap or expensive!

And what happens when you have to value a multi-asset portfolio? How do you determine whether real estate, international stocks, euros, or emerging market bonds are cheaper? The lack of agreed-upon metrics between markets, let alone within an individual market, make this kind of decision a nightmare from a valuation perspective.

Since prior performance is precise, relative strength measurements can be made easily and without argument. Relative strength can easily be used for multi-asset portfolios as well. And finally, the returns from using the relative strength factor are high–as high as the returns from the value factor, and in many studies, higher. Luckily for us, efficient-market types normally decline to use relative strength because theoretically it shouldn’t work! Their loss is our gain.

—-this article originally appeared 4/26/2010. The broad market is up about 40% since this article was written and the debate about valuations is still raging! That alone should tell you that trying to figure out who is right—and whether to invest or not—is not a productive use of time. Relative strength simply indicates what the strongest markets are at any given time.