Volatility ETPs misbehaving?

stevenpaul · Jan 24, 2018

Just curious about how these things work. I notice that often VXX, XIV, and the others in the space sometimes move by different percentage amounts (and in different directions) than the Vix futures.

For example, early today the Vix futures front and second month were down by about 1%, yet VXX was up by close to 1% and XIV was down by about the same amount. The VIX itself was down, and the ES futures were up. I would expect VXX to be down and XIV to be up.

If these ETPs track the performance of the Vix futures, how can they be out of line with the those futures so dramatically? Also interesting is that the entire space was uniformly out of line (VXX, TVIX, UVXY, SVXY, XIV, TVIZ, and ZIV). Thus, across ETP companies, the aberration from the stated tracking goals has presented itself, often persistently. (And not just today either, but ubiquitously in recent days and weeks.)

Any thoughts on how and why this happens?

JackRab · Jan 24, 2018

Are you sure you're looking at the correct data? I don't see anything weird about VXX in the past week. Pretty much tracks the Feb and March futures of VIX.

jeb9999 · Jan 24, 2018

"For example, early today the Vix futures front and second month were down by about 1%, yet VXX was up by close to 1% and XIV was down by about the same amount."

That is exactly the correct behavior. Don't see why you have a problem with it.

JackRab · Jan 24, 2018

jeb9999 said:
"For example, early today the Vix futures front and second month were down by about 1%, yet VXX was up by close to 1% and XIV was down by about the same amount."

That is exactly the correct behavior. Don't see why you have a problem with it.

No... If Vix drops than VXX should go down as well and XIV up, since XIV is the reverse ETN.
OP noticed the reverse... but I can't see that... in my data I see both move correctly.

jeb9999 · Jan 24, 2018

JackRab said:
No... If Vix drops than VXX should go down as well and XIV up, since XIV is the reverse ETN.
OP noticed the reverse... but I can't see that... in my data I see both move correctly.

Yes you are correct, that quote I quoted is incorrect. Thanks for pointing out my clear error.

My brain was working with what actually happened this morning. Look at a chart, the VIX futures were UP 1% in early trading not DOWN 1%.

The correct statement should be: "For example, early today the Vix futures front and second month were UP by about 1%, yet VXX was up by close to 1% and XIV was down by about the same amount."

stevenpaul · Jan 25, 2018

Well, maybe I just wasn’t patient enough. I’m absolutely positive that at various times in the last few days the futures and the ETPs have been off significantly, to the tune of a 2% spread, but perhaps eventually closed up. What was striking to me is that all the ETPs were in line with each other yet way off from the futures. If it happens again I’ll have to take a screenshot.

JackRab · Jan 25, 2018

So at what time and day roughly did you notice that? And for how long?

ajensen · Jan 25, 2018

stevenpaul said:
Well, maybe I just wasn’t patient enough. I’m absolutely positive that at various times in the last few days the futures and the ETPs have been off significantly, to the tune of a 2% spread, but perhaps eventually closed up. What was striking to me is that all the ETPs were in line with each other yet way off from the futures. If it happens again I’ll have to take a screenshot.

Regular trading in VIX futures ends at 4:15pm EST, 15 minutes later than for ETFs, so if futures prices move in that time, the returns measured from yesterday's close will differ for futures and ETFs.

stevenpaul · Jan 26, 2018

No this was all during the regular trading day when all the markets were open. It was most pronounced in the premarket hours, but that could relate to illiquidity.

I noticed the effect again today, although in a more muted way. The first shot is from 10:19 am Eastern. VXX is unchanged here, although had been fluctuating, XIV is up, the VIX is down, and the VIX futures are down in the front month. The VIX is irrelevant, I suppose.

Half an hour later, the markets have moved, but the tracking error has persisted. VXX pulled back less than the futures, while XIV was up more than the futures were down.

I’m probably making a mountain out of a mole hill, but the gap between XIV and VXX closed by the end of the day. Actually XIV was down and the VXX was also down. None of it resembling the futures very closely.

JackRab · Jan 28, 2018

I think you're not aware of how VXX is calculated.

All VIX related ETN's want to track the VIX index, but since you can't trade the index itself, they track a mix of the futures. That mix depends on the timeline. VXX needs the 30 day VIX, which currently is about 60% Feb future and 40% March...

So, if you look at the prices at those times...
VXX 27.82 to 27.54 = -1%
Feb 12.40 to 12.25 = -1.2%
Mar 13.00 to 12.90 = -0.75%
So VXX is tracking fine
XIV 139.24 to 141.22 = +1.4% seems a bit low... but I don't know if it uses the same mix of futures.

The tracking errors can also be bid/ask error... last traded price... I don't know how liquid those are and if the spread is wide. The spread in futures is also 0.05 cents, which is 0.4% in the March future.. so just a tick down or up means a move that likely is not reflected directly in the ETN's.

Again, looking at longer timeline it seems fine to me... but if you think you can arb the error... go for it!