The VIX - VXO a.k.a. old VIX

::::This document is an archive, although still very present in the search engines. ::::: The arguments developped here have been valid for a long time. ::::: Current success of exchanges traded funds has change the overall picture. ::::: Hence the need for a new VIX based on the SP500 options prices on Septembre 22, 2003. ::::: This will be addressed in a new paper in the near future.  ::::: In the meantime, please open this link to the CBOE VIX minisite.:::::::

Archived from Oct 20, 2000

Introduced by the CBOE in 1993, this index is calculated by taking a weighted average of the implied volatilities of eight OEX calls and puts. (OEX is the index of the Standard & Poors 100).
The chosen options have an average time to maturity of 30 days.
Consequently, the VIX is intended to indicate the implied volatility of 30-day index options.
What the CBOE itself explains: Volatility Indices
It is used by options traders as a gauge of investors fear: high numbers mean excessive bearishness, hence a bullish indicator.

Some authors prefer the inverted VIX, easyer to apprehend: a high value of a VIX corresponds to a bottom in the market. It is easier to apprehend an indicator if the highs corresponds to highs of the market. Many charting programms allow you this.

According to the most current interpretation, usual levels are:

Needless to say that I am not satisfied with this interpretation. Most analysts still are referring to a piece by Bernie Schaeffer of Oct 1997.

Quotes and charts
a chart of the VIX can be found at: StockCharts
Streaming chart delayed from Quote.com , type VIX.X as the symbol.

Other sources:
OEX Trader has a proprietary plot on the VIX

Historical values: (this kind of table leads the commentators hasty conclusions.)
Check this reminder.
 

Consecutive bottom signals? What happens between bottoms? IMHO, the table shows the limitations of the simplistic approach.
Checking the CBOE VIX statistics (although the VIX is recent, 1993, CBOE has back-checked it's values till 1993 it show an average value for the VIX of 20.37. This number should be considered as normal.
Low readings of the VIX implies low volatility in the market, low volatility implies small changes in the prices, also range trading.
In an upward range, low volatility should then mean that we stay in the range: upward. Same for a downward range.
If volatility was to change drastically, this would mean larger changes in prices,( up and down price changes, also a break of the trading range): break, as we all know, can be upwards or downwards.
The traditional interpretation is, IMHO, that the peaks are the representation of panic sell-offs, as the throughs are the representation of lower prices bottoming. Chart reading in a small window (3 to 6 months) during 1998 and 1999 would have lead to that interpretation.
As the VIX has been orderly printing lower prices in a long period of time (from April till now), the interpretation I come with, is that the market is driven by an orderly bullish order flow (not seen in months or years). When the VIX will show increased values (still only my interpretation) this could be the sign of the re-entry of speculative or momentum players, consequently, time to get out.
At what value? I don't know: from March 1991 through Dec 1996, VIX very rarely printed 20 or more. Bearish sign, keep out of the market? VIX below 20= you must sell ?
From Mar 91 to Dec 96, the SP500 climbed from 340ish to 620ish.
Better:
Early Dec 94, SP500= 443, VIX=14.91
Mid Feb 96, SP500=664, VIX=17.06

My current reading:
I analyze the VIX with the indicators available on StockCharts:
with these settings:

Interpretation:

     

• when CCI is above 100 and StochRSI is above .80, go short
• when CCI is below -100 and StochRSI is below .20 go long
• when other values, abstain

No matter which is the quote of the VIX. See the table at the bottom of this page.

These settings have caught most of the trends.
But read Bernie Schaeffer's first article cited here below: VIX could be obsoleted as an indicator.

Recent articles about the VIX

• Bernie Schaeffer has noted the decline in daily volume of OEX options. To the contrary, options on the Nasdaq 100 Trust equity (QQQ) has attracted a lot of traders: time to check QIX (QQQ's VIX) ? Jul 07, 2000
   
• Rex Takasugi piece of Apr 07, 00 on Stockcharts.com I am of the opinion that a significant low has been put in by the NASDAQ and technology stocks in general. It would be surprising to me to see this spike low broken over the near term.
   
• Rex Takasugi piece of Apr 21, 00 on Stockcharts.com
  VIX Trading Model Update: The sharp increase in volatility out of normal parameters has caused my VIX Trading Model to be thrown off, as the latest buy signal appears to be either wrong or early…
  During the autumn of 1998, a similar situation also occurred. When this happens I step up to the next time frame, which would be a weekly chart…
  The weekly VIX is getting as close to the Buy Line as we've seen since that 1998 low. The VIX lows of 1999 have been exceeded, which would suggest that a major low is imminent, although we're not quite there yet. However, it is possible that we could see a reversal without reaching the Buy Line, as in 1999. That is my current expectation.

Most analysts still are referring to a piece by Bernie Schaeffer of Oct 1997.

• Consensus has a document dated Oct 16, 1997 interesting as the VIX still was a relatively "new indicator" then.
  SCHAEFFER'S RESEARCH REVIEW
  Prepared by
  Investment Research Institute, Inc.
  This week marks the 10-year anniversary of the 1987 stock market crash.
  No doubt you are well aware of this fact in light of the extensive media attention given to remembering this event.
  Many analysts have been offering their opinions on whether a similar crash could decimate market prices again some time in the future.
  Most of the analysts have referenced fundamental or technical research to make their point (on both sides of the issue), but in our first segment, we are going to discuss the relevant sentiment factors of then versus now.
  The next topic in this issue of Schaeffer's Research Review explores a very effective sentiment analysis of the Dow Jones Industrial Average (DJIA).
  As you may be aware, options on the DJX (an index 1/100 the price of the DJIA) have recently begun trading on the Chicago Board Options Exchange.
  The sentiment analysis we present here will likely provide a very effective addition to a DJX option trading strategy.
  Finally, we move to a new analysis of one of our favorite sentiment indicators–the CBOE Market Volatility Index (VIX). Long-time readers have become familiar with our analysis of the VIX, but the study we discuss in this issue is slightly different from any mentioned previously"
  Jerry Wang
  October 16, 1997 Investment Research Institute, Inc.
   
• The PitBull Investor referred (in 1999) to the Bollinger Band analysis of the VIX described in the above article.
  **VOLATILITY INDEX**One of the indicators I have talked about is the VIX. This is another indicator from CBOE whose full name is Market Volatility Index. This index, while it has no nice fixed number, (like <0 = oversold >0 =overbought), can be best evaluated with the use of simple 20 period Bollinger bands. When the VIX rises to meet or exceed the 20 day upper band, you can generally count on a rally within the next two days....conversely, when the VIX drops near or through the lower band, you can expect the market to reverse and fail within the next two days.... It doesn't ALWAYS work that way, but when used in conjunction with some other simple indicators can give you a pretty good idea of market direction near term."
   
• Teresa Lo on a piece of (dead link www.trading-ideas.com/guest/speculator/speculator-snapshot070299.html) July 02, 1999 discussed the Bollinger Band approach and notes the limitations of that index.

Return table of a VIX based system

It can be noted that the above table is not based on sufficient data to be statistically acceptable.
But as note above, VIX should better be replaced, as indicator, by QQQ based, or QIX based data.
For my own use, 2 bad signals out of ten (one 2 day's long, one 3 day's long) is good enough when this indicator is to be used with other indicators as the put/call ratios.