The other day, I heard about a new volatility index introduced last March by the CBOE: the VVIX.

This CBOE micro-page on VVIX introduces this index by saying, "Every asset class deserves its own volatility index, including volatility itself."

The VVIX measures the anticipated volatility of the 30-day forward price of the VIX. Just as the calculations for the VIX depend on an algorithm applied to nearby SPX options prices, the VVIX depends on the same algorithm applied to nearby VIX options prices. It's the expected volatility of the volatility.

How can we use this information? The CBOE wants us to consider selling a VVIX portfolio--a portfolio of VIX options--regularly to "capture a volatility risk premium." Not me. Or perhaps, as a white paper found on the site suggests, we could buy "a short-dated VVIX portfolio" to ameliorate losses that could occur from an explosion higher in VIX (R) futures prices ("Double the Fun with CBOE's VVIX (SM) Index"). I'm not going to be doing that, either.

I have at least two reasons for not acting on those suggestions. Although the white paper authors have reconstructed the VVIX going back to June, 2006, the product has actually been available only since March of this year. The CBOE-published FAQs conclude that "there has been little correlation between the VVIX and the VIX (R) except at extreme values of the VIX (R)." We just don't have a lot of experience observing it yet, so I'm not going to be found either buying or selling a VVIX portfolio. Furthermore, those who have traded VIX-related products have found that its non-standard expirations and tie to VIX futures rather than VIX spot prices have made for some wonky correlations.

However, others have proposed another use for the VVIX. Someone out in the cyberspace supposedly proposed that any time the VVIX dipped below 90, the SPX was at risk of rolling over.

I recently told someone that I should have been born in Missouri, the "show me" state, because I don't tend to believe what people tell me until I verify it for myself. What I learned after going to the CBOE's site was that the VVIX is mean-reverting and the long-term mean is either 85 (white paper) or 86 (FAQs). "Since the flash crash of May 2010, the VVIX has rarely dropped below 80," the white paper also concluded.

Taken together, some implications might be proposed, with a goal of ascertaining if they're valid. First, if the mean is 85-86, it would seem that the 90 benchmark is too high. I also took a look at a chart comparing the VVIX level to the SPX level. Remember that my articles are roughed out a week to two weeks before publication, so the charts do not reflect current values. There are pros and cons to this, as we can look back and test whether current conditions followed the predictions made or not made by those charts.

VVIX Versus SPX:

Let's see what we can see. The top black horizontal line crosses at 90; the bottom one, at 86. The green, red and blue columns identify times when the VVIX dipped below 90 (left edge of the column) and then crossed it again heading up again (right edge).

The green column marked a time when the SPX, indeed, was about to roll over and roll over sharply. The dip below 90 in late April perhaps did offer a warning that the SPX was about to roll lower. The warning was a bit early, but the SPX did not make much upward progress after the VVIX dipped below 90. If the benchmark were lowered to 86, the warning was even more timely.

No such signal was offered in early April, however, before the green column, when the SPX was peaking. The bearish trader who could have taken advantage of the rollover from the late April/early May period likely would not have cared, when all was said and done. That trader still had lots of gains to make. The bullish trader is the one who would have liked to have been prepared for that first jolt lower.

The red column also marked a time when the VVIX dropped below 90 and then soon below its mean of 85-86. The VVIX dropped below 90 on 6/29 and the SPX peaked that time on 7/3. However, the pullback was brief.

On 8/2, the VVIX dropped below 90 again, in the period marked by the blue column. It quickly dropped below the 85-86 mean before turning up again. The SPX did not, however, roll down. It continued higher until the middle of August. At the time that chart was snapped, on 8/23, a small pullback had begun. Since then, the SPX dropped a bit lower, rose and has dropped again, but the pullbacks have all been small ones. We're now almost a month past that so-called warning offered by the VVIX. If it offers a warning, it's not a particularly timely one.

What can we conclude? Remember that quoted sentence from the CBOE white paper: "Since the flash crash of May 2010, the VVIX has rarely dropped below 80." The white paper's authors and researchers might have had access to the data from which to calculate those values, but we retail traders have only been able to watch the VVIX for a few months. What that sentence tells us is that those few months come under the umbrella of the time since the flash crash of May 2010, and the charts we're able to examine might not tell the whole story. We should be hesitant to draw too many conclusions based on these few months of data available to us on our regular charting programs.

Perhaps it's a good idea to get in the habit of watching the VVIX. It may become a useful tool. I know people who have set up their own proprietary indicators to show when volatilities have become more volatile, and they feel that those times offer a solid warning to the retail trader. I'd classify it in the "Hmm, interesting, and I'll watch this to see if anything comes of it" category for now.

You may hear the VVIX talked about as I did. When you hear things such as the "SPX will roll over when the VVIX dips under 90" admonitions, always check it out for yourself. We all love this stuff or we wouldn't be trading options, so we're vulnerable to jumping on the next new indicator.

It will be helpful to know when the VIX itself becomes more volatile. I don't know if we yet know how useful this particular indicator will be. But that's just me, always a bit skeptical. I am also interested, so I'll watch.