Dispersion, Correlation, and 0DTE 

Mandy Xu, Head of Derivatives Market Intelligence at Cboe, once again joins Steve Sosnick, IBKR’s Chief Strategist, for a wide-ranging discussion about volatility, options trading, and Cboe products that track market correlation and dispersion.

Summary – IBKR Podcasts Ep. 123

The following is a summary of a live audio recording and may contain errors in spelling or grammar. Although IBKR has edited for clarity no material changes have been made. 

Steve Sosnick 

Hi everybody, welcome to the latest edition of IBKR podcasts. I’m your host today, Steve Sosnick, chief strategist here at Interactive Brokers, and I’m thrilled to have as my guest today, a return guest: Mandy Xu, head of derivatives market intelligence at the Cboe. Hi, Mandy, how are you? 

Mandy Xu 

Hey Steve, thanks for having me back on. 

Steve Sosnick 

It’s my pleasure. I enjoyed our conversation so much last time that I’m glad we’re able to do this again. 

Mandy Xu 

Yeah, same here. 

Steve Sosnick 

Well, let’s just jump right into it. When we look at markets, the bond market is saying one thing, particularly in terms of volatility, while the stock market is saying something else. When we look at something like the MOVE index, that’s definitely getting more elevated in recent sessions. By the way, just so everybody knows, we’re taping this on Thursday afternoon, December 7th, in case you’re listening to this later on. But since the start of December, we’ve seen the MOVE index really leap higher while VIX just remains in the doldrums. I’d love to hear your thoughts on that, Mandy. 

Mandy Xu 

That’s a great question and I think — especially coming off of a month like November where everything kind of rallied together, we had this historic rally in bonds and equities, and credit across asset classes — the divergences we’re actually seeing in the volatility markets of these various asset classes, I think it’s very interesting, right? So, as you pointed out, on one hand we have equity, and I would also put credit in the same camp, where volatility has just completely collapsed.  VIX went from the mid-20s in October to a low of 12.  Same with credit volatility. If we look at our credit VIX indices, investment grade VIX IG as well as high yield VIX HY index both fell by over 50% in November. So those two asset classes are clearly in the more bullish camp in terms of where they’re pricing for risk.  And then on the other hand you have the bond market’s MOVE index.  It did decline in November but very modestly. And as you pointed out in recent days it’s going higher. So, what’s going on? 

To me the divergence is really around the shift and the consensus expectation for Fed policy, but also, the persistent uncertainty around that expectation. So, what do I mean?  Over the past month, markets have clearly gotten more comfortable with where rates are and where the Fed’s about to go. The expectations of potential Fed cuts as soon as March of next year. So, the modal expectation for Fed policy has certainly gotten more constructive for risk assets such as equities and credit. I think that’s part of why those asset classes rallied and volatility has declined. But the uncertainty band right around that expectation is as wide as it has ever been. Yes, we can definitely see a situation where we’re headed for soft landing.  The Fed starts to cut rates if inflation remains low, but there’s also a scenario where inflation starts picking up higher, or growth actually falls even more, right?  The Fed could cut even more severely, so that uncertainty band around Fed policy and the path of policy rates is extremely wide, and that is what is driving interest rate volatility higher and why it has remained extremely elevated throughout this rally. So, to me that divergence is significant, and, in that sense, it tells you there’s still a lot of uncertainty regarding that policy. But the other divergence that I want to highlight, that may be flying under the radar a little bit, is also what we’re seeing in the in the commodities market with oil.  It’s one asset class that notably didn’t participate in the rally in November, as oil prices actually fell, and oil volatility remained very sticky during that period and did not participate in the decline that we saw in equity and credit volatility. And yes, you can potentially attribute part of the decline in oil prices to dissipating geopolitical risk or supply issues. But I would say oil prices near a one-year low signals concerns around the growth picture, the outlook for growth. 

You contrast that again with equities and the rally that we saw in November very much led by cyclical stocks, with small caps outpacing large caps, value outpacing growth, cyclical stocks outpacing the defensive. That’s very much an equity market that is positioning for pickup in economic growth rather than potential downside concerns. So, to me these divergences in asset class volatility and asset class performance really just kind of highlights the still very elevated uncertainty around the macro outlook. 

Steve Sosnick 

There are times I wish we disagreed more because I’m sitting there nodding my head. Obviously, the listeners can’t see that. But my thesis has been — I wrote about this over the last few days — is that the bond market and the short term fixed income expectations are reflecting probably a hard landing since the last Fed meeting. So, I did this from the from the from November 2nd, the day after the Fed meeting, the last FOMC meeting. We’ve pushed up expectations for rate cuts by two months from June to March. We were about 75% in June then, about 75% in March now. And we’ve come from three cuts in 2024 to 5 cuts in 2024. That is not a soft landing. That’s a pretty hard landing.  Stocks seem to be ignoring this prospect. And as you say, when you see, like, the Russell taking off, which are the most economically sensitive stocks of them all, it comes down to are stocks just sort of…  One of the questions I asked basically is has the stock market front run Santa Claus? They’re just so content with what’s happened before that they’re really willing to overlook some of the quote, “the known unknowns”, including — we’re taping this before the payrolls numbers and the FOMC meeting, and I’m looking at something like VIX9D, which obviously you’re familiar with. I’ll let you explain it better as the CBOE representative, but even the nine-day volatility expectations as opposed to the 30 days incorporated in the main VIX product are not really showing much concern ahead of these events. 

Mandy Xu 

Yeah, to a point I think short-dated volatility, whether you look at one month, one-day, nine-day — pick your tenor — those volatilities have clearly fallen. I think they’ve fallen because we’re not gonna know this whole soft landing versus recession or hard landing.  The outcome is not gonna be clear anytime soon, right?  It’s gonna be sometime next year where we’re going to look back and say, OK, yes, this is one way or the other and that’s why actually the part of the volatility curve that it is reflecting more risk. It’s longer-dated tenors, not shorter-dated. So, VIX one-day, nine-day, one-month, all the short-dated tenors have fallen. But if you look at one-year implied volatility in the S&P, that spread relative to say the VIX, it’s near the highs. So that tells you people are still concerned over the longer-term outlook and it’s just not being reflected in a short-dated measure such as the VIX. But one thing I do want to kind of maybe touch upon, you talk about the kind of the rally that we’ve seen in November. I do think part of it is positioning related.  Yes, part of it is fundamental, right?  Like the better-than-expected inflation numbers, the Fed speak, the Fed meetings, they all kind of shifted the fundamental picture and the narrative a bit. 

But also, positioning is a big part and people, investors, I think consensus expectation coming to the fall was so bearish, right. And I think that’s why markets really had a hard time selling off.   In September, October the volatility was not on the downside; it was in the rally in in November that we saw the sharp moves to the upside.  That tells you the pain point that’s really markets going higher and you saw that also in the derivatives market.  Options volumes picked up in November.  On the CPI Day, we had the record single day volume in S&P 4.4 million contracts. It was dominated by the calls, right, particularly longer data calls expiring year end, December expiry. That tells you people were really managing their risks and adjusting their portfolios very dynamically and really concerned about hedging that right tail.  Institutional investors have underperformed all year, not positioned for this rally, a sharp rally going to year end. We’re scrambling for those upside calls, and you can see it also in, for example, like S&P put-call ratios falling to a one year low in November. Also, various measures of skew in S&P and Russell all flattening to kind of to the lows. This tells you really, that the scramble is really for that upside exposure because people were not positioned that way. 

Steve Sosnick 

I think I referred to it as weaponized FOMO.  The thing is, if you’re an individual investor listening in on this, FOMO is really greed, actually, because your fear of missing out is really just the greed aspect of it, or your idea of missing or seeing a move and not participating, that’s regrets.  But if you’re an institutional investor, FOMO is real, and it is fear because you can’t… no institutional investor wants to go into year-end by underperforming his or her benchmark, or by or by being in the lowest decile, quartile, whatever. And so, when that when that rally hit no one wanted to be caught off guard, especially because the change in sentiment was so abrupt. And I think your the statistics you highlight point to that and I think and I love the fact that you pointed out the year-end SPX volume because I think it’s very easy for a lot of people to say, oh,  what, everybody’s playing the SPX , zero-dated options and that’s all they care about, but that’s the fact that the fact that you saw high volumes going into year-end told us that’s institutional investors needing at least some upside coverage and needing it in a hurry. 

Mandy Xu 

Exactly. Yeah. So, I would say like obviously zero-day options have seen the highest amount of growth given that it’s a relatively new product, but oftentimes when people talk about the growth in 0DTE the narrative has a tendency to talk about how it is kind of detracting or taking away from other tenors, and that’s not at all what we’ve seen. We’ve seen kind of the entire pie grow for SPX volumes growing across all tenors.  It’s just that they’re growing faster for zero-days and so that’s something worth highlighting. And then to the point about if you look at what is really driving volume growth outside of zero day options, I really tie it to the macro backdrop.  In a time of elevated uncertainty with regard to the inflation and the growth backdrop, investors are turning more to options to kind of help manage both the volatility of the portfolio but also to really quantify kind of the risk-reward of taking a directional view.  Like, if everything is clear and for sure market is going up, you just buy the underlying, right?  But it’s when you’re unsure, when there’s a lot of risks out there that could derail that view, that’s when you turn to options.  Because going in if you buy a call option, your risk if you’re wrong is just premium paid. So that I think that part of option trading really appeals to institutional investors as well as retail during periods like this where they need, I guess, more clarity and better tools to manage the volatility of their portfolio. 

Steve Sosnick 

Yet at the same time, while that is really the use case for options — and you’ve articulated wonderfully — it’s reflective of greater uncertainty. Yet when you look at other measures, as we go back to the volatility indices, they’re not reflecting a lot of uncertainty. It’s a very strange dilemma, conundrum, oxymoron. I’m not sure what word is the right one here, but I think that’s an interesting part of it. 

Mandy Xu 

Yeah, for sure. I mean I wouldn’t say every part of the volatility curve is low, but certainly at the front end of the curve, as we discussed, the lack of demand or the lack of pricing for downside risk in the near term certainly stands out right now in terms of what is being priced in. 

Steve Sosnick 

And without going down too deep down the 0DTE rabbit hole because I think you and I can have a higher level discussion than that.  It’s interesting. One thing I just want to point out to the listeners is that I focus a lot on the popularity of the SPX index option itself because what I love about that… I was on a panel with your colleague Henry Schwartz when we discussed the same thing, and the question came up, “why is everybody so in love with SPX [index options] all of a sudden?”  I love the fact that all types of options traders and investors have woken up to the benefits of cash settled options.  That you go home, you don’t have to worry about assignment risk, you don’t have to worry about pin risk. You’re just…  Money changes hands and as long as long as you’re well capitalized enough and your brokerage firm has done a good enough job maintaining your margins, that’s it. You just settle up at the end of the day and count your winnings or your losses as need be. Am I putting words in your mouth? 

Mandy Xu 

You’re preaching to the choir for sure. I mean, that’s definitely a huge benefit to trading index options and I think we talk about it.  Obviously, a popular question that we get is, “will we see zero-dated options on other underlyings?”  I think the assignment risk, the settlement risk around physically settled options, that’s gonna be something that people need to figure out in order for it to really broaden out. But specifically, with cash settled zero-day option trading, one of the things I like to emphasize, especially when people talk about concerns around 0DTE trading, is Volmageddon.  I think we’ve gotten past that concern. 

I think at this point people understand 0DTE options and not to realize that is an overblown headline, but one of the things that I do push back on is that it’s unlike what we saw in 2017 going into 2018’s Volmageddon. That period of very low volatility let all these inverse VIX ETF’s and ETN’s really build up leverage.  With zero-day trading, the risk resets every single day. You don’t get that build up of leverage over a period of days, weeks, months or over a course of a year. So, we talked about like Volmageddon and leverage. It’s a very different environment and scenario now versus what we saw in 2017-2018. 

Steve Sosnick 

I do feel bad for the third panelist on that panel, because he was one of the guys who threw out the Volmageddon argument, and Henry and I kind of basically said back to him what you just said.  History rhymes, it doesn’t repeat, and I don’t see it.  My, you’re not going to like this analogy, your colleagues aren’t going to like the analogy, and but I said in the very beginning that the advent of 0DTE options was really that the casinos opened more tables. And I know you don’t want to, you do not have to, do not comment on that because I don’t want to get you in trouble, but basically there were no new games invented. They just opened a lot more $5 blackjack tables, so to speak, instead of inventing some game where no one knew the risks, and no one knew the outcomes.  And I don’t want to go too, too deep down the 0DTE place because you and I could talk about other stuff beyond that and… 

Mandy Xu 

So I would say for people who are using options to speculate, certainly, that analogy  is fitting, but I would say — and this is something that we’ve been trying to hammer home — is that the activity that we’re seeing in 0DTE is actually very diverse and very balanced. And that’s because there are different use cases, a diversity of use cases for zero day options. There are people buying it to speculate. There are people buying these options to hedge either event risk, overnight risk, intraday risk, but also there are a lot of people selling these options for income or for yield.  So, because it’s so balanced, because it’s not just speculators, hedgers, momentum chasers, reversion, people planning for reversion, because it’s so diverse. At the end of the day, when you look at the net risk across different strikes of these zero-day options, it is extremely balanced. And I think that’s why a lot of the headlines are overblown. Because they are assuming that the notional volume, the notional risk is what you should be looking at when it’s in fact the net, and they assume that all the activity is in one direction which would leave market makers obviously exposed when in reality the activity is actually very balanced. 

Steve Sosnick 

Agreed.  I’m going to change gears entirely. We’ll get off this topic and onto one that I found to be an eye opener when you and I talked a few months ago.  You brought up the topic of low volatility readings partially being influenced by correlation or lack thereof. 

Mandy Xu 

Yeah. 

Steve Sosnick 

And one of the things that I hadn’t really appreciated before our discussion.  I knew the mathematics of it. And for those listening and who didn’t hear the first time: simple thing being, if you have an index and half the stocks go up and half the stocks go down when you calculate the index, it didn’t move.  I’m oversimplifying greatly, but and if you have high correlation and they’re all moving in the same direction one way or the other, that increases the volatility. Oversimplifying, but I really wasn’t aware of the various indices that the CBOE had in its suite. COR1M being my favorite because it’s the same tenor as VIX. But you teased at the time that you had some new stuff in the works, and sure enough, like a week or two later you all introduced DSPX. 

Mandy Xu 

Right, yep. 

Steve Sosnick 

And so, what I would like is for you to sort of compare and contrast the various correlation and dispersion indices that you have in your product suite. 

Mandy Xu 

Yeah, sure. So maybe before I can compare the two, let me just quickly go over what is dispersion and what is correlation, right. They are related, but different concepts. So, dispersion at its core kind of measures how stocks are moving relative to each other.  You can think of it as like a measure of idiosyncratic or fundamental risk in the market, and I would say particularly if you are an equity long-short investor or stock picker, dispersion is something that you want to track because it measures how much stock picking alpha there is to be had, right.  In a high dispersion environment where there’s a lot of differentiation between stocks, right, a big spread between best performing versus worst performing stocks, that’s when it’s most attractive to be a stock picker, when there’s the big movements and between winners and losers. Whereas in a low dispersion environment with not much differentiation, well, you might as well be a passive index investor in that scenario, right? Because there’s not much alpha to be generated from stock picking. So, dispersion is an absolute measure of how stocks are moving relative to each other. Correlation, on the other hand, is a relative measure of how stocks are moving, but it’s relative and that it doesn’t really measure the magnitude. So that’s where stocks could all be moving, could be all be selling off today, but it matters how much the individual stocks are moving, right, relative to each other. So, you can definitely have that situation. I think that’s where people can get a little bit confused is that conceptually you think dispersion and correlation are inverse of each other. High correlation must mean low dispersion, vice versa, but actually there are often times where you get high correlation, high dispersion and high VIX.  That’s for example, what we saw in March of 2020.  And the best way to think about is that all of these measures are measures of risk in the market.  When risk is high — and whether it’s macro market risk, whether it’s idiosyncratic risk — it’s gonna impact all of these indices at the same time. So, to me, having dispersion and DSPX index is complementary to the existing indices we have out there.  For example, VIX and COR1M, as you mentioned, really completes the understanding of what is going on in the market.  How much of the risk is coming from individual stocks versus the macro, the market level, how much of it is stocks moving relative to each other versus the index. 

Steve Sosnick 

Yeah. Because one of the reasons I’ve not yet been able to come up with good signals DSPX versus VIX, while I have, while I have been able to come up with some — and those who , those who suffer through reading my stuff every day, have seen this before — is I found that COR1M has given me some good signals. Particularly in September, when the correlations started to return to more normal levels, having been very low, yet VIX was just mired along and then all of a sudden, boom, they caught up to each other and that is what we’re seeing now.  We’ll see if we go two for two on that one. But I have to really come up with a good way of sort of — hate to use the word correlating — but coming up with a signal that that links together DSPX and something a bit more tangible in terms of trading one vis-a-vis the other using a signal DSPX as a signal for VIX or some other volatility type of trading. 

Mandy Xu 

Yeah. So to me it’s like I think of this as a complementary measure to the VIX because it can tell you really what is driving the VIX right, and in the piece that I published on the SPX the example that I used was in 2018 where we had two big VIX spikes that year, February 2018 and then Q4 of 2018, right? In both cases, VIX spiked over 30.  Similar magnitude of moves, but what was driving the VIX was very different, right?  In February of 2018, DSPX actually fell to 0.  So, at the time if you had DSPX on that day when VIX blew up and the VIX ETN’s, that blew up, it would have told you it was a technical index led move that had no bearing on underlying fundamentals.  Single stock options didn’t see a similar bid right. And volatility that we saw was at the index level, it was very little fundamental risk that was being increased in the market. Whereas in Q4 of 2018 it was dispersion.  Actually, DSPX actually started rising before the VIX and it was rising alongside the VIX. That was a period where the VIX was going higher because of increasing fundamental uncertainty with the market.  I think at the time it was Trump trade wars, there was Powell monetary policy uncertainty — he was a lot more hawkish than people expected. There was angst around earnings Q4 earnings that could be particularly bad that year, or a lot of uncertainty around that. So that was a very much a fundamentally driven market selloff. 

Whereas February 2018 was kind of a technical index-led move and having DSPX alongside the VIX would have allowed you to discern between the two much better.  And one of the things I think that’s interesting recently is that VIX has fallen a lot and DSPX, relatively speaking, has been more elevated. Now it fell a lot in in November, partly due to seasonality issues. But overall, if you look on the average over the past couple of months, DSPX has been a lot more elevated than the VIX.  And this has kind of been true for the past 18 months and it’s really a reflection of the fact that a lot of the macro risks that we’re talking about are having disproportionate impact at the sector and the sector and then the stock level, rather than at the market level, right?  So, when we talk about higher rates, that has kind of resulted in in the equity market having huge swings over the past two years in growth versus value, cyclical versus defensive stocks, old economy versus tech. So, we’re seeing big swings at the sector and the single stock level, but because of low correlation like we discussed last time, it hasn’t really translated into higher levels of in depth. 

Steve Sosnick 

Yeah, I think that’s also interesting to unpack.  We’ve all heard the 7 versus 493 argument, but it is a lot richer than that. And do you plan, without revealing any proprietary secrets, on expanding the suite of dispersion indices over the coming over the coming times?  Is that something you’re looking at doing based on what you’ve seen, the adoption of DSPX so far? 

Mandy Xu 

Yeah, that’s something we’re definitely gonna be building upon. So right now, all that is out is the DSPX index, which is not tradable. It’s meant to be an indicator very much like the VIX. But our plan sometime next year is to launch futures on the index very much like VIX futures on the underlying VIX index. That will allow investors to actually express a view on future implied dispersion.  In case it wasn’t clear to anyone listening, DSPX like the VIX, is a measure of forward-looking dispersion, implied dispersion. How much stocks are expected to move relative to each other, and then over the next 30 days. So, once we launch the futures, we will actually be able to trade that level similarly to how people trade VIX futures. That’s something we’re very excited about. So, stay tuned for that. 

Steve Sosnick 

Because I think that would be a very interesting type of trading where you could trade DSPX futures versus VIX futures and see how those curves set up, because, as you alluded to earlier, longer term volatility is much higher than shorter term volatility. There’s a certain…  the I find the VIX curve right now to be very steep. Some of it is strictly positive contango. 

Mandy Xu 

Yeah. 

Steve Sosnick 

Let me unpack that for those who are not futures traders.  As you look at a curve, it slopes up left to right. Some of that is interest rate effects, while at some level it reflects the relative plentifulness versus scarcity of the underlying commodity. In this case, one of the things it tells me when you see a very steep VIX curve is that the commodity in question is available volatility protection, and as of now it is quite plentiful and easy to come by. When you see the VIX curve invert — think about when that happens, it’s usually in times of stress — that tells you that there is scarcity.  Sorry, that was not for you Mandy. That was for all the others.  But I think it’s important to keep those frames in mind and I think it’ll be fascinating to see how a DSPX curve sets up with what is a normal contango for that curve.  How will that be? 

Mandy Xu 

No, absolutely, you touched upon something that we’re very excited about. One of the things that make us excited about the launch of DSPX futures is that we think they will trade with a little bit of different curve and relationship versus the VIX.  Because as we all know, VIX typically is upward sloping, right, that the futures curve is in contango as you just explained. But think about DSPX. Forward-looking dispersion is a big part of ordinary stock dispersion. Actually, the most reliable catalyst is earnings, right?  So, as we go through earnings season, you would expect to see more backwardation in the DSPX futures curve than you do for VIX. And what does that mean for carrying a long dispersion position versus a long vol or VIX position?  It should be when it’s in backwardation– obviously the carry cost is a lot lower. So, there are definitely gonna be a lot of nuances between DSPX futures trading versus VIX trading, and something we’re very excited about for next year. 

Steve Sosnick 

Well, I think we’ve just set up a time frame for our next discussion if we don’t do it sooner, because I can’t wait to have it.  I’m getting something on my screen that we’ve blown through the time that we’ve had allotted to us, which I think is great because it means you and I have just been chatting away, and as I’ve said I enjoy our conversations immensely. So, I think unless you have anything else to add, Mandy, I think we’re going to wrap it up here. And again, I want to thank everybody for listening. My guest today has been Mandy Xu, head of derivatives market intelligence at Cboe. I am Steve Sosnick, chief strategist here at Interactive Brokers.  You can find all our podcasts at ibkrpodcasts.com and on your favorite podcast source – all the usual suspects — and we hope to see you soon. And Mandy, what a pleasure talking to you as always. 

Mandy Xu 

Thank you, Steve. 

Steve Sosnick 

Take care everybody. Bye bye. 

(Link to previous podcast with Mandy Xu: Might Correlation be the Key to Understanding VIX?)