VIX vs. the IV Index - iVolatility Builds a Better Mousetrap for Capturing Market Volatility Measurements

March 1, 2001

For the past year, NASDAQ stocks have been hit by wave after wave of bad news. Maybe this sell-off was triggered by the curtailment of margin trading, or was it when the price of a barrel of oil jumped over $30? Insider profit-taking after their lockups expired may also be blamed. Some people say it is valuation corrections exacerbated by earnings revisions and warnings, and finally, bankruptcies and restructurings. Whatever the reason(s) the damage is undeniable-the NASDAQ Composite Index (COMPQ) has fallen more than 60% from its all-time intra-day high of 5132, set last March. Although much of the buzz is gone from the dot com go-go days, there is still a large degree of uncertainty in markets today. Just how low can these markets fall? How much pain can investors endure? How high can volatility levels rise to? When has the market moved to an extreme?

One indicator of market sentiment and extreme moves that many investors use is the CBOE's S&P 100 Implied Volatility Index (VIX). Launched in 1993, the VIX has now become a fixture on many a trader's monitor and the home page to many financial web-sites. Calculated by using implied volatilities of options on the S&P100 Index, the VIX is often looked at as an indicator of overall market sentiment. Extreme values in either direction signal a reversal in the current trend. tracks implied volatilities on every stock trading listed options in the US, and also covers many indices as well. Like the VIX, our 30-day IV Index looks at actual option premiums quoted in the market. We have taken great care to perfect our analysis of this data, resulting in better quality signals. Any given expiration month can have over 50 different strikes, with each strike exhibiting differences in implied volatilities. Add to that, differences in implied volatilities between puts and calls, and you have a very complex problem in trying to estimate the market's expectation of future volatility.

We believe that our approach to calculating implied volatilities produces a more accurate estimate of the market's expectations, is more valid in making comparisons from one day to the next, and provides a better indication of market sentiment based on option premiums. A look a recent market conditions and a little back testing proves our point.


On February 23, the S&P 100 Implied Volatility Index (VIX) had become oversold and thus presented a positive indication for the market following a trend change. The index then turned back down and declined to a low of 26.25% on March 8, not even reaching the overbought range before the market turned negative once again, as illustrated on the chart below.

S&P 100 Implied Volatility Index

By contrast, the IV Index for the OEX filters out more of the noise than VIX. For example, while the VIX got whipsawed in February giving several oversold signals, our IV Index never once gave a false signal to get into the market. Not only did it remain well below oversold levels for the entire month, it actually gave a clear warning that there could be more carnage directly ahead as it signaled, within one day, the high close for the month. On February 14th the IV Index dipped below 20% and thereby signaled that the market had reached overbought conditions. The very next day, the OEX registered its high closing price for the month at 688.50.

IV Mean for OEX vs. 30 day Historical Volatility for S&P100


The last quarter of 2000 was rocky for many investors, but if you were looking to the VIX for guidance you would have received many false cues. On the other hand, the IV Index signals were stellar. While the VIX gave oversold conditions throughout October, November and December, the IV Index had only three. On October 12th, the IV peaked just above the 30% level (30.51) - the very day of the low close for OEX that month (703.68). In November, the Index peaked just above the 30% oversold level on the 29th, this time it was one day ahead of the underlying index which bottomed at 697.28 on November 30th. December registered 2 oversold days; the first and highest reading was on the 19th when it spike to 32.84%. The next day, when the S&P100 collapsed to 661.50, but the IV index improved to 30.53% and the S&P100 recovered 25 points during the waning days of the year. Not a bad signal to end out a year that was hard on many investors.


On January 22, 2001, the CBOE launched the VXN, a sister index to VIX. The VXN tracks volatility and option premiums on the NDX - which at the time of this writing has just reached levels last seen in March 1998. The VXN, like its underlying instrument is a far more volatile instrument than its sibling, the S&P100 Index. With readings typically in the 60% to 70% range, VXN is currently over 75% and clearly needs to be scaled to a different range. The problem is limited historical data on the VXN, which begins on January 2, 2001 (and coincidently its highest level to date, 84.95%), certainly not enough time to justify any significant back testing.

The CBOE Anounces VXN

IV Mean for the NDX vs. 30 Day Historical Volatility for the NASDAQ100

On the other hand, IVolatility's database of implied volatilities goes back to May 1999, but even with the aid of a much longer time series, it is difficult to draw any statistically valid conclusions about what constitutes oversold or overbought readings. Statistically speaking, the IV Index average over this period was about 46% with a standard deviation of roughly 12%, that would put any readings over 70% as an extreme, but it turns out that 75% worked better. In the year 2000, there were three days in which the IV Index on the NDX peaked above the 75 mark: April 14th, , May10th, and then on December 20th. The price action around these dates suggests that, at best, these readings were good for signals of corrections lasting 3 to 5 days. Perhaps using Bollinger Bands will prove more valuable in capturing extreme moves in the NDX than a static oversold/overbought indicator. We will take a look at applying this study to our data and report back to our registered readers in a subsequent issue.


At the time of this writing, the VIX is back above 32%, another oversold reading, but the IV Index for the OEX is only 24.56%, portending that the decline in prices is hast not yet reached an extreme. The IV Index for the NDX is at 57.40%, pretty close to its average reading, which for so far this year, has been 58.41%. The VXN has already recorded 12 days this year with readings over 75%. By comparison, the IV Index has yet to register a single 75%+ day this year, once again proving that our approach to measuring volatility provides better quality readings than the VXN and VIX. Just one reason more reason you should be "examining your options" with iVolatility on a daily basis.


We are adding more to our web-site and are pleased to provide market commentary by Scott 'the Strategist' Fullman. Please visit our new section of the web-site, The Strategist, on our navigation toolbar for a free 2 week trial.

Scott has over 20 years of option market experience and is a registered option principal. He shares his current market view with us here, "We continue to suggest the use of the covered combination strategy as an attractive alternative to purchasing the underlying shares. While the negative trend of the NASDAQ has dominated financial headlines, there is an opportunity for taking positions in depressed shares for the long-term horizon. Short-term traders should use close stops to limit losses as quick trend reversals are possible. Holders of short-put and long-call positions should adjust expiring positions early and not wait for expiration pressures, which should become evident later in the week. Caution is still urged for all positions!"