U.S. Markets close in 6 hrs 26 mins

When Low Volatility Bites Back

Elisabeth Kashner


As my colleague Ugo Egbunike and I pointed out in our blog "Why Low Volatility Is Losing Its Alpha," SPLV had a bad week on the heels of serious corrections in high-yielding utilities stocks.

Returns of the PowerShares S'P 500 Low Volatility Portfolio (SPLV) versus those of its parent index, the S'P 500, suddenly look unappealing. After a long run of outperformance, low-volatility funds have fallen out of bed with a thud.

All Low-Vol fund returns through 5-30-13


Sector exposure mostly explains SPLV's fall, but it doesn't fully explain the drop many of the other funds.

The iShares MSCI USA Minimum Volatility Index Fund (USMV) was hit almost as hard by its financials as it was by its utilities. The SPDR Russell 1000 Low Volatility ETF (LGLV) suffered partly because of its allocation to consumer non-cyclicals. But the causes are broader than sector allocation.

USMV, which selects from a large-to-midcap universe, saw weighted average returns equivalent to the 2,024 th -worst security in the Russell 3000.

LGLV, which selects 100 of the lowest-volatility stocks from the Russell 1000, was overrepresented in the losers' column, with the average performance rank of its selections coming in at 625 out of 1,000, rather than the 500 that would be suggested by chance.

Of last month's 100 worst performers in the Russell 1000, LGLV held 18, which is twice what would be expected by chance.

Low-volatility investing became popular only recently, after investors saw the resilience of low-volatility strategies during the Great Recession. Despite years of rising markets, investors have been seeking defensive strategies, and have been strangely rewarded by low vol in a market environment that ought to favor risk-taking. What's going on?

Let's explore the following arguments:

  1. Much of low volatility's outperformance isn't "real" in the sense that it fails to reach statistical significance. The so-called anomaly is hooey. Volatility metrics don't account for price direction or for "ump-risk," the term for sudden, unheralded price moves. If the outperformance wasn't "real," then its disappearance was inevitable.
  2. Low volatility traditionally underperforms during market bubbles. Market watchers, take heed.
  3. The "discovery" of low volatility has ruined everything. The anomaly has been eliminated by heavy buying in the low-vol space.
  4. Low volatility rewards investors during time of market stress, when volatility is high. Today's volatility is very low.

Is the anomaly real?



Since the 1970s, accelerating after the launch of SPLV, just over two years ago, low-vol proponents have been citing academic studies that support the so-called low-volatility anomaly—namely that, contrary to financial theory, securities with low risk (as measured by historic volatility) deliver high returns over time, while securities with high risk deliver low returns.

They have documented, though not completely explained, that low-volatility funds have had far higher risk-adjusted returns than their cap-weighted brethren—in nearly every size bucket and every country.

Has low vol delivered outsized returns by being less risky, and therefore losing less in bad times, even as it gains less in good times? In other words, has low vol delivered excess risk-adjusted returns?

This chart, which contrasts the rolling 12-month return differential between the S'P Low Volatility Index and the S'P 500 with the statistical significance of the low-volatility version's outperformance (or underperformance), suggests that the answer is "sometimes."

These analyses are quite sensitive to the time periods chosen. Other analyses, using more variables and longer time periods, have come to different conclusions. But in terms of the simple question, "Could I have done just as well by investing 70 percent of my cash in the S'P 500 and the balance in T-bills?", the answer is a definite maybe.

Is Alpha Significant

The telltale sign of a bubble?

The hallmark of low volatility is underperformance during market bubbles, and outperformance when the bubbles burst.

The chart below shows the rolling 12-month returns difference between the S'P Low Volatility Index and its parent, the S'P 500 index.

The low-volatility strategy took it on the chin during the Internet bubble, but delivered in spades after the dot-com bust, only to drop again after the market bottom in 2002. Low volatility was a good refuge during the housing bubble crash, but lagged badly during the initial phase of the recovery.

The recent underperformance of low vol might signal the onset of another market bubble.

Low-Vol Excess Returns

Which bubble is it?



Can we say that, after four-plus years of rising markets, we have somehow crossed a line into bubble territory during the past month?

According to theory, low volatility should underperform during rising markets, as it did for most of the 1990s and during a significant chunk of the housing bubble years. But that's not what's been happening.

Instead, the cumulative performance gap between the S'P 500 Low Volatility Index and the cap-weighted S'P 500 has been widening. You can see this in the chart above and the one below, which shows the cumulative performance of both the S'P 500 and its low-volatility version since October 2006.

So, what explains low volatility's outperformance for much of the past four years?

500 vs low vol:2006 - 2013

The principle "the observer affects the observed" could be at work here.

Simply put, after the housing crash, playing defense is the new offense, and investors have been piling in. Many portfolio managers have hopped on the defensive bandwagon, adding allocations to high-yielding dividend stocks and to low volatility. In the process, a strange thing has happened to low volatility valuations—they're really not cheap anymore.

As of May 2, 2013, U.S.-focused large-cap and total-market volatility funds have posted some chunky price-to-earnings (P/E) ratios:


Fund/Index P/E Ratio
MSCI USA Investable Markets Index 19.63
MSCI USA Large Cap Index 17.48
iShares MSCI USA Minimum Volatility Index Fund (USMV) 19.18
SPDR Russell 1000 Low Volatility Fund (LGLV) 19.80
PowerShares S'P 500 Low Volatility Fund (SPLV) 20.63


Indeed, SPLV's and LGLV's P/E ratios are notably higher than those of their parent index's large-cap universes. To be fair, the Russell 1000 dips well into the midcap space by many definitions, but excludes small-caps. Even compared with the overall U.S. equity market, LGLV looked expensive.

This wasn't always so.



Historically, low-volatility portfolios have carried low P/E ratios—so much so that academics have argued about whether low volatility is a separate anomaly from the value effect, or if they are two sides of the same coin. Could the recent uptick in low-volatility P/Es be a telltale sign of a low-volatility bubble?

Is low volatility, in fact, just low volatility?

Low volatility could just be doing its job, which is to reward investors during volatility spikes.

Take a look at this chart, which superimposes the 12-month rolling difference in returns to the S'P Low Volatility strategy versus the S'P 500, and the trailing 30-day realized volatility of the S'P 500 index:

30-day Trailing Volatility Plus Excess Returns

As of May 28, the trailing 30-day realized standard deviation of closing levels of the S'P 500—a figure commonly quoted to represent market volatility, stood at 9.22 percent, well off its 20-year average of 16.7 percent.

This volatility statistic has been below 10 percent on only 10.8 percent of the days since the October 2007 market peak, while remaining below its current levels of less than 8.0 percent of the days since that peak.

Isn't low volatility supposed to lag when markets are calm? Maybe there is no story here after all.

We have a few theories to bounce around.

Whether low vol's sudden underperformance heralds the beginning of an equity bubble, is a byproduct of today's low-volatility environment—in the U.S. equity markets, at least—or, alternatively, is the kickoff of a long-overdue valuation correction remains to be seen.

If the low-volatility anomaly has been eliminated or reduced because of increased investor awareness, then investors can again expect to be fairly compensated for risk, and will no longer be able to—sometimes—pick up oversized risk-adjusted returns.

Riddle me this:If you expect equity markets to rise, do you want to be in a defensive position?

If you expect them to fall, why not just go to cash?

At the time this article was written, the author held no positions in the securities mentioned. Contact Elisabeth Kashner at ekashner@indexuniverse.com .


Permalink | ' Copyright 2013 IndexUniverse LLC. All rights reserved