Thursday, April 13, 2017

Buy When There's Blood in the Streets? Market Timing with Volatility Triggers

An 18th century British nobleman, Baron Rothschild, was rumored to have made his fortune buying during the panic that followed the Battle of Waterloo against Napoleon. He is behind the often quoted saying "Buy when there's blood in the streets”, which he continued “even if the blood is your own." This post will share a framework that may identify regimes that benefit from buying when there's blood in the streets, as well as when the market is at the greatest risk of underperformance (potentially allowing investors to reduce risk before there is blood in the streets).


Market volatility has long been one of the most disruptive aspects to investing and investor behavior. Over the past few decades, U.S. investors have seen large portions of their equity wealth evaporate as market volatility spiked on multiple occasions.

January 31st, 1993* - December 31st, 2016

  • 4 Distinct S&P 500 drawdowns greater than 30%  
  • 15% realized monthly standard deviation of the S&P 500 (annualized) 
  • 5% to 29% range of one year rolling standard deviation of monthly S&P 500 return  
  • 20 average level of the VIX Index
  • 67 Month-ends where the VIX Index was > 30
  • 89.5 peak VIX value during 2008-2009 financial crisis
* The CBOE first started publishing the VIX Index in January 1993

Poor investor reaction to market volatility has contributed to the poor returns investors have captured. Analysis of Morningstar investor returns relative to the returns produced across US equity styles reveals the extent to which investors have undermined their own investment performance over time. Over the last decade the average investor realized returns that were as much as two percentage points lower each year than the relevant Morningstar category.

In the case of the Morningstar large-cap value category, investor behavior reduced compounded dollar returns by 40% over this ten year period relative to what the average fund within the category produced. If / when markets exhibit another period of heightened volatility – which at some point is bound to happen – and if investors continue to undermine their investment performance during these periods – which is likely to happen – a systematic approach that can reduce the impact of these swings in price and volatility may help investors stay on track.


We’ve all seen the caveat that “past performance is not indicative of future returns”. The same has been true regarding past levels of volatility (high or low) and future returns (i.e. the relationship between the VIX and forward returns is weak). On the other hand, past levels of market volatility has been correlated with future levels of market volatility (i.e. when volatility is high... it tends to stay high). Given returns have been similar irrespective of the VIX, while volatility has been lower when the VIX is low / higher when the VIX is high, risk-adjusted performance (return per unit of risk) has been higher when VIX (and volatility) has been low (the return numerator stays roughly the same, while the standard deviation denominator is higher in high volatility regimes).

To identify high / low volatility regimes, we can use the current level of the VIX (in the examples below, when the VIX is above or below the historical average of 20).
  • High Volatility Regime (VIX > 20): volatility is more likely to remain elevated
  • Low Volatility Regime (VIX < 20): volatility is more likely to remain low

But wait… there’s more.

The returns within high volatility regimes (those when the VIX ended the previous month above 20) can be broken down further, in this case split between periods when the VIX had declined or increased month-over-month. Since the VIX Index inception in 1993, during high volatility regimes when the VIX declined month-over-month (i.e. VIX was above 20, but the VIX was less than the previous month-end), returns have been materially higher than when the VIX was elevated and had increased, while the risk was also greatly reduced when the VIX had declined. In fact, the risk-adjusted returns when the VIX was elevated and declining closely match the high levels of those generated within low volatility regimes.


Given historical risk-adjusted returns have been much more favorable when the VIX ended the previous month below 20 or when the VIX was above 20 and declining, we can test the hypothetical performance of a model rebalancd monthly that has a risk-on allocation (stocks) when the VIX is low or declining and a risk-averse allocation (intermediate bonds) when the VIX is high and increasing.
  • Low volatility regime (VIX < 20 or declining): 100% S&P 500 Index
  • High volatility regime (VIX > 20 and increasing): 100% Bloomberg Barclays US Intermediate Treasury Index
The model’s results are promising. Not only have the returns been similar (in fact slightly higher) than a buy and hold allocation to the S&P 500, risk was greatly reduced resulting in materially higher risk-adjusted returns.


Although Baron Rothschild may have had the fortitude to buy when there was blood in the streets, the above framework reveals there may be other, potentially less stressful, ways to capture the opportunity. While this example is simplified – and of course hypothetical – a similar framework may help protect investors from undermining their own financial progress by reducing equity exposure before fear fully materializes and/or increasing equity exposure when fear is high and improving.

No comments:

Post a Comment