In the "Second Key to the VIX", instead of using RVX/VIX exhaustion, I used the ((average of VXN and RVX)/VIX)'s percent change from a year ago (exhaustion) and compared it to the VIX and S&P 500 on a daily, weekly and bi-weekly basis. I like how it's not noisy on the daily chart (chart 1). After the daily indicator spiked through 30% in the second half of 2006, the VIX bottomed out in early 2007 and began its up-cycle. The weekly and bi-weekly charts are better at predicting corrections in the VIX and stock market. VXN is the CBOE Nasdaq 100 Volatility Index, RVX is the CBOE Russell 2000 Volatility Index, and VIX is the CBOE Volatility Index (for the S&P 500).
On the weekly chart, every time the indicator spiked or pierced through 20%, it accurately predicted when the VIX would prepare for a major market correction (see the beginning of 2011). However, it looks like a longer lag occurred between the indicator, VIX and S&P 500 at the end of the previous bull market, which makes sense because a cycle top takes longer to form. But all cycles have different forces driving them. And this one isn't normal by any means. So based on the weekly chart, there's a high probability that the VIX will start to rise in the near-term and the market will correct between now and the next few months.
The bi-weekly indicator is now higher than it was in 2011 before major volatility erupted in the market. Because the bi-weekly indicator smooths out the year-over-year percent change a bit, I think it's the most important time frame to watch. Make sure to watch RVX/VIX's year-over-year percent change as well (The Key to the VIX).
Go to the first Key to the VIX Indicator
(The FRED charts were reformatted on 11/17/2016.)