By Lawrence G. McMillan
This article was originally published in The Option Strategist Newsletter Volume 13, No. 19 on October 15, 2004.
Despite a modest, recent rise in $VIX, the CBOEs Volatility Index remains very subdued – as it has since March of 2003, and especially for most of this year. There are some general relationships between the broad market and $VIX, and there is a good deal of price history to justify those relationships. However, there have been recent articles published in several forums that suggest many traders seem to think it will be different this time – that $VIX isn’t predicting the same sorts of things that have happened in the past. In this article, we’ll explore those suppositions and try to outline some things to look for – from both $VIX and from the broad stock market.
$VIX As A Buy Signal
As a means of background – although it isn’t applicable to today’s market – let’s establish that $VIX has a good track record as a buy signal. When a market is collapsing, and if it is accompanied by a swiftly rising $VIX (as it usually is), then a buy signal sets up. That buy signal arrives when $VIX peaks. There are numerous instances of such activity and – while some traders often try to anticipate the peak and get buried as selling continues to drive the market to its nadir – a conservative approach of waiting until it is clear that $VIX has peaked is a tried and true bullish strategy. Not only has this approach worked at major bottoms, such as October 1998, September 2001, July and October 2002, March 2003, and so forth, but it has also worked at more minor bottoms such as the ones we’ve had this year (see the $VIX chart on page 11).
Thus, we can clearly say that a peak in $VIX is a buy signal. Almost all observers of $VIX agree with this fact, although I should point out that at the 2002 and 2003 bottoms, most of them got it wrong because they bought way too early, when they thought $VIX was “too high,” rather than simply waiting for it to peak – thus letting the marketplace show the buy signal.
Actually, I believe it was the excellent timing of the $VIX buy signals in 2001 and 2002 that enticed a lot of would-be volatility analysts to begin watching $VIX.
What Does A Low $VIX Mean?
Once most of them understood that a high $VIX leads to a buy signal, they jumped to the conclusion that a low $VIX leads to a sell signal. Not only is that the wrong conclusion, but most of these newbie volatility analysts didn’t have a clue as to when $VIX was “too low.”
Thus, as $VIX declined steadily all during the 2003 market rally, they kept issuing sell signals because of it. Now that $VIX has hit rock bottom (or has it?), and the market hasn’t collapsed, they are writing articles that say a low $VIX clearly isn’t a sell signal and so doesn’t tell us much of anything. They are apparently going to retreat to the sidelines until some day $VIX is high again and they can rush in to buy (but a cynic might say that they will probably buy too soon).
One of the interesting things about contrary indicators is that if everyone begins to believe in them and thus embraces them, they are no longer contrary, for the interpreters have thus become the majority – and will thus be wrong, by contrarian logic. This is a little bit of what is going on with $VIX analysis.
We have written many articles about $VIX in the past, and I won’t go into great detail on all of the points that have been made, but here are a few of the important ones that are necessary for an understanding and interpretation of $VIX.
A) $VIX is not merely a contrary indicator; it is a measure of implied volatility. And implied volatility is tied to historical volatility to a certain extent, so trying to interpret $VIX as a contrary indicator at all times is not a correct thing to do.
B) When $VIX is applicable as a contrary indicator, it measures public sentiment about the upcoming volatility of the market – not necessarily the market’s direction. To understand this, one must delve into what the public is doing. When $VIX is skyrocketing while the market is falling, what is really happening is that traders are panicking to buy puts, and will pay most any price. So put premiums rise (and, through arbitrage, call premiums do, too) and $VIX rises. This action can be interpreted in a contrarian fashion: when the last panicky trader has paid top dollar for the last put, the market stops falling and rallies. Thus, the buy signal from a peak in $VIX.
But, when $VIX is low, as it is now, it implies something completely different from a contrarian’s viewpoint. What causes $VIX to be low? Option sellers are dominant, option buyers are timid, and option premiums decline (and so does $VIX). What do the majority of people expect the market to do at these times? Nothing. They expect the market to remain dull and thus they feel comfortable selling (naked) options. When the majority expects nothing to happen, you can rest assured that something will happen.
Usually that something is a large market explosion – typically in the direction of the major trend of the market. We have written many articles and shown many charts in the past to demonstrate this. Thus, a low $VIX is not a sell signal, but rather a harbinger of higher volatility in the marketplace – and that volatility can manifest itself as a rally or a decline. Classic examples were the 9- month rally in the first half of 1995 after $VIX traded down to 10, or the market crashes in 1987, 1998 and 2001 after very low (relative) $VIX readings preceded them. On a more routine basis, in the 1994-2000 time frame, a low $VIX reading generally preceded a rise in the stock market, but since 2000, the opposite has been true as low $VIX readings have often preceded market declines.
C) a low in $VIX can’t be determined except in hindsight – where $VIX has either risen sharply afterwards, or has not made a new low for 30 days.
Most of the columnists whose work I am quibbling with have been too quick to call a “low” in $VIX. We have had only two occasions when $VIX did not make a new low for 30 days this year – and both led to market declines. So, right now, we’re either waiting for $VIX to rise above 16 (a rise of 3 points from a low is usually the signal that $VIX has bottomed) or for it to not make a new low for 30 days (another sign of a $VIX bottom) in order to declare that the expected market explosion is about to occur.
Most of the columnists whose work I am quibbling with have been too quick to call a “low” in $VIX. We have had only two occasions when $VIX did not make a new low for 30 days this year – and both led to market declines. So, right now, we’re either waiting for $VIX to rise above 16 (a rise of 3 points from a low is usually the signal that $VIX has bottomed) or for it to not make a new low for 30 days (another sign of a $VIX bottom) in order to declare that the expected market explosion is about to occur.
This article was originally published in The Option Strategist Newsletter Volume 13, No. 19 on October 15, 2004.
© 2023 The Option Strategist | McMillan Analysis Corporation