Get a FREE subscriber login and read more articles

The low volatility anomaly - miracle or mirage

(An article from Equity Jul 2013 by Dominic McCormick)

The “low volatility” or “low beta” anomaly seems to be the new Holy Grail in investment markets. This is the premise, well documented by studies of historical returns in most major sharemarkets, that low volatility or low beta stocks outperform high volatility/high beta stocks over the long term. One study of US stocks between 1968-2008 showed that stocks in the bottom 20% based on past volatility outperformed those in the top 20% by a compounded return of around 11.2% per year.  (Baker, Bradley and Wurgler: Benchmarks as Limits to Arbitrage, Understanding the Low Volatility Anomaly 2010).

Of course, this is not what the Capital Asset Pricing Model (CAPM) taught us.  Beta is not only dead but instead seems to work in totally the opposite way that Modern Portfolio Theory suggested. High risk stocks (whether defined by standard deviation and/or beta) seem to actually exhibit a lower return than low risk stocks over the long term.

What explains this anomaly?  The academics have come up with a range of explanations relating mainly to behavioural factors or industry structural constraints. One focuses on the tendency of investors generally to chase volatile stocks.  This is the so called “lottery effect” where investors are willing to overpay for more volatile stocks because of the chance/perception of higher returns. Overconfidence is another behavioural bias that comes into play here. Related to this is the view that mutual fund investors chase recent performance, which pressures their fund managers to hold high beta stocks, particularly when they have to offset their small cash holdings held for redemptions.  

Another explanation is that too much of the rational institutional investor universe is focused on benchmarking to traditional capitalisation weighted indexes and therefore unable to take much advantage of this anomaly. Focused on outperforming their particular benchmark they have little incentive to overweight less volatile stocks and sectors because it expands their tracking error greatly and increases their chance of significantly underperforming that benchmark in the short term.        

Although this anomaly has been studied and well known for decades it is only in recent years that fund managers and investors have moved to explicitly exploit it.   A range of managers have set up low volatility or minimum variance funds and even a range of ETFs have been set up.  According to Ned Davis Research total assets in low-volatility ETFs have grown tenfold in the past 15 months to just over $10.4 billion. A number of funds have also been launched in Australia recently.  Many of these funds are not aiming to outperform the market but to provide around the market return but with lower risk. The dramatic growth of Risk Parity funds is also relevant as they may also incorporate the low volatility phenomena for their equity components. 

There has been much discussion in recent years of the benefits of building less volatile portfolios, particularly given the “sequencing risk” for those in retirement and drawing down their capital.  Given this are low volatility equity funds something institutional and retail investors should be including in their portfolios? In my mind there are a range of issues to consider.

  • Are investors in low volatility equity funds simply loading up on sectors/assets that they already have significant exposure to in a well diversified portfolio?
  • Are they already benefiting from this anomaly through some of their better active equity fund managers?
  • Has the anomaly been accentuated by the GFC and post GFC period and may be less prominent in coming years? i.e. Have low volatility stocks become too expensive?  
  • Is this anomaly likely to be “arbitraged” away now that it is being widely exploited? 

I will touch on each of these points below.

One problem I see is that two of the prominent “low volatility” sectors that dominate in sharemarkets, and particularly narrow markets like Australia, are infrastructure/utilities and property/AREITa (despite the latter’s high volatility through the GFC). These sectors may well make sense in a total portfolio but typically, in Australia at least, investors often already have quite significant explicit exposures to these “assets” through infrastructure and AREIT fund/manager allocations.  For example, one recently launched Australian low volatility fund currently has an allocation of almost one third to these categories and they dominate their top holdings.

If a major “advantage” of low volatility funds is just giving you large exposures to these particular sectors you need to ask whether you are not just as well or better off with exposure through active sector specialists in these sectors/asset classes’ who know these sectors well, or alternatively to use lower cost passive exposure.    

This doesn’t negate the concept of low volatility but suggests that selecting from the widest possible universe where there is a diverse range of sectors to build sensible diversification and reduce such sector overlaps.  This suggests that low volatility global equity funds make more sense than narrower country based ones.          

Another valid question is whether the active managers you are using are already benefiting from this anomaly, perhaps not explicitly but as part of their search for undervalued stocks? Low volatility stocks typically have value characteristics, pay higher dividends, are large in size and quality focused. In particular, good active managers are good at avoiding high volatility, overhyped stocks and sectors.  After all, to a large extent this is what Warren Buffett does – although also effectively gearing this portfolio through the benefit of his insurance float.  There are a number of prominent active global managers offering funds in Australia that implicitly have this low volatility element as part of their strategy.  Of course, these funds come with higher fees than many of the new low volatility funds.  

Another issue is whether low volatility has overly benefited from the risk aversion generated by the GFC. The risk is that some excess returns have been frontloaded into this period as investors have shied away from heavily geared, particularly cyclical and risky businesses.  Perhaps many of these “safe” low volatility stocks are now expensive versus history suggesting that future absolute and relative returns may be subdued. While the evidence seems to suggest that the anomaly has worked over extended periods through bull and bear cycles it is clear low volatility stocks can perform poorly for extended periods.  I suspect that an element of the recent good performance of, and increased demand for, low volatility funds reflects the move to defensive stocks in response to the GFC. This in turn has further improved the marketing “story” for these low volatility funds.       

Finally, is this anomaly ultimately likely to be exploited away for good? I think this is likely to some degree at least.  Whenever the investment industry comes up with a new concept, particularly one that has caught on as quickly as low-volatility, it pays to be sceptical.   As the growth of low volatility funds and ETFs, as well as use in risk parity portfolios, is showing this is one anomaly where money is easily able to move to exploit it.  Passive volatility funds don’t care what price they pay for that low volatility.   Even the broader movement to loosen manager benchmark constraints on many active equity managers will be a factor lessening this anomaly. 

Perhaps the key lesson in the volatility anomaly is not so much the merits of low volatility stocks themselves but rather the dangers in chasing high volatility stocks.  The Baker, Bradley & Wurgler paper mentioned above showed that between January 1968 and December 2008 investors who put $1 in high volatility stocks ended up with just 58 cents but when the inflation effect over the period (eroding $1 to 0.17 cents) was incorporated one ends up with less than 10 cents in real terms – i.e. a loss of more than 90% in real terms. In contrast $1 in low volatility stocks was worth $59.55 or a $10.12 gain in real terms.

While money is going into long only low volatility equity funds and ETFs, and this may end up eroding the anomaly for these funds, there are relatively few funds that are specifically shorting high volatility stocks so it is difficult to see how the behavioural characteristics that result in many investors chasing volatile, overpriced growth stocks are going to be eliminated any time soon.  

Whether one invests in the new specialist low volatility funds or not, tracking the persistence or demise of, and considering the on-going drivers for, the low volatility anomaly is well worthwhile for those building investment portfolios.  Doing so can help ensure that portfolios are positioned to benefit from any persistence in this anomaly but, perhaps more importantly, highlight the risk of potential losses from poorly considered or excessive exposure to high volatility stocks. 

Dominic McCormick is the Chief Investment Officer at Select Investment Partners. This article was first published Money Management in April 2013.     


There are no comments. Be the first to post