Interpreting Strange Movement in Options Volatility
10/13/2010 12:01 am EST
If there’s one point I would emphasize when VIX watching, it’s to pay the most attention when it does something unusual.
Most times the VIX does the “expected.” It lifts when the market tanks. It tanks when the market lifts. It rises ahead of an expected news event and then declines after the news comes out. It declines on Friday afternoon as traders seek to avoid paying too much for weekend decay on SPX options, and then it rallies on Monday as the calendar catches up to the options.
But Monday, it did the unexpected.
On Friday, the VIX behaved as it usually does, i.e., it declined on a Friday into a strong market. In fact, it declined to a five-month, post-flash-crash low. Then, on Monday, the market opened up modestly and traded within a small range for the rest of day.
If you had me take VIX off my screen and then speculate as to what it did based on the market action, I would have guessed that it was up small early in the day and then drifted. And I would have said that the only reason for the rise in the morning was the offset of the Friday decline.
Yet, the VIX did nothing of the sort. It opened weak and got weaker. And, again, this was off five-week lows. This is one of those times when we need to sit up and take notice.
What should we notice? Well, that’s a trickier question.
The VIX is exhibiting clear signs of complacency. We consider the VIX oversold when it closes 10% or more below its ten-day simple moving average (SMA). Well, it closed about 14% below its SMA. That’s bearish for the market, but I wouldn’t go loading up on SPDR S&P 500 shorts just yet.
“Complacency” does not work as well as an indicator as its evil cousin, “fear.” Fear tends to resolve relatively quickly, while complacency can linger and linger. Likewise, complacency is consistent with the larger uptrend in the market, and thus, in a way just confirms what we already know. A fear spike within a market uptrend provides a better short-term signal than a complacency spike.
And, finally, VIX futures still don’t buy the VIX complacency angle. November futures carry a $6.50 premium to the actual VIX, which is at or near a record level of premium for a future this close to expiration.
So I’d suggest that while the ugly VIX might signal a pause in the rally, it’s highly unlikely it predicts a top.
Article Continues on Page 2|pagebreak|
Differential Research notes that all was not as it appeared on the VIX decline.
“The front month October 2010 VIX future fell only .65, a substantial difference considering the VIX index derives most of its value from the same November 2010 SPX options that are used to value October VIX futures.
The balance is the difference between short, shorter, and shortest implied volatility, the shape of the implied volatility forward curve, and a unique feature of the way that the VIX Index is calculated. Today was the morning that the SPX options used in the VIX Index calculation switched from the first and second to the second and third monthly series. Today, December SPX options replaced October SPX options, joining November SPX options in the VIX Index calculation. This switch happens every month ten days before VIX expiration, every month!
While the intent is to avoid extra volatility in the calculation as we near SPX expiration by removing the expiring series, it renders a momentary incongruity where the VIX Index on Monday morning is not the same as the VIX Index the previous Friday. More specifically, for most of the monthly cycle, the VIX index adds the calculated contribution of the first and second months’ SPX Index options for the last ten days before VIX expiration, the first month is dropped, the second month weighted over 100%, and the third month’s contribution is subtracted.”
Is he correct? Well, that dip did seem to just happen, as opposed to developing over the course of a day. So that suggests he’s correct in that there’s some sort of calculation quirk. I believe VIX changes over eight days prior to SPX expiration, which was last Thursday into Friday. So why did it apparently hit Monday? Not sure. He notes that it’s ten days ahead of “VIX-piration.”
Again, the price action in VIX is consistent with a calculation “blip.” VIX futures do not care why VIX does what it does. And they dipped yesterday, but not to the extent of the VIX itself. So that too suggests VIX overstated the drop. Then again, VIX futures perpetually carry large premiums, so what’s another $1 with VIX down here?
So let’s just conclude the VIX is a statistic, subject to the mercy of all sorts of quirks that make it too imperfect to read gospel into every single tick. It is demonstrably low versus any level we’ve seen since the (May 6) flash crash. It is demonstrably high versus realized volatility that sits in the 12-13 range. And yet VIX futures perpetually anticipate a significant VIX lift.By Adam Warner of DailyOptionsReports.com