Tuesday, January 08, 2008

Another look at the VIX

I haven't looked at the VIX for a while, but this 3 linear regression study (30,11,3) of the weekly bars gives something to ponder. Given the upslope of the VIX since the beginning of 2007, the 30 bar channel has been a good overbought/oversold tell. Disregarding the current oversold character of the markets, the position of the current VIX bar suggests that increased volatility is the likely trend for the near term. Based on the 3 LRs model, a reversion to the mean would take the VIX to the 27.50 area. . which would translate to a significant decline from current market levels. This concept is at odds with my previous VIX trading methodology which revolved around the VIX's propensity to revert at least 50% back towards the mean within 4 days. . .the mean being the 10 and 20 day SMAs. . when an 18% boundary had been violated. Although that system worked well going into late summer (82% reliable for the IWM)), the August and November performance was rather pitiful as the inverse correlation between the VIX and market returns failed to hold up. In hindsight, (always the best way to trade) the 3 LRS study would have proven the better choice to time my swing trades (2-4 days) and I am adopting this directional model for my current short term trades in the Qs. If nothing else, the 3 LRs model suggests exercising caution before entering those "can't lose" near term longs. "Better safe than sorry", as my ole Pappy use to say.

No comments: