As the stock market spiking to new heights, the volatility index, being inversely correlated with stock indices, has been descending to new lows. Today, it took an 8% tumble and fell to 11.56. It is the lowest close since February 26, 2007. As luck would have it, on the very next day, February 27, 2007, S&P 500 fell 3.5% on concern over slowing growth in China, which was then the growth engine of the world economy. Those who would like a refresher on that fateful day six years ago, can read this article from CNNMoney.
The fact that the VIX is at a six year low worries many market participants. We all know that volatility index is inversely correlated with stock indices. So as low volatility inevitably revert to the mean, there must be a subsequent drop in stock price. To assuage my own uneasiness, I downloaded the historical VIX data from Yahoo Finance and decided to find out what actually happened when VIX had fallen to such low levels.
There had been 5844 trading days since Jan 2, 1990. There were 372 days on which VIX closed between 11 and 12. VIX closed between 10 and 11 on 106 days and below 10 on only 9 occasions.
During those 22 plus years, in any given 90 days, there is a 33% chance that S&P 500 would record a negative price return. For 45 day periods, the odds of negative return for S&P 500 increased slightly to 36%. On the day that VIX closed between 11 and 12, there were a 31% chance of S&P 500 incurring a loss during the next 90 days and 28% chance of such events for the next 45 days. For VIX closing between 10 and 11, the odds of loss over the next 90 days was 17% and over the next 45 days was 31%. So the empirical results did indicate that on days VIX closed at a level between 10 and 12, the subsequent returns were in fact not worse off compared to any random days.
It was indeed very rare for VIX to close below 10. On the 9 occasions that it did over the past 22 years, the subsequent 90 day return were negative on 5 of them and positive on only 4, which meant the odds of adverse development is over 50%. For 45 day returns, there were 6 negative occurrences and only 3 positive ones. However, for the 90 day periods, the average of the negative returns was about 4% and non of the returns was worse than 5%. These were hardly calamitous. Furthermore, of the 9 days VIX closed below 10, there were really only two clusters. All the single digit VIX numbers were produced either during the end of 2006 and beginning of 2007 and the end of 1993 and the beginning of 1994. So we really only have two data points from which no statistically meaning conclusion can really be drawn.
When market is doing well and investors are calm, VIX is low; when market is in turmoil and investors are nervous, VIX is high. These are simply two different aspects of the market and even though VIX is inversely correlated with stock indices, history shows that one can not use VIX to predict market return. Just as rainy days and sunny days are two different types of weather conditions, the fact that sunny days inevitably follow rainy days, does not make rain a predictor of Sun since rain itself can not forecast the during of rain thus the subsequent ascent of the Sun. In the market, turmoils do always follow calm, but we can not use calm to predict turmoil as calm can last for a long time. So VIX as a market timing tool is simply not all that useful.
No comments:
Post a Comment