Is Low Net Actually Low Risk? A “Buy Strength, Sell Weakness” Example.

28 Sep Is Low Net Actually Low Risk? A “Buy Strength, Sell Weakness” Example.

In previous posts, I presented portfolio examples based on earnings and book value to show that net exposure has risk management benefits relative to zero net exposure. This post will make it a trend with a third portfolio example based on momentum stocks – i.e., buy strength and sell weakness. The conclusion is the same: the lowest risk portfolio has a net exposure greater than zero. In addition, the sensitivity to the market increases.

A Momentum Stock Portfolio Example

I created an equity long/short portfolio using “Momentum” monthly return data from the Kenneth French data library. The long portfolio is the top decile of securities ranked by lagged prior 12-month return and the short portfolio contains the bottom decile of securities. The securities are equal-weighted in the long and short portfolios.

The dot plot below shows that the portfolio with net exposure (“nee”) equal to 45% has the lowest volatility of all scenarios when varying nee from -100% to 100%. The volatility declines from nee = -100% to nee = 45% then increases as nee approaches 100%.  Interestingly, the volatilities at nee = 100% and nee = 0% are very similar.  The difference between the two scenarios lies in the higher Market Beta and higher performance at nee = 100%.

The volatility of the portfolio at nee = 45% (“low volatility portfolio”) is two-thirds the volatility at nee = 0% (“base case portfolio”) and roughly equal to the volatilities of the broad market and the S&P 500 indices.

The cumulative performance and drawdown plots of the base case and low volatility portfolios (less cash) show that the low volatility portfolio outperforms with shallower drawdowns. The losses were less severe and shorter in duration.

The lower volatility and improved performance of the lower volatility portfolio result in higher risk-adjusted returns. As volatility decreases, average performance increases to 11.07% from 5.89%. Consider the following table containing a couple of risk-adjusted return measures:

The market beta increases to 0.47 from -0.14. This is a sizeable increase and highlights the issue of how much market beta is too much. In this case, the market volatility is much lower than the base case portfolio’s volatility.  If I designate that the base case volatility is the inherent volatility for the momentum example, then this trend shows that it may be necessary to increase market exposure to reduce the volatility.

How much of the low volatility portfolio performance can I attribute to the increased market beta? I estimate that the increased market beta accounts for 67.67% of the performance increase – the product of the change in market beta (0.61) and the market return (6.43%) divided by the change in portfolio performance (5.81%). The value is 0.89% assuming that the S&P 500 is the benchmark. Similar to previous examples, the index choice is important when gauging market sensitivity.

Minimizing Drawdowns

In addition to lowering volatility, minimizing drawdown events – either in number or magnitude – is a gauge of the effectiveness of a risk management scheme. The Calmar ratio (or Drawdown ratio) comparison below shows that the low volatility portfolio dominates – outperforming 80% of the time with an average spread of 0.91. When the line is above zero, the low volatility portfolio’s ratio is greater (or better) than the base case portfolio’s ratio.

Interestingly, the base case portfolio outperforms, based on the Calmar ratio, from Dec. 2007 to Sept. 2011 with the greatest outperformance in the first month of the period. But, the base case drawdown was greater than the low volatility drawdown. The reason may be that the low volatility portfolio performance turned negative sooner. The low volatility portfolio performance caught up after 17 months and continued to outperform for the remainder of the observation period.  Regardless, the low volatility portfolio performance exceeds the base case as the time horizon increases.

To summarize, I have shown that you can increase net exposure while increasing risk-adjusted return with lower volatility and higher return. In addition, I show that a minority of the performance increase is due to increased market exposure.

Common Factor Risk

As in earlier examples, I seek to understand some of the drivers of performance by regressing the returns over the following Fama/French benchmark factors: Market, Value/Growth, Size, and Momentum. I included the momentum factor to maintain consistency with previous analyses expecting that the factor sensitivity would be high. The market return is the value-weighted US return in excess of cash.

The t-statistics significant at the 1% level are as follows:

Note: Number of observations = 762

The adjusted R² for the nee = 0% and the nee = 45% are 0.72 and 0.83, respectively.

There are a several notable changes to factor sensitivities. All factor sensitivities increase except for momentum and unexplained return becomes negative. Like the earnings and book-value examples, the low volatility portfolio’s sensitivity to the size factor increases.  The low-volatility portfolios are highlighting the well-documented “size effect” – small firms have higher risk- adjusted returns than large firms.

The same analysis applied to a value-weighted version of the momentum portfolio produces similar results. The lowest volatility portfolio is at nee = 55%. Volatility and returns improve resulting in better, consistent risk-adjusted performance.

Regarding the common factor analysis, exposure to the market and size factors increases along with marginally higher adjusted R², 0.86 compared with 0.84.

The t-statistics significant at the 1% level are as follows:

Note: Number of observations = 762

Up & Down Markets Performance
In performance terms, an ideal long/short portfolio would generate positive performance in positive and negative markets. An improvement on this ideal portfolio would be one that captures a larger portion of the upside in a positive market. In an interesting turn, the low volatility portfolio in this momentum example does not fit this ideal, but the base case portfolio does. In fact, the base case portfolio captures only 16% of the broad market up return and none of the broad market down return (I observe similar results for the S&P500). On the other hand, the low volatility portfolio captures 75% of the up market return and 35% of the down market return.

Conclusion

A long/short portfolio constructed by investing in momentum stocks has a lower risk profile at nee > 0%. This aligns with the conclusion in the previous examples. The key observations have been:

1.  A portfolio with net equity exposure different from zero has a lower volatility profile than the same portfolio at nee = 0%
2.  Changes in other common factor exposures may be just as relevant as the changes in market exposure
3.  The size effect is observable in the lower volatility portfolios

I will use the next post to identify commonalities among the earnings, book value and momentum examples.