What is the VIX? With so many indicators out there, would this be one to add? Trading the stock market is fun to think about. When times are tough, knowing how to trade can be extremely beneficial. As a result, a lot of people jump on the trading bandwagon. Many think they’re gonna get rich quick. But it doesn’t turn out that way. So what if there’s a way to help measure volatility?
- The Volatility Index (VIX) measures differences between prices on future calls and puts. If call options are being purchased for dates several months in the future for wildly varying prices, the VIX should have a high number; typically in the twenties or thirties. If calls are trading at similar prices (not necessarily the same as the current stock prices), the VIX should have a low number; in the single digits or teens. Interestingly, the market seems to react to the same information the VIX reacts to. In that days when the market swings seem to correlate to days when the VIX is high.
This article attempts to answer three major questions about the VIX index.First, how reliable is the VIX? Meaning specifically, can an index that measures unpredictability, in fact, measure it reliably? Is there a difference between, say, predictably unpredictable, unpredictable, and very unpredictable?
Second, how useful is the VIX? Specifically, does a change in the VIX index provide enough information to take action in the market? Or are changes in the VIX priced in immediately?
What does reliability mean? Naively, if the VIX index is low, one would expect the volatility in the market several months in the future to be low.
If the index is higher, limited future volatility would be expected. And if the index is very high, extreme volatility would be expected.
Since it’s measuring future chaos, one would want it to be reliable. However, not overly reliable, since chaos is inherently unpredictable.
To test reliability, we look at three ranges of values for the VIX:1. 0 < VIX < 152. 15 <= VIX < 193. VIX >= 19
Our hypothesis is that average market volatility in the six months succeeding a VIX measurement will be very limited for case 1, limited in case 2, and unlimited in case 3.
The hypothesis, if not rejected, will supply specific volatility measurements we may be able to use in trading.
The Easiest Way to Measure Reliability
The easiest and most standard way to measure reliability is in measuring combined returns. We took the ending monthly values of the VIX from January 1990 through September 2019; since we’re looking at several months of future activity, daily swings in the VIX would not matter.
We then took the combined monthly returns of the S&P 500 for the same period, and created a data set comparing returns to volatility in six month periods, looking out six months from each month to select and normalize combined returns of the S&P to each VIX category, generating several thousand data points.
We hypothesized that the VIX category would not predict combined returns.
Regressing normalized combined returns against VIX categories revealed that the coefficients of the categories were significant with a confidence level of 95% or higher.
Thus, we were able to reject the hypothesis, and proceed under the assumption that VIX values above 19 correlate with different combined returns than VIX values between 15 and 19 (VIX values below 15 were assumed to have little to no market volatility and were ignored).
The hypothesis test has also answered the question of usefulness. While volatility predicted by the VIX is probably priced in immediately by the market, the fact that the hypothesis was rejected indicates that not all the volatility is priced in immediately.
Therefore, opportunities to profit exist for several months after a change in the VIX. You can also use something like the TTM Squeeze to make quick trading decisions.
Is the VIX a Leading Indicator?
- When it comes to indicators, we get leading or lagging. When an indicator is lagging, it reads past data. The VIX would be a leading indicator. Because it’s economic data that’s used to predict and forecast future events.
A pattern screener uses several parameters to decide if a pattern is present (or likely to be present) in stock prices. For example, an ascending or descending channel is defined by the distance between the trend lines, the number of breakouts, the entrance criteria to the pattern (when prices begin to move within the trend lines), and the exit criteria (when have prices diverged far enough from the trend lines to be called an exit).
If you were to look at several months of stock prices and apply a particular pattern screener, you would identify a number of instances of the pattern in the data.
If you take the very same data, and adjust the prices to reflect an increase in volatility while still maintaining the general trends of the data, the screener ought to find fewer instances of the pattern because:
- More breakouts will appear to be exits because breakouts will occur more often and move further above or below the trend lines, and fail to match the screener’s breakout parameters.
- The trend lines will be further apart, also failing to match the screener’s parameters.
- The entry points may not be identifiable to the screener since the trend lines are further apart.
- The exit points may be prematurely identified (really breakouts).
Thus, if the VIX were to increase today, the number of patterns identified by screeners would be reduced, and with the same number of traders trading the stocks, these patterns would be overtraded, and less profitable.
The obvious suggestion is to react to a change in the VIX by modifying your screener parameters to pick up the patterns that would not be found and trade those patterns.
For example, suppose the VIX goes up a given percent and you start a new screener, adjusting the parameters accordingly. You then run the old and new screener in parallel.
Any pattern found by the new screener and not the old screener is likely to be thinly traded and more profitable.
Given a particular change in the VIX, how would you adjust the parameters? Since pattern trading is stock specific, changes in the VIX will be manifested in individual stock prices differently. We recommend the following approach:
First, select stocks of interest to you. Then grab several months of price data and calculate the monthly (or daily, or weekly) standard deviation in excel. In a second column, get several months of VIX values (actual value or change in value) and compare the volatility of your stock to the VIX.
The end result you want to calculate is: for any given value or change of value in the VIX, a likely standard deviation in price changes in your stock. It doesn’t have to be scientific. A blue chip stock is going to have fewer price swings than a small cap stock and you merely want to understand how much your stock price moves as opposed to the overall market.
Then, select a time period in the past, look at the VIX, select a time period in the future, and decide what the likely volatility of your stock will be.
For example, suppose my stock symbol is ABC and its average daily standard deviation is 9%. But I’ve decided, based on looking at the recent values of the VIX, that the likely standard deviation in the near future may be closer to 18%.
This is the same as predicting that breakouts will be twice as big (18% divided by 9% = 2), trend lines will be further apart by 9% (18% – 9% = 9%). And exits will naturally need to be larger than breakouts.
If I take my standard screener for, say, an ascending or descending channel, create a new screener with the new parameters, run both in parallel, choosing only the patterns found by the new screener, it’s likely that:
a) such patterns will exist (you did the calculations, after all), andb) fewer traders will trade the new patterns,
Leaving you with less competition at those particular points in time for trading the stock.
The VIX is a great leading indicator for volatility with options. Options are a great way to grow a small account as well. When you learn how to the VIX with your trading, you’ll be that much better as a trader.