With its 52% decline since July7th, XIV has shown again this is not an investment for weak stomachs.
How long might it be before XIV surpasses its recent (and all time high) of 19.43? I looked at the last three significant corrections in the market—the Flash Crash, the Fukushima Crash, and the recent Cap Flash. As the starting point I used the recent high of XIV right before the the correction, the finish point was when the XIV matched that high after the crash. In the graph below I show a line to show the crash/recovery and a spike to show the bottom. I’m assuming the current Cap flash is into recovery…
.
| Start | Finish | Recovery Duration (days) | Days till XIV bottom | Percent Decline | Effective XIV growth rate from bottom | |
| Flash Crash | 31-Mar-2011 (XIV 8.44) | 30-Jun-2011 | 129 | 23 (XIV 3.99 -est.) | 53% | 17% / month |
| Fukushima Crash | 16-Feb-2011 (XIV 15.78) | 20-April-2011 | 47 | 22 (XIV 11.2) | 29% | 45% / month |
| Cap Crash | 7-July ( XIV 19.4) | ?? | ?? | 25 (XIV 9.4) | 52% | ?? |
.
XIV’s recovery from the Fukushima crash was quite a bit faster than from the Flash Crash. I suspect that is because the VIX fall-off was quite a bit cleaner with the Fukushima crash, whereas the Flash Crash had some aftershocks. The number that surprised me was the number of trading days from the XIV high before the decline to the bottom. If this mid 20ish number is a typical number, it could be a significant aid to determining when the market has bottomed. All of these crashes were in a bull market. In a bear market it will probably take significantly longer for XIV to recover.
.
.
.
.
.
.
.
.
.
Looking at historical term structure, this time range (July-Aug 2011) is in backwardation. One would be asking for trouble holding XIV in such periods. Since 87% of trading days are in contango, better stay on the sidelines here.
Agree. Not worth the risk when term structure is in backwardation. Also no contango headwind for XIV at that point.
Lance, let assume that XIV fell 80% during the 1987 crash. Based on your looking at the data how long would it taken to recover then? Also does 9/11/2001 and the 2009 Bear Stearns/Lehman events recover quickly or slow? Thank you
Thanks for the support guys, Actually I dont have a problem feel heavy loses if I believe i’m in the right position, Actually that was exactly what happened here, I was plannig on scaling my price when vix keeps going up but I really didnt take under consideration the huge decay, my average is also around 11 and I still have some more money but now I’m waiting because xiv right now is the worst position in the market no doubt about it, vix goes down , vix goes up xiv is in red:(
Vance, in the current pace of decay I think macro doens’t really matter any more because soon your position would be like 60 -70 % down and then even with things getting back to normal it would take at least one year for xiv to go back to 11, not even talking about see some profit in this position.
What doesn’t kill you makes you stronger… Anyway, I did shift some of my XIV position away to some volatility alternatives. I sold a 26-Aug SPY straddle which derives its premium from a higher short term volatility of 41% compared to 35% on Sept VIX futures which is what XIV is exposed to. Also I shifted some money over to ZIV which seems pretty liquid and does not nearly have the same backwardation losses per unit time as XIV currently has.
Jones, I’m not a happy camper either. I sold at the highs, but then bought back in way too early. Using the Bollinger bands as a guide for where the low end of the range was was clearly a mistake. Significant backwardation is back as of yesterday and long XIV can only win if volatility drops enough during a month to counteract that affect. Moreover if volatility goes up and down day after day, there will be the drag from daily rebalancing. Since the Sept contract is so high at 34.35 as I write, there is a good chance that volatility will drop significantly enough to counteract backwardation and rebalancing losses. In that sense, XIV is ‘fairly valued’. It is not a screaming bargain or an outright rip-off. But if it doesn’t fit your risk profile, it may be time to move to a different investment. Of course, no one wants to hear that, but looking at a case study like 2008/2009 implies things can get worse or take a long time during which you are not sleeping very well.
come one guys, this xiv won’t let me sleep at night. vix is down 3 % as I write and xiv is also down 3 %!!! it’s like each day xiv is losing 5 % regardless to any thing else.
At this pace we gonna reach 2-3 in one month. Ok, I understand backwardation , rebalancing but god, it’s costing the position like 5% each day!!
How are you dealing with this situation? I’m really worried.
Hi Jones, I put less than 10% of my working capital into XIV, and got in for an average price of around 11. Since the rest of my capital was essentially in cash, my overall decline is a lot less than if I have just been invested in the general market. I’m kicking myself that I didn’t anticipate the double peak in VIX. I have mostly been watching VIX since 2009 after the big blow-off. Bill Luby at vixandmore talks about it some.
As I’ve said I think the key issue is macro. No one can know for sure which way the market is going to go. If you think we are going into a recession then getting out and taking your lumps would probably be the best course, other-wise I suspect that XIV will rebound relatively quickly. To make money you have to take a position–there are no no-risk opportunities out there. It is not worth your health in any event.
One thing I will do sometimes is buy something that gives me something to feel good about, even if the market in general is not going my way. For example I just bought some VXX and sold 26-Aug S41 calls against it. It doesn’t cancel out my XIV position, but at least it gives me something positive if VXX goes up or stays flat.
— Vance
The problem is the backwardation and decay and whatever else is just killing xiv, another spike to 50 can take us to 5.0, half of the money of the original trade gone and it could take years until we go back abouv 10 🙁
Vance, with the new low in XIV today, we are 42 days in with no bottom in sight. So staying out for a month is no guarantee and in fact as you computed we have that the 2008/2009 situation had the bottom actually being 131 days in. Maybe the current situation is closer to 2008/2009 than we think.
Hi Andrew,
Yep. I was premature in calling the bottom on XIV. In retrospect the big VIX spikes usually have a double top sort of structure–with the 2nd spike often bigger than the first. I still think the macro-economic outlook does not warrant a bear market, but the market often proves me wrong….
— Vance
Vance, This is great work! If we can understand the character and recovery time for these volatility jumps, we have a shot at avoiding buying back in too early. So it looks like the rule of thumb is to stay out for a solid month after the sh*t hits the fan. Would be nice if data on these volatility jumps could be synthesized back to when VIX futures began trading… One of your previous posts did give a further back insight: In 1987, XIV would have achieved a 100% loss… but maybe that can’t happen anymore with so many traders willing to take the risky side of volatility futures when things go bad…
Hi Jones,
I used the SPVXSTR index values for my simulation, which is virtually identical from a percentage move standpoint to the SPVXSP index that XIV is based on. The index includes all the volatility future effects (overall volatility, contango/backwardation, and daily rebalancing) plus interest earned on unneeded cash put in Treasury bills (invisible!). It does not include the 1.35% annual fee that VelocityShares charges. I don’t include the annual fee in my simulation.
I have a lot of confidence in my simulation because it very closely matches the actual XIV values (if you look closely at the chart you can see tiny bits of red where the simulated and actual XIV values divergeslightly).
I can do a XIV simulation on the 2008 crash–probably pretty close to a worst case situation for XIV.
— Vance
Thanks a lot, this is a great article. Are all parameters here taken under consideration (backwardation, rebalancing, time decay and maybe some other I’m not familiar with)? Is it even possible to know them when xiv was not around during the flash crash?
Can you also give a simulation of the 2008 crisis? that would be the ultimate test because from I under stand vix picked 96 and we had 64 days of backwardation.