The VIX and the financial crisis
By David BukeyJust days before Christmas, many investors likely breathed a sigh of relief as the CBOE Volatility Index (VIX), a measure of S&P 500 option implied volatility, dropped below 20 for the first time since the financial crisis struck in September 2008.
The VIX is a widely watched stock-market barometer that tends to rise as investors anticipate more volatile markets and drop as they expect them to settle down. In periods of uncertainty, traders bid up options prices as they scramble to protect stock portfolios by purchasing put options. In rising markets, however, traders often don’t buy puts because they believe a significant market decline is unlikely.
Because the VIX moves inversely to stocks, traders typically assume high VIX values correspond with market bottoms and low VIX values appear near market tops. That conclusion isn’t necessarily wrong — the VIX and SPX often mirror each other.
But timing is everything. The VIX quadrupled in value in September and October 2008, peaking at an all-time high of 89.53 on Oct. 24, 2008 as investors panicked. Meanwhile, the S&P 500 lost one-third of its value during this period. History shows buying stocks after such extreme conditions has tended to pay off; in this instance, SPX rallied 18 percent within seven days. But the bounce was short-lived — the upswing topped on Nov. 4 — and the stock market dropped another 33 percent over the next four months. Like most “sentiment” measures, the VIX has rarely identified exact tops or bottoms, and has tended to hit extremes before the market reverses direction.
A previous Market Pulse study examined the relationship between the volatility index and S&P 500 Depositary Receipts (SPY) and found the ETF rebounded roughly 1.2 percent immediately following “unusual spikes” in the VIX from 1993 to 2005 (“Tracking VIX swings,” Active Trader, January 2006). But SPY’s behavior over the past couple of years appears to have diverged from historical norms.
Identifying markets on edgeThe earlier Pulse article measured VIX changes over various look-back periods (i.e., 60 and 120 days) and ranked those moves against previous moves of the same length and direction. For example, each 60-day move was compared to the past 120 same-length moves, and each 120-day move was compared to the past 240 same-length moves.
Each move was then ranked on a scale of zero to 100, representing its size as a percentage of prior moves. A 100-percent move means it is larger than all past moves, and a zero-percent move means it is smaller than all previous moves. The analysis focused only on moves with a percentile rank of 95 or higher. In addition, each VIX move had to exceed its most recent moves. For example, a 60-day VIX gain with a percentile rank of at least 95 also had to exceed the highest VIX high of the past 60 days, and so on.
Figure 1 shows a daily chart of the VIX (top) and SPY (bottom) from August to December 2008. Again, as the VIX climbed to new highs, the market slid to new lows. Anyone who bought stocks in response to the 60- and 120-day VIX highs in September and early October 2008 would have been decimated. Buying stocks after volatility spikes might have worked in calmer markets, but not during the greatest financial panic in 80 years. (However, if you look closely at Figure 1, you will notice SPY managed to climb briefly the day after the 60- and 120-day VIX highs. This anecdotal evidence is no reason to jump into the market as it sinks, but if extreme highs in the VIX prompt you to sell stocks, you might want to reconsider and wait for a brief rebound first.)
For the complete article, see the March 2010 issue of Active Trader magazine. Click here to subscribe.
|
|
