Thursday, May 7, 2009

Testing VIX Model

Headknocker is finished with the preparations for the softball game tomorrow.  I thought about walking around the block, took out the various SPRIs from their boxes, and glanced at the cover of the Exercise Guide Book.  So I am ready to play ball now.

The volatility indexes are up today more than the market is down, a frequently observed phenomenon.   Conversely, I would expect a fall in the volatility indexes with a rise in the market.   The rise in the market since March has been correlated with a fall in the VIX which is a positive under the model.   

The volatility indexes may meander around unchanged when the market is close to flat.  It is not unusual when the DJIA is up say 15 points and its volatility index to be up some.   I would not expect however to see VXD up 3% with the DJIA down 15 or up 15.    On down days, like today, I would anticipate seeing the volatility indexes moving harder in the opposite direction, meaning up in price.  

For purposes of the VIX allocation model, I am interested in the level of volatility, its direction, and the duration of  significant events.    A significant move out of the Stable range of below 20 is extremely important in the model.  It is signaling that there is more risk in the market, volatility has increased, and with increased volatility and risk there is a significant danger of lower prices ahead, probably a bear market on the horizon. 

 I challenged the readers in a prior post to make stock allocation decisions from 1990 to date based on the VIX Asset Allocation Model, which I view as obvious, as you move the cursor over an interactive chart starting in 1990, adding to stock holdings or reducing them based solely on what you see in the VIX chart, with no other consideration.   Based on the model then add to positions when movement below 20 occurs, waiting 3 months for continuous movement below 20 (others may want to vary that period), and then follow what I call Trigger Events, taking stocks off when the movement falls back below 20, and so on.   

The VIX is signaling that equities have more risk than you might think when you see a Whipsaw Pattern at elevated levels emerging out of a Stable VIX Pattern.

I had a comment that the Stable VIX Pattern shown in the charts contained in this post do not look Stable. Vix Charts from 2004 2005 2006 Stable VIX Patterns Phase 1 and Phase 2 Anything below 20 on the VIX is stable.  I am only interested in movements in and out of the two Phases in the Stable VIX Pattern.  I would generally anticipate a move first into a Phase 1 Stable Vix Pattern range first, coming out of the current  Unstable Vix Pattern.  Then, as people become more comfortable with the bull market, which should then be underway, the VIX can meander down into Phase 2, the most stable phase, with occasional spurts back into Phase 1 when something spooks investors.  The movement out of Phase 2 to Phase 1 is just an Alert requiring some thinking person to try to learn everything possible about the event causing the Alert and how that might impact future security selection and asset allocation.

It was a snap to figure out what caused the Alert in February 2007.  I had not opened the VIX chart by then, I believe.  It was around this time of year in 2007.  But readers of this post know that I read a great deal, and can reasonably assume that I was very aware of what was happening in the mortgage market in early 2007.  I had to actually look at the VIX chart to see the pattern, and then the model was as I say obvious. The model was formulated in about 10 minutes.  This is not something that I view as difficult.   The work has been trying to figure out what to do once you see that increased volatility means stocks as an asset class may need some adjustment.  

ADDED 3:10 5/7 :  A more appropriate term to use in characterizing  a spike in volatility out of the Stable VIX Pattern to anywhere around 30, would be a CRITICAL event under the Model rather than just "extremely important."   There have not been many of the Trigger events since 1990.  Ignore them at your peril. If you were sleeping before they happened, then someone needs to slap you and scream at you to wake up.  Trigger events occurred in two phases prior to the bear market starting  in 2000 and there were two prior to the onset of the current bear market in 2007.  
Prior to data on the VIX, which goes back to 1990, there was volatility data on the S & P 100.  There was one Trigger event before the crash in October 1987. 

No comments:

Post a Comment