The Limits of Low Volatility and the Deviancy of Standard Deviation

Low volatility alone leads to no portfolio diversification benefits

Originally published in the September 2015 issue

Launched in 2009, the Argonaut Absolute Return has achieved a positive return in every calendar year to date, with the price of the fund now up more than 100% since launch.[1] Whilst these returns are intrinsically attractive, investing in Absolute Return should also be about delivering attractive returns combined with a risk profile which differs from traditional long-only funds and therefore provides portfolio diversification.[2] We think that risk profile is more than just standard deviation and that the debate around Absolute Return is overly focused on low volatility characteristics. We would point out that low volatility by itself has no portfolio diversification benefits. We are concerned that focus on low volatility in isolation can often lead to too many funds that can only deliver cash-plus returns without the safety of cash or the cost-free fee structure of cash as an asset class. Far better, we think, to view volatility as necessary to achieve a stated target return and to measure return, not just against volatility, but also against correlation to risk assets, which is far more important in constructing a diversified portfolio.

The trade-off between risk and return is well known. A fund manager’s skill in converting units of volatility into units of return is most often expressed using a Sharpe ratio,[3] which assumes a linear trade off (units of excess return are divided by units of risk). The higher the Sharpe ratio the more attractive the risk-adjusted return. Any fund manager who consistently achieves a Sharpe ratio in excess of one (whereby units of return exceed units of risk) is commonly regarded as a skilful converter of risk into return. As such, a fund manager who consistently delivers a consistent excess return of 2% with 2% standard deviation is as meritorious as a fund manager who delivers a consistent excess return of 12% with 12% standard deviation. Despite the different risk and return of the two funds, both fund managers have delivered the same Sharpe ratio. The range of annualized returns and volatility of the funds within the Targeted Return sector is illustrated by Fig.1. Since launch, for example, the Argonaut Absolute Return fund has achieved annualized rate of return of 11% with a realized volatility of 9%, (which after deducting the risk-free rate from the annual rate of return[4]) corresponds to a Sharpe ratio of 1.15, and has been amongst the better funds at converting volatility (as measured by standard deviation) into returns.

The most common objection to use of the Sharpe ratio is that it fails to distinguish between downside volatility (generally considered undesirable) and upside volatility (which seldom draws complaints from investors). The modification of the Sharpe ratio to take into account only downside volatility is the Sortino ratio.[6] Again, the higher the Sortino ratio the more attractive the risk-adjusted return. The range of annualized returns and downside volatility of the funds within the Targeted Return sector is illustrated by Fig.2. Since launch, for example, the Argonaut Absolute Return fund has achieved annualized rate of return of 11% with a realized downside volatility of 8% (which after deducting the risk-free rate from the annual rate of return[7]) corresponds to a Sortino ratio of 1.4.[8] So stripping out its upside volatility (not generally thought of as undesirable) and focusing on undesirable downside volatility only, we can see that the fund has been even more efficient at converting downside risk into return, than just focusing on all volatility would suggest, with a Sortino ratio of 1.4, against a Sharpe ratio of 1.1.

Although we think we are efficient converters of risk (particularly downside risk) into return, we do not believe in marketing our Absolute Return fund as a “low volatility” fund. We target an annualized (ex-anti) standard deviation of 8% (although our ex-post volatility since launch has been on average nearer 9%). We believe that this has been necessary volatility to enable us to achieve our annual target return of 5-8%. Whether this volatility is regarded as “high” or “low” relative to the market or the peer group is of secondary importance. One of the difficulties in marketing a fund as “low volatility” is that this definition of volatility is itself most often defined relative to market volatility, which is never constant. As we can see from Fig.3, when we launched, market volatility was well over 20%, so that our fund volatility was just a third of the market volatility. More recently market volatility has dropped to 10-15%. So the fund has gone from only having approximately one third of market volatility to two thirds. In order to maintain consistent volatility relative to the market, we would have to target lower volatility and therefore lower returns. As such marketing a fund as “low volatility” makes no sense unless we were to consistently adjust our target return. Far better to market the fund with a stated targeted return and take appropriate volatility to achieve that return.

What strikes us about the Target Absolute Return sector is the number of funds whose volatility implies that they can only be targeting “cash plus” returns with approximately half of the sector displaying a standard deviation of less than 4% (see Fig.4). But are “cash plus” returns worth the extra bother? Cash is the ultimate low volatility investment. Unlike low volatility funds, cash never delivers a negative absolute return (unless of course the bank where you have deposited your money becomes insolvent), so that the risk/return analysis of cash as an asset class is always consistently strong, simply owing to the absence of risk. But by far the biggest limitation of “cash plus” low volatility funds is that their low volatility by itself offers no diversification benefits. An investor might believe that by selecting a dozen low volatility funds they are achieving diversification only to find that all of the funds selected are highly correlated to the stock market (and/or each other) and therefore achieve no or only limited diversification. The investor could replicate the same return profile by splitting his assets between cash and the market index, which would probably be cheaper to replicate after fees.

This brings us nicely onto the fee structure of low-volatility “cash plus” products. Whereas cash doesn’t charge an annual management fee (or performance fee), this is not the case with the universe of low volatility funds. In fact we wonder whether it is appropriate for low volatility funds to apply annual management charges (AMC) which although are commonplace in the industry as a whole would seem rich for funds targeted such meagre returns. If for example a fund has a standard deviation of 2%, is it really appropriate for it to apply an annual management charge of 75bps, when this implies that even in a year where the fund manager has demonstrated skill in converting risk into return (with a Sharpe ratio of 1) that this charge would likely conserve nearly 40% of the targeted return? Our analysis across the Targeted Return sector (Fig.5) suggests that approximately one quarter of the funds have an AMC, which is at least half of their standard deviation, implying that even in a good year fees will eat away at least half of the implied return. This begs the question as to whether low-volatility Targeted Absolute Return products offer value for money, as their fees would seem to consume an unhealthy proportion of the targeted return.

We think that the most attractive attribute risk characteristic of an Absolute Return fund is not low-volatility but low correlation[9] to risk assets and ability to deliver returns in all market environments. After all, the least difficult skill in investing is delivering a positive return at the same time as the market and everyone else. The most difficult skill in investing is delivering a positive return at different times to the market and everyone else, whilst still delivering an attractive return profile overall (given the tendency of stock markets to deliver positive returns over market cycles). This requires investing in a non-correlated asset (in our case a short book of equities) and having some skill in generating short alpha as well as long alpha. It also requires net exposure to never get too aggressive. Our analysis of the Targeted Absolute Return sector (Fig.6) shows that only three funds have a negative correlation to the European stock market but that the overall return record of these funds is poor. By contrast over half of the funds in the sector a correlation of over 0.5 with the European stock market, which suggests limited diversification benefits. There are only two funds – including the Argonaut Absolute Return Fund – that would appear to combine an attractive return profile with a low correlation to the European stock market.

Whilst low correlation to risk-assets is desirable, rather like the Sharpe ratio with upside volatility, it is rare for investors to complain about correlation to bull markets. As such a fund’s ability to perform in negative market environments in isolation is best measured by its downside capture ratio[10] with the figure illustrating the percentage of the market drawdown typically suffered by the fund and with a negative figure indicating overall an ability to generate positive returns for the fund in negative market environments. Over the past five years, our analysis of the Target Absolute Return sector (which must come with the caveat that the sector is also full of multi-asset and bond funds whose correlation to the European stock market should not be as high as equity-based funds in any case) suggests that 30% of funds have generated a negative downside capture ratio, with this figure rising to 40% over three and one years. Only two funds in the sector (of which the Argonaut Absolute Return is one) since our launch have combined a downside capture ratio in excess of 10% with an annualized positive return in excess of 10% (Fig.7). It is also important to point out that ability to generate positive returns in negative market months must also mean ability to generate negative returns in positive months. We are also not arguing that Absolute Return funds should derive their returns equally from negative and positive months for the stock-market as given the tendency of markets to rise over an economic cycle, this would likely lead to unattractive return profiles overall.

The IMA Target Absolute Return sector is composed of a variety of different types of fund: multi-asset, bond and long/short equity. Whilst this means different risk and return profiles, the framework for analysing the attractiveness of their profiles is the same. Return needs to be analysed relative to risk but solely measuring risk in units of standard deviation, which is too common is crude and misleading. Moreover, low volatility (as measured by low standard deviation) by itself has no diversification benefits and low volatility funds can offer poor value for money given their fee structures. So whilst returns need to be analysed relative to risk, bear in mind the deviancy of standard deviation and the limitations of purely focusing on low volatility.


  1. Asness, C. S., T. J. Moskowitz, and L. H. Pedersen (2013). Value and Momentum Everywhere. Journal of Finance 68 (3), 929-985.
  2. Cochrane, J. 2005. Asset Pricing Theory. Princeton University Press. Revised Edition.
  3. Duffie, D. 2001. Dynamic Asset Pricing Theory. Princeton University Press, Princeton.
  4. Fama, E. F. and K. R. French. 1993. Common Risk Factors in the Returns on Stocks and Bonds. Journal of Financial Economics 33 (1), 3-56.
  5. Harvey, C., Y. Liu, and H. Zhu. 2013. … and the cross-section of expected returns. Available at SSRN 2249314.
  6. Kan, R. and G. Zhou (2012). Tests of Mean-Variance Spanning. Annals of Economics and Finance 13 (1), 139-187.
  7. Martellini L. and V. Milhau, July 2015, Factor Investing, EDHEC-Risk Institute Publication produced as part of the Lyxor “Risk Allocation Solutions” research chair at EDHEC-Risk Institute.
  8. Merton, R. 1973. An Intertemporal Capital Asset Pricing Model. Econometrica 41 (5), 867-887.
  9. Ross, S. 1976. The Arbitrage Theory of Capital Asset Pricing. Journal of Economic Theory 13 (3), 341-360.
  10. Sharpe, W. 1963. A Simplified Model for Portfolio Analysis. Management Science 9 (2), 277-293.
  11. Sharpe, W. 1964. Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk. Journal of Finance 19 (3), 425-442.