## [tech-spec] Again, with apologies for cross-posting from SpecList

• From: Steve Wisdom <swisdom@xxxxxxxxx>
• To: tech-spec@xxxxxxxxxxxxx
• Date: Tue, 26 Oct 2004 17:03:27 -0400

--
Gregg van Kipnis [mailto:gvankipnis@xxxxxxxxxxxxxxxxxx] :

> As I recall, vol moves up November in election years

I know Dr van Kipnis plans to augment this one-liner with some
numbers, but I'll take the liberty of beating him to the punch

I put monthly VIX at http://dailyspeculations.com/VIX.txt

The R script below reads the file and calculates average monthly
changes in VIX.  Looks like VIX has risen 2 vols in pre-election
Octobers, albeit with only 3 obs, which bucks the tendency of VIX to
fall late in the year, on average.  Probably correlates a bit with
Y-le's `sell in May and go away' effect, but I'll leave that for him
to study

--

VIXE<-  VIX[!VIX\$Year%%4,]
matrix(with(VIX ,by(Chg,Month,mean))) #All years
matrix(with(VIXE,by(Chg,Month,mean))) #Election only
R.version.string

>
> VIXE<-  VIX[!VIX\$Year%%4,]
> matrix(with(VIX ,by(Chg,Month,mean))) #All years
[,1]
[1,] -1.2114286
[2,]  0.4086667
[3,]  0.1293333
[4,] -1.2213333
[5,] -1.0426667
[6,] -0.6560000
[7,]  1.7006667
[8,]  1.7806667
[9,]  1.4413333
[10,] -0.7842857
[11,] -1.4721429
[12,] -0.1128571
> matrix(with(VIXE,by(Chg,Month,mean))) #Election only
[,1]
[1,] -0.8175000
[2,]  0.0325000
[3,]  1.0675000
[4,] -0.2900000
[5,] -1.4175000
[6,] -2.0425000
[7,]  1.9450000
[8,] -1.4925000
[9,]  0.6050000
[10,]  2.0300000
[11,]  0.6366667
[12,]  0.1800000
> R.version.string
[1] "R version 2.0.0, 2004-10-04"
>

### Other related posts:

• » [tech-spec] Again, with apologies for cross-posting from SpecList