Well some of us are clearly not interviewing to be market makers so care to illuminate on this?
unusual VIX term structure
- Thread starter ajensen
- Start date
Well some of us are clearly not interviewing to be market makers so care to illuminate on this?
Sure, no disagreement there. My small issue is with J getting massively bid up over K, that's basically it. By Wednesday of next week, said events will still be near and J will be off the calendar.
What's your take on mid-month cal moves starting around 3/28-3/29, e.g. MN, NQ, etc? General repricing of the curve that's unrelated or at best coincidentally related?
I agree with SIV. In general, the term structure in volatility follows the general theory of uncertainty. That is, in general vol is in contango because we are more certain of the short term then the long term. And the tenor in vol reflects that. That gets distorted in general around earnings, FED meetings, FDA meetings and stuff like perhaps the debt ceiling or a major election. In these examples, the short term offers more uncertainty then the long term and you get some form of either backwardation or a move towards it.
With regards to the French election, logic would follow that as we get more information, the tenor in vol should reflect that and it does and it will. Now I'm not speaking about vol in general, just the shape of the curve. After the election the curve will move "towards contango". Maybe it takes a day or two or even a week to get fully back there.
Bayesian probability is a huge topic for option traders and it's difficult to get hired anywhere without demonstrating at least an intuitive understanding of it. It's basically conditional probability. What is the probability of B happening given that A happens. It's NOT the same probability of B happening alone. In laymens terms, you have to account for variable change. Think of the famous Monty Hall interview question that was featured in the movie "21" about the blackjack card counting team in which the question was asked, if there are 3 doors and behind one door is a million dollars and the other two doors have a goat and you choose one of the doors. Now Monty will open one of the remaining two doors (obviously one without the money) and ask you if you would like to switch doors. The question is, should you switch? It's actually an extremely complicated question mathematically speaking but the general theory is yes, you should always switch. With the simple reason being that you have more information now then you did before the reveal and the probability actually increases.
People often get confused with probability and information and an interpretation of that information. Case in point our election. Obviously vol was bid up going into the election but many felt if Trump won, that the event outcome itself would generate even more vol. The fact is, the event outcome is usually immaterial. The fact is, after the event, you now have the information and the uncertainty has been greatly reduced. The result was vol got crushed.
So back to the French election, right now we have the most uncertainty. After April 23rd most of that uncertainty is gone yet we still most likely will not have a winner. But we will know exactly who didn't win. The fact that there will be only two candidates and that runnoffs usually go to the favored, in this case Macron, vol should reflect this. The one trickier part that can happen in runnoffs that happen frequently in US elections is weather. If there is really bad weather on the day of the runnoff, it CAN lead to a surprise. This is not expected to be an issue. The other uncertainty I can think of is obviously a terrorist attack which many feel would give a boost to LePen. Obviously these things can't be predicted and so more often then not the market doesn't price that stuff in.
To sum up, it's not the magnitude of the event that matters but merely the condition that such event happens. And that is what Bayesian Probability is all about. Making the proper adjustment to the probability given that you now have new information. This is actually a good conversation and important topic. There is no right or wrong here, just theory and opinions. I'm giving you mine. And as I like to say, I'm often wrong.
quite honestly the middle, back imo is underpriced when u see oct @16 handle for instance. all i can try to do is extract money by selling the humps in the curve, buying the dip or cheaper parts and hoping they adjust. this is what me and rally have done for years; but it seems a lot harder on this super low vol regime.
If you understand forward vol then what is your issue with J being bid up massively over K? On Apr 19th, the vix will settle to options that encompass both rounds of voting and since J settles to that same value it's getting bid up. On that same day and despite the fact that J expired and K is now front it will still, theoretically, reflect the expected vol in mid June. June! That's a month after the elections and seeing how the Brexit and the U.S. election selloffs got massive reversals within days is it a surprise that no one is interested in paying up for protection a month after an event?
Having said that, I shared your frustration a week ago when K was at 13. Here at 15 and a half it's probably fairly valued all things considered. If not a bit rich should Frexit turn out to be a non event. As sellindexvol said, I'd be a buyer of longer term vol like sep or oct. IMO, that's where the value is if you are worried about event risks over the next few months.
Last edited:
I guess where my thought process is coming from when it comes to the "theory vs reality" part is that yes these things are priced off of 1mo forward-vol - but pragmatically if shit hits the fan uncertainty-wise, even if for less than a week, we'll no doubt see spot VIX spike and K going along with it. I do get Mav's comments that most of the premium is being driven by uncertainty and even simply removing uncertainty is enough to change the outcome - atleast to get to the next uncertainty after that (event uncertainty vs outcome uncertainty). So while it does make sense that J might be getting the lion share of the bid given where it happens to expire vs the election dates and the whole 1month thing, it does seem a bit odd that K isn't just a bit higher than it actually is right now. But then again, this is all priced off of *expectations* of future volatility based on a strip of options, so it's easy to get trapped by unintuitive factors.
Yep, I'm already long VXU7. I'm also long K via May 20 calls I picked up a few weeks back as well. Might be thinking harder about dumping a good portion of those by the end of next week.
The real question is who has the balls to go against the crowd and put on short vol position believing that after the election becomes much ado about nothing, the curve reverts... I have short vols hoping for such a scenerio being a contrarian myself. Sometimes when everyone goes right and right seems incorrect after the news the stampede left can be profitable.
If not...close and move on to next month
If not...close and move on to next month
I hear you. If there is a repeat of Brexit K could go >20. And yet here we are at 15. I think the prevailing buy the dip mentality is having an effect. No one seems to care about path dependency anymore as long as they feel confident it will mean revert fairly quickly. It sure is frustrating if you are on the long vol side of the tide which is why no one bothers buying vol anymore. sellindexvol is long vol for the first time in like 5 years and i give him 50/50 odds of a win at best
Last edited:
This is literally what the hordes of vol sellers are actually doing and why I, and some others, think it's a good time to actually be long vol - and if not at the front, further out as @rallymode and @sellindexvol66 mentioned.
By short KM and KMN I mean KM cal or KMN fly not a strip. If indeed the curve is going to steepen back into contango we'd want to exploit that curve shift.