Is Low Net Actually Low Risk?

29 Oct Is Low Net Actually Low Risk?

Some Common Characteristics among the 3 scenarios.

In previous posts, I presented 3 equity long/short portfolios constructed based on earnings, book value, and momentum.  The objective was to show that a portfolio with zero net exposure (or dollar neutral) can have realized volatility that is higher than the same portfolio with non-zero net exposure.  In each example, the lowest risk portfolio has a non-zero net exposure.

Beyond the net exposure, there are a number of characteristics that are common among the examples.  First, performance increases and risk decreases for the low volatility portfolios compared with the base case portfolios.  Second, the low volatility portfolios experience less frequent and less severe drawdown events.  Third, the low volatility portfolios have greater risk-adjusted returns. Finally, the low volatility returns are more sensitive to changes in the market return (depending on how we define the market) and less sensitive to large company returns.

Less return variability

As the label implies, the low volatility portfolios are the portfolios with the lowest standard deviation of monthly returns for each example.  The low volatility portfolios’ risk values tend to be 2/3 of the base case portfolios’ risk values.

The difference in risk between the portfolios is more apparent when viewed over time.  The graphs below depict the cumulative variances for the base case and low volatility portfolios for each example.  The low volatility variance (in red) is consistently lower than the base case variance (in blue).

Better return profile

The low volatility portfolios have better return profiles.  Without passing judgment on the merits of the base case portfolios, the increase in performance from base case to low volatility is clear and unambiguous.  Consider the following graphs of the natural log of cumulative returns:

Fewer, shorter and less severe drawdown events

The low volatility portfolios have fewer drawdown events and the events do not last as long.  In addition, the maximum drawdown events are not as severe.

Higher risk-adjusted returns

The low volatility portfolios have better risk-adjusted returns – Sharpe and Sortino ratios.  This is not surprising given the increased return and coincident lower volatility.  The Sharpe and Sortino ratios for the base case and low volatility portfolios are included in the table below.  The Minimum Acceptable Return for the Sortino ratio is the annualized time-weighted monthly return to the 3-month t-bill over the period.

More sensitive to the broad market with some size effect

The low volatility portfolios are more sensitive to the market and show some hint of the size effect.  In each example, the low volatility portfolios are more sensitive to the market; however, this result is conditional on the choice of the market index.  In the foregoing analyses, the market is a value-weighted index that includes all U.S.-listed equity securities on the NYSE, AMEX or NASDAQ. In practice, the S&P 500 is the market index of choice for performance and risk comparisons;  it is a value-weighted index of 500 large companies listed on the NYSE or NASDAQ.

The S&P 500 results differ from the broad market index results presented below.  As I mentioned in earlier posts, the choice of index appears to matter.  When I repeat the foregoing analyses substituting the S&P 500 as the market index, the base case and low volatility portfolio returns have little sensitive to the market. In fact, the sensitivities differ little between the base case and low volatility portfolios.  Because the S&P 500 is used most commonly in practice, this result is more than interesting; it is more relevant.

Another characteristic that is common among the low volatility portfolios is the lower sensitivity to large company returns (Note: the size factor return is the small company returns minus large company returns). The portfolios become more neutral or slightly positive on the size factor.

Conclusion

The low volatility portfolios produce risk-adjusted returns that are better than the base case portfolios.  The improvement is due to a number of factors including a better performance profile, less variability in monthly returns and a better drawdown profile. While the improved performance is welcome, it is clear that the low volatility portfolio returns could be more sensitive to common factor risk, specifically market return and size, than the base case (non-zero net exposure) portfolio returns.  However, this result is contingent on the choice of market index. Nonetheless, the lower sensitivity to large company returns is a redeeming trait in light of the return premium associated with small company returns.

The next posts will highlight the results using the S&P 500 as a market index and will evaluate how the results vary over time.