Fight-or-Flight Response to Increasing Volatility

02 Feb Fight-or-Flight Response to Increasing Volatility

The first two weeks of 2016 were the worst start to a year for the equities market since the 1930’s.  We have convincingly entered a period of increased volatility which began in September 2015.  When faced with increased volatility, an investor’s natural reaction might be to decrease portfolio net equity exposure to decrease variance in returns.  An analysis of periods of higher market volatility shows this to be a questionable approach.  Net equity exposures of 40 – 50% produced less volatile returns.

Recognizing that increased volatility occurs in clusters, how should equity long/short investors position their portfolios’ net exposures to manage risk in the face of what may be a prolonged period of steadily increasing volatility?

Volatility increases tend to cluster

There were 10 periods of above average 3-year volatility from 1954 to 2014.  The average duration was 3 years with the shortest period being 3 months and the longest being 82 months.  The duration included the time that volatility increased from the long-term average to the duration peak, then decreased to the long-term average.  For our purpose, the increase from the long-term average to the peak, instead of from the previous trough to the peak, is informative because the initial increase from the trough is usually during a period of below-average volatility.

The table below shows that the increase tends to be briefer (and sharper) than the decline.  So, the time to adjust is relatively short but the increase can be significant – the median peak was 15% higher than the long-term average.  Excluding periods from 1954 to 1975 results in a median peak that was 1/3 higher.  So, it is reasonable to assume that average peaks are trending higher.

Zero net exposure may not pay  off

Faced with increasing volatility, many investors may be tempted to adjust their net equity exposures to zero and assume that return risk has been neutralized.  Instead, a non-zero exposure may be a better option to reduce risk inherent in the portfolio (given the weights of each security in the long and short portfolios).  Given a set of long/short portfolios, the table below shows that portfolios with net exposure were less risky, without sacrificing return.  Consider the summary statistics and net equity exposure (“net”) for long/short portfolios based on Momentum, Book-to-Price and Earnings Yield styles with holdings weighted equally during the periods of increasing volatility described above.

Earnings Yield

Note:  Volatility is the trailing 36-month volatility as of the month of peak market volatility during the period.  Return is the portfolio return from the start of the period to the peak month.

Book Value-to-Price

Note:  Volatility is the trailing 36-month volatility as of the month of peak market volatility during the period.  Return is the portfolio return from the start of the period to the peak month. 

Momentum

Note:  Volatility is the trailing 36-month volatility as of the month of peak market volatility during the period.  Return is the portfolio return from the start of the period to the peak month.

These results show clearly that a net exposure different from 0 would have produced a less risky portfolio assuming that the investor was not going to change its holdings.  It is happenstance that the net exposures are positive and approximately 45 – 55%.  More importantly, the takeaway is that the zero net exposure would not have consistently insured the investor from the increasing market volatility, and often would have increased volatility.