Showing a monthly VIX chart with only 1 indicator, a 10 period MA in aqua, and a red line drawn at the lows of that level in 1993 and 1995.
Main point is that the VIX low in December 1995 was a very long way from a meaningful top in stocks, and if you were bearish stocks because of rising VIX that was a massive mistake. This would have cost gains on one of the epic stock market runs. VIX was rising quite decently in 1996 and 1997, then actually declining in 1999 and early 2000 when you'd think it would be rising.
In other words, even though I've been a big fan of VIX in the last few years, it didn't seem to work that well in the 1990s as a supplemental indicator.
Next chart from 2000+ is same idea, with the VIX low in early 2007, rising meaningfully in the second half of the year, then VIX highs right where they should be in 2008. Seems fine in this millennium.
So, VIX doesn't have to work, but has been in the 2000s which I will take. The 10MA is on the lower side of the last 17 years, but not as low as it was in 2005 yet.
Using another technique and one that I actually prefer, VIX has reached levels of previous quarterly close lows. Here is a quarterly chart with a line drawn at 11.26 (or as close as i could get it), which was the low in 1993 Q2. All closes since then have been higher, and you can see clear support near this level on many bars.
In the current 2017 Q1, in addition, VIX tagged its quarterly Bollinger band, which seems to have never happened before.
I rather dislike sounding bearish alarm, because often these turn out to be wrong, and a bull market will find a way to come back again and again. But I'd add this quarterly VIX chart to a list of a few concerns that the next drop could be more than the mild 3-5% range that was my initial expectation.