Note: I had published this material in August of 2009 on my former blog. I "discovered" these indicators on my own, but after the fact I did find that the VIX:VXV ratio has been conceived by others in the past. The charts are out of date, but they still illustrate the major points.

The following is a research report on two unique technical indicators for timing swing highs and lows in the S&P 500. The report analyzes the constitution of volatility indexes, and then uses the insights drawn from that analysis to develop market timing indicators.

INTRODUCTION 

The mechanics behind the VIX are taken for granted. The VIX, or Volatility Index, is constructed by the CBOE to index the implied 30-day volatility of the S&P 500. This means that if the VIX is at x, certain arithmetic yields that the market is supposed to move +/- y% in the next 30 days. That is all that needs to be said of this indexes’ intended purpose.

The VIX is known as a fear gauge because it spikes during panics, which coincide with market bottoms. Analyzing the reason behind this fact and then discovering and exploiting its predictive value is the purpose of this research report.  Surprisingly, the answer to why the VIX is a fear gauge is: a bi-product of simple number properties.

THE EQUATION

Above is the equation for how the VIX is calculated. Don't be intimidated. Don't be distracted. You don't even need to know what all the variables stand for to get the key takeaway. All you need to know to understand the volatility equation is in the green circle and red circle that are drawn.

The green circle shows you that the VIX is based on the weighted sum of many values. Those values are based on S&P 500 option prices and strikes. Essentially, each S&P option with a bid greater than zero contributes its weight to the VIX. In the red circle, K is the variable for the strike price. A mathematical property of 1/K^2 is that the greater the value of K, the smaller the weight, and vice-versa.Think of deep out of the money put strikes versus deep out of the money call strikes.

The former will have a much greater weight; e.g. 1/600^2 >> 1/1500^2. Therefore, put options must have a greater weight that call options in the calculation of the VIX. So, when enough investors are buying puts that expire within 30-days with incredibly low strikes, you can expect the VIX to soar to heights as lofty as the horror is profound. That behavior is important because it signals profitable buy points– that is the heart of the matter. Most other volatility indexes, such as those on other equity indexes or commodities, are calculated with the same formula.

One last thing I would like to add to further illustrate the point is a simple schematic I have made. It gives a visual of the result of the mathematical property described above.
The Y-axis is for the weight of an option in the VIX calculation. The X-axis represents the strike prices, ranging from out of the money put options, to at the money options, to out of the money call options. Notice how the weighing mechanism is not symmetrical.

When a mad scientist creates a beast with an intended purpose, that beast often turns and
surprises its maker with displays of its own unique will.

CONCLUSIVE EVIDENCE FOR BEHAVIORAL INTERPRETATION OF THE VIX

In a university lecture on technical analysis, Robert Prechter said, "The stock market is efficient at expressing the herd mentality."

For those of unfamiliar with Google Trends, it tracks the volume of Google keyword searches. I am not the first to write about it as being a useful sentiment indicator, nor the first to present the chart below. That chart is a plot of the weekly VIX against the search volume for the term "stock market". It dates from early 2005 to the end of August 2009 (Google Trends publishes delayed data.) Most analogous terms, like "Dow Jones," "stocks," or "recession" produce similar results.

Note the incredible, high correlation between Google keyword search volume and a financial index with a complex equation and dozens of inputs. I have yet to encounter better evidence that the market is an expression of behavior, as opposed to the continual pricing of a stream of material information. With the VIX, the behavioral element behind levels is very pure.

THE HEART OF THE MATTER

The prior sections on volatility indexes have served as an introduction. The insights I have drawn have led me to ask: since the VIX is a very pure expression of behavior, and behavior can aid in the prediction of short term (ST) and intermediate term (IT) market movements, how can I best isolate the predictive value of volatility indexes, just as VIX at 30 was a reliable buy signal in the past?

S&P 100 VS S&P 500 OPTIONS PARTICIPANTS, PART I

The basic sentiment interpretation of volatility indexes' peaks and troughs is that at peaks, fear is high and one should buy, and at troughs, complacency forebodes some sort of decline in the prices of the stock market. This strategy works because markets tend to turn directions at the extremes of both emotional states.

The original volatility index, VXO, was based on the S&P 100 (OEX). Later, the VIX, based on the S&P 500 (SPX) was created and gained prominence.  Comparing these two different groups of options participants revealed that consistent behavioral qualities can be identified.

In the chart below, I took the ratio of VXO to VIX, and plotted it against SPX, (to be exact, the 9-day SMA of the ratio, for clarity's sake.) You can see that the ratio and SPX are nearly mirror images. During market troughs, this ratio is higher, indicating that VXO is out-performing the VIX. In other words, at these instances, OEX option traders are more fearful than SPX option traders, as they purchase more or deeper out-of-the-money puts.

The same is true for relative highs in the market; OEX traders are more complacent than SPX traders. This behavior is consistent throughout the history of these volatility indexes. It is not random. One cannot attribute this difference to a higher volatility in OEX; if this were the case, there is no reason why this ratio should be lower at market highs. We can therefore induce a specific quality of OEX vs. SPX traders, as well as another indication that these indexes very accurately quantify the emotional states of the collective whole.

The next section demonstrates how this ratio can give buy and sell signals.

S&P 100 VS S&P 500 OPTIONS PARTICIPANTS, PART II:
THE VXO TO VIX INDICATOR

Just as price-based horizontal resistance and support lines as well as parallel channels provide accurate signals in financial indexes and commodities, so do they with sentiment. Emotions can be relative; humans build tolerances and are adaptive. During a sustained run up you will see higher highs in optimism; in sustained declines, higher levels of pessimism. There are also certain absolute levels of fear, such as the former VIX 30 level. Channels capture relative levels whereas horizontal lines capture absolute levels.

VXO:VIX best provides signals with channels. For this indicator, I like using the 9-day EMA as opposed to the actual ratio. I drew some circles to make it simpler to detect some the signals.

THE VIX TO VXV INDICATOR

The VXV is the 90-day volatility index of the S&P 500. The ratio of VIX:VXV (below) is the expression of sentiment about the near future relative to sentiment of a more distant future. Since VIX and VXV are both based on SPX options, there is surely an apples to apples comparison. In times of panic, people become a lot more focused on the immediate future, therefore, the shorter term index, the VIX, should out-perform VXV, raising the value of the ratio to indicate panic.

This indicator provides good buy and sell signals on horizontal levels and on channels. (However, it gave a false sell signal in July.) I believe that when it reaches that level again, it will be a potent sell signal. This chart plots the actual ratio, not a moving average. Notice the sell signals near the bottom of the ratio’s range, and the buy signals along the descending green trend line.

COMPLACENCY AND PANIC SPIKES

Throughout my blog you will find various examples of using bars that are buy and sell signals. There occur infrequently, which improves their accuracy. With the creation of a liquid ETF to trade the VIX, ticker VXX, I expect these spikes can be taken advantage of for a trade to a greater degree, as they indicate absolute and relative overbought and oversold levels for volatility.

That concludes this study. My goals have been to 1) provide insight into the workings of volatility indexes 2) use these insights to argue that they are a very pure expressions of behavior based on the simple emotions of fear during stock market drops and elation/complacency during rises and to 3) introduce valuable timing indicators.