VXST and the “Weekend Effect”

Whenever we have a Friday or Monday market holiday and a three day weekend is on the horizon I will often remind traders about the impact that the extra day off should have on VIX.  For those new to VIX, the calculation that determines the value of VIX is based on calendar days, not trading days.  The result is a bit of a headwind that will compress the value of VIX into a weekend.  For three day weekends that headwind is exacerbated by the extra day off.  VXST is calculated using the same methodology as VIX, but the result is a nine day measure.  With VXST measuring a shorter time period that VIX, the weekend effect appears to be a little more dramatic.

CBOE compiled closing prices for VXST going back to the first day of 2011 which allows for studies of the price behavior of VXST over a three plus year period.  For those keeping score (or checking my work) the specific history I used was January 1, 2011 through February 14, 2014.  Over this time period there have been 142 ‘normal’ weekends.  A normal weekend is a two day, non-holiday weekend.  Using that data I came up with the following results –

VXST was lower 90 of the 142 Fridays before a normal weekend, or 63.38% of those days.

VXST was higher 103 of the 142 Mondays following a 2 day weekend, or 72.54% of those days.

Just to get an idea of how these numbers stack up relative to an average trading day I took a look at every Tuesday, Wednesday, or Thursday that did not fall before or after a holiday.  There were 447 of these ‘normal’ trading days over the time period in the study.  Here’s the outcome from looking at those days –

The day over day close for VXST was higher 177 of 447 of those trading days or 39.60% of the time.

VXST closed lower 268 of those trading days or 59.96% of the time.

To me, what stands out is the difference between the percentage of higher closing prices on a Tuesday, Wednesday, or Thursday relative to the number of Mondays where VXST is higher on the day.   Over 70% of post two day weekend Mondays we see VXST close higher, while under 40% of non-holiday mid-week days the VXST one day changes is higher on the day.

OK, the numbers are fun, but why is this important?  Knowing the weekend impact on the spot VXST index is important for traders taking positions in VXST futures (and pending regulatory approval options).  We are only a few weeks into trading for VXST futures, but so far it appears that the futures markets are properly adjusting for this weekend effect.  As time goes along and we have more data, I’ll do more digging and share any findings in this space.

  • Numan Ulku

    Hi Russell, Can you check the accuracy of your sentence ” …, the calculation that determines the value of VIX is based on calendar days, not trading days.” In early 1990s it was calculated based calendar days; several articles published have showed problems with this calculation; now (for about 15 years) CBOE calculates it based on trading days, as far as I know ?