Friday, May 8, 2009

More on the Vix Model: What it Does not Predict is as Important as What it Does/Parallels to VXO 1987-1988

LB hopes that the Lord has not decided to wake it up every morning with a thunderstorm at 4 o'clock in the morning. Sleep deprivation is not what concerns LB, having to listen to RB another 2 hours is starting to wear LB down.  The very first words coming from RB this morning, and I kid you not, was "I told you to go all-in on National Dentex at $1".  LB replied "I thought that was a joke" And then RB said "Swap Land for Water". So begins the day here at HQ before the Headknocker even opens his eyes to greet another rain soaked day. Headknocker did wonder, possibly trying to change the subject away from RB's endless banter, whether the Mexicans had forgotten to mow his yard, since it looked like the fescue was turning to seed.    

The good news is that the softball game will most likely be rained out. Headknocker was suspicious that his training exercises yesterday were a sufficient warm-up for all of that running stuff. 

Once there is a Trigger in the VIX Asset Allocation model, the model simply says that, no matter what you think or believe, stocks have become considerably more riskier as an asset class, and the investor needs to do something. The Model does not predict how severe the downturn will be or its length.  It is possible that the VIX will in the future find stability soon after a Trigger Event, with the S & P 500 continuing a bull run. In which case, the Model would then flash a green light again. Another important point is that the Model does not predict whether an investor will be given an opportunity to lighten up at better prices.  Historically, an investor could wait for better prices after the Trigger Event to sell stocks and to change the asset allocation. 

The 1987 crash was preceded by a Trigger Event in the volatility index for the  S & P 100.  I have not studied the volatility index for the S & P 100 except for the period prior to the start of the CBOE data on the VIX. I prefer to make judgments based on the VIX rather than VXO, since the VIX is more representative and has a larger sample. But I did look at the VXO in 2007 after developing the model on the VIX and the other major indexes.

On October 19, 1987 the VXO closed at  150.19. Well, that had to be a bad day for those long stocks. On October 7th VXO closed at 22.09. So, then I wanted to know about any Trigger Events that may have occurred  before the Cataclysmic Event, moving directly from a reading in the mid 20s into a Phase 2 of the Unstable Vix Pattern. I am just looking at the historical prices for the VXO at Yahoo Finance and scrolling back in history. I am seeing a pattern as I scroll back through September and October 1987 of worrisome readings in the low to mid 20s. Then I see readings in August and July 1987 that look stable in the Phase 1 Stable VIX Pattern. This would indicate to me without looking at the S & P 100 average that the market was moving up. Then scrolling back to May and June, I now see the problem.  The VXO is out of its Stable Pattern, with a lot of movement in the mid to high 20 range. Then, I see the Trigger Event:

May 4  27.18
May 1  28.22
Ap 30  28.45
Ap29   29.22
Ap 28  31.2
Ap 27  31.46
Ap 23  29.13
Ap 22  27.95
Ap 21   27.22
Ap 20  27.18
Ap 16   27.48
Ap 15   27.59
Ap 14  28.97

The severity and duration of this event qualifies it as a Trigger Event under the Model. In this particular case, the market was not letting off some steam but sounding a Warning. On this occasion, there would not be another warning, one and done. There was an opportunity to sell at higher prices as noted above before the Crash in October 1987. This may or may not happen again, and LB does not make assumptions about the past repeating itself, just because it has done so more than a couple of times. 

The period from 1987 was a major disruption in the calm that preceded it. The volatility readings were just off the charts, much worse than anything seen in the VIX in 2007 to 2008. Possibly the VIX would have had similar readings in 1987. But, with these major disruptions in the calm and stability, the 1987 event does suggest that time will heal the damage, but it will take time. Calm was not restored in first half of 1988. By September 1988, the VXD manages to find its way to readings below 20 but it is still on edge. So it took about a year. The one year anniversary of our current Cataclysmic Event will be in the later part of September 2009.