We have noted many times the poor structure and performance of the iPath S&P 500 VIX Short Term Futures ETN (NYSE: VXX). But as you can see in the chart below of VXX (black) versus VIX (yellow), it has performed even worse than we could have imagined in the past three months.

Since July 21, VXX has pared a jaw-dropping 45%, while VIX has lost only 20%. That’s incredible when you consider VXX should only move about half of VIX on a given day. That guideline apparently only works on days when the VIX is up, and we have not seen too many of those since July.

Let’s say you were unlucky enough to own VXX back then at 27. How far would VIX, which was at 24 in mid-July, have to rally to get you back to even? The answer is pretty darn far.

The course of VXX is very path dependent. That is, there’s no rule that says if VIX is “X,” then VXX will go to “Y.” It depends how long it would take VIX to get to “X,” and how closely nearer-term VIX futures track the move to “X.” In other words, if futures believe VIX will stay at “X,” they will rally sharply. If they believe the move to “X” is a one-off pop that won’t sustain, they won’t track so closely.

So let’s say VIX doubles or so to 40 in the next week. It’s highly unlikely, but it could happen. The VIX nearly did that in May, albeit with a lower starting level.

Right here, right now, VXX predominantly consists of November VIX futures, so we can use those as a proxy for VXX.

What would November futures do if VIX soared to 40 in fairly quick order? They would track some of that move, but likely maintain a decent discount. No one is assuming a 40 VIX will hold for a month right off the bat. So let’s say November VIX futures go to something like 33. Well, conveniently, they trade at 22 now, so that’s a 50% lift. If VXX exactly proxies that move, it will lift 50% … all the way back to 20.

Yes, that’s right. In mid-July, VIX was 24 and VIX was 27. In a late October crash, that moves VIX to 40, and VXX will have trouble even getting over 20. And trust me, I leaned toward optimistic assumptions here.

In a week, VXX will contain more December VIX futures than it does now. And if VIX explodes, the VIX term structure will invert from the current contango to more of a backwardation. And December VIX futures now trade 24.5. So they may only rally about 33% or so in this VIX pop, further dampening a potential VIX move.

To get VXX back to that magical 27, you might need to see VIX pop to the 50s, or even the 60s. It all depends on the timing. A slower pop in VIX would ultimately work better for VXX, but of course, that also means you would need to wait longer for the VXX pop.

In short, VXX sux as a trading vehicle!

Follow Adam Warner on Twitter @agwarner.