Analytical Toolbox: Volatility Indexes
March 19, 2009
The CBOE Volatility Index® (VIX) is probably the most watched measure of volatility in the US equities markets, followed by the CBOE Nasdaq-100 Volatility IndexSM (VXN). VIX measures the implied volatility levels for a select group of S&P 500® Index (SPX) options using a construction methodology that was revised in 2003. Historical data from CBOE provides VIX levels using this revised method on pre-2003 option data. The former VIX was calculated using the S&P 100® Index (OEX) which is now available as VXO. VXN measures implied volatilities for the NASDAQ-100® Index (NDX).
Increased volatility generally accompanies increased fear as markets decline. This results in a strong negative correlation between the SPX and the VIX with a 1% move down in SPX corresponding to a 0.74% move up in the VIX, using a little more than 10 years worth of data. Keep in mind, however, that the relationship describes the current day's movement. Is there anything a volatility index can tell us about the future for the underlying index?
In his revised July 2003 paper, "On the relationships between implied volatility and indices and stock index returns", Pierre Giot of the University of Namur [Belgium] uses two index-volatility pairs to assess the same day relationship between the pairs, as well as lagged results. That is, the performance of the underlying index a certain number of days after the volatility index is measured.
Giot used OEX-VXO and VXN-NDX pairs to confirm the negative correlation between each volatility index and the underlying stock index, noting the relationship was statistically significant. In addition, moves down in the stock market translated to bigger moves up in the volatility index, particularly during low volatility periods. This characteristic was stronger in the OEX-VXO pair.
In terms of lagged results, Giot noted that, "There is some evidence that positive (negative) forward looking returns are to be expected for long positions triggered by extremely high (low) levels of the implied volatility indices. However one must wait for extremely high levels of implied volatility to get attractive positive forward looking returns."1
While the study was completed prior to the new methodology, the CBOE reports in its product description that the revised VIX behaves similarly to VXO (old VIX) and provides annual high and low data for comparison purposes. When viewing a scatter plot of the two relative to daily SPX changes, the regression line for the new VIX is slightly steeper than VXO, displaying an increase in the correlation to SPX.
A view of the current weekly VIX-SPY relationship of returns can be seen in the scatter plot provided in Figure 1. VIX data is on the x-axis and is more dispersed than the SPY data.
Figure 1: VIX-SPY Weekly Returns
VIX Charts
When analyzing the VIX to assess the health of the underlying index, or in this case its proxy SPY, it does appear that large spikes upward in the VIX does correspond to positive future returns in SPY, but those returns are not necessarily sustainable on a long-term basis. At the same time, it doesn't appear that significant market reversals occur without a series of VIX spikes which includes a last blow-off and a subsequent successful test of previous levels. Unfortunately, this is based on a relatively short history for the pairing (16 years on the charts).
Since the VIX is a measure that is not range-bound, there isn't a specific level that people can point to as oversold. This was seen in the Fall of 2008 when the index was reaching historic highs and continued to move upward. A view of the monthly VIX chart with its 12-month exponential moving average [EMA] and an SPY overlay appears in Figure 2.
Figure 2: Monthly VIX with 12-Month EMA and SPY Overlay
It appears a more range-bound Relative Strength ratio of the VIX/SPY corresponds to sustained upward moves in the underlying index. This can include sideways trending and moderate declines.
Shortening the interval to a weekly basis for the VIX and removing some of the spike noise by removing the index data and displaying the 10-week EMA for the index with the SPY overlay provides a bit of a clearer view of the volatility trends. Figure 3 displays this chart on a logarithmic scale.
Figure 3: Weekly 10-Week EMA of VIX with SPY Overlay on Log-Scale
Here you can see accelerating positive returns in the 1990s with fewer volatility spikes, along with successive volatility peaks in declining markets.
To Monitor in the Weeks Ahead
It seems hard to imagine that Wednesday's post-Fed move after a week-long rally will completely remove fear from the market. With that in mind, look for smaller move upward in the VIX before any significant uptrend develops (intermediate-term as viewed on weekly charts.) A strong move down in the VIX may suggest more weakness to come, while a more moderate decline may also provide bullish fuel in the intermediate term.
Since I tend to stay focused on the current trend in place, rather than potential counter-trend moves, remember there are no guarantees in the markets. No one knows what next month, next week or the end of the week will bring.
1Giot, Pierre, 2003. "On the relationships between implied volatility and indices and stock index returns," Center for Operations Research and Econometrics [CORE], at Universit´e catholique de Louvain, Belgium. [pp 12-13].
Current Volatility and Index Correlations
To access other articles written by Clare White, please click here.
Clare White
Contributing Writer and Options Strategist
Optionetics.com ~ Your Options Education Site
Questions for Clare? Visit the Optionetics.com Discussion Board
Email This Article
View Full HTML
Digg This
Delicious
Subscribe© Copyright 1995-2010 Optionetics. All rights reserved. This material is for personal use only. Republication and re-dissemination, including posting to newsgroups, is expressly prohibited without the prior written consent of Optionetics. Optionetics is a registered trademark of Optionetics, Inc.