Dan and Matt take VS Pro questions on the Sep 18 expiration book: charm flow against positive gamma, the overnight range, how a 5,000-lot order gets worked, and the view into October.
Dan explains that selectable expirations in the VS3D gradient charts are too much compute to serve to everyone, so he built the expiration webpage as the guide instead.
He steps through the morning 0DTE position, the 7610 test level and the 7640 to 7645 pin zone, and argues the most useful profile excludes 0DTE volume so the overnight position stays as it was set.
He shows how to advance the expiration page hour by hour to find where charm crosses zero, lands on roughly 7680 into Friday, and notes market makers and market makers plus firms both have delta to buy.
Drawing on his market making years, he describes orders that moved nothing and hedges that got run over, and points out that much of the observed index volume is arbitrage locking rather than directional flow.
For large size he prefers crossing the market or taking a committed price from a bank over working the order and signaling direction.
The dashed series is a fixed 2:50 PM reference against the updating line, and while a few firms can take down a 10,000 lot without hedging, charm applies a persistent pressure inside a range that recycles through gamma hedging.
He answers whether charm and positive gamma cancel by describing the position as a risk reversal whose gamma collapses onto the strikes into expiration, and notes gamma scales with implied vol so cheap upside calls carry more of it.
He plays the trapdoor timeline and explains the watch condition as a front two week VIX option shape that creates vega to buy as time passes, independent of where implied vol is now.
He publishes the expiration charm view for the bigger quarterly and sub-quarterly cycles rather than every day, and recommends Natenberg, then Colin Bennett, then Taleb.
Selecting the September 18 AM series in VS3D, he reads a decay up shape toward 7680 with a possible stall near 7650 and says he will refresh it after the close.
He applies the exposure rule to the AM and PM charm curves to locate buying below and selling above the cross, and states his position as a 7660 to 7680 call spread for Friday with a bearish bias beyond.
He describes the option distribution in three parts, notes that charm and vanna vanish at a 50 delta option until the settlement moment, and says hedging style varies by desk and reinforces the regime each desk is set up for.
He stays bearish through October, plans far out of the money October puts with wing convexity, and reviews why the September 23 put spread was the wrong expression for his own flush and bounce hypothesis.
Reviewing the day from the 7610 test back toward 7645 and a 7637 close, he tells the 1220 to 1300 put fly story to explain why he does not carry an expiration position into settlement.
What was said, in order. Each line opens the player at the moment it is said.
The profile with the most traction leaves today's 0DTE volume out, so the 0DTE position stays as it was set rather than moving with intraday flow.
3:28 · openInside the morning framework only 7640 to 7645 had pin potential, which made it the most likely place to land.
4:30 · openWhere charm flips from supportive to suppressive is where consolidation is implied, and that is how the overnight range gets framed.
7:40 · openWhen market makers and market makers plus firm accounts show the same picture into an expiration, the read is cleaner because both have the same delta to trade.
8:27 · openHow an order is worked matters more than its size, because signaling direction on the complex order book moves the price against you more than crossing does.
15:51 · openFor a 5,000 lot and up, hit the market or take a committed price from a bank rather than staging it on an observable frequency.
17:10 · openGamma scales with implied vol, so upside calls at lower vol carry more gamma than downside puts of the same vega, and a flat vol spread is not flat gamma.
24:58 · openNegative exposure gives a positive hedge contract to trade, so negative charm exposure marks the area where buying pressure should show up.
37:02 · openDan is long the 7660 to 7680 call spread for Friday and keeps a bearish bias for the rest of the month and through October.
39:36 · openTyped in the Discord stage and the Zoom chat during the session. Each answered question links to where the answer starts.
Has today's px action changed ur view for next week and into month end?
Bearish for the rest of the month and into October, and he would no longer bet on 7775 the way he did earlier in the week. answer at 39:36
Can you make the expiration selection for the gradient charts available? Would like to see Greeks sans today’s expo
Selectable expirations in the gradient charts are too much compute to serve, so he built the expiration webpage, and a view without today's expiry is coming. answer at 0:28
is charm's delta impact much larger for quarterly opex than regular monthly opex and weekly opex?
Yes, the quarterly and sub-quarterly positions are bigger, which is why he publishes this view for them and leaves ordinary days to the VS3D gradient. answer at 31:57
Does $500 million really move the index? +$0.6bn/day seems like it wouldn't move the index?
Much of the observed volume is arbitrage and locking trades, so the directional imbalance works against a much smaller book and execution style moves price more than size. answer at 11:26
What's the dotted line Vs. the full one?
The dashed series is a fixed 2:50 PM reference and the solid line is the current shape, so the two show how it is changing. answer at 18:16
is there a way to see this visual you are using with the moving tab on the website? i've tried looking at the volsignals.com site but i can't find it
It is live at the link he posted, and he does not want it circulated. answer at 12:38
So does charm flow completely overpower positive gamma? If the gamma is positive while charm flow is dealer buying, wouldn't they kinda cancel each other out?
They do not cancel, because the position is a risk reversal whose gamma shrinks onto the strikes into expiration while the charm pressure continues. answer at 22:57
If 6k contracts flow got crushed, wouldn't that mean the same for the charm flow? How does it apply directional pressure that is visible?
Charm is a constant bleed inside a range rather than one block, so it keeps applying pressure and the move gets recycled through gamma hedging. answer at 20:08
so whats correct way to execute larger orders 5,000+ lots?
Cross the market or get a committed price from a bank instead of working it and signaling direction. answer at 17:10
What is the “Set Aside - 60” column at the bottom?
Not covered on the recording.
So there are different channels of liquidity? So different ways of making a trade change the impact on the price? Is the way that charm is a constant bleed mean that the impact is strogner than a huge block trade?
Yes, a firm with the balance sheet can absorb a 10,000 lot with nothing moving, while charm's persistent pressure inside a range shows up more than a single block. answer at 19:28
Guys, Sep aside, anything interesting into Oct opex?
He plans far out of the money October puts and put spreads with wing convexity, plus higher strike puts expiring next week. answer at 43:21
Can you talk about vanna trapdoor? What is quiet watch pressure?
The trapdoor combines a proprietary position condition with three market measures, and when it triggers it flags about three weeks of elevated risk. answer at 27:41
This is based off yesterdays position correct?
No, the view is live as of 2:50 PM, though the page also lets you compare the prior close and the day's trades. answer at 21:21
is charm's delta impact much larger for quarterly opex than regular monthly opex and weekly opex?
Yes, the quarterly positions are bigger, which is why he renders the expiration charm view for them. answer at 31:57
After expiration, charm will reset, the data (plot) still be available with coming positions?
He will run it again when it is meaningful, for quarterly and sub-quarterly cycles rather than every expiration. answer at 31:57
Dan, could you recommend a book for a newbie?
Natenberg first, then Colin Bennett's Trading Volatility, then Taleb's Dynamic Hedging. answer at 32:12
Dan - are you able to speak on how the data in VS3D as well as volsignals.com/fomc-sep16 informs the overnight trading range?
Select the September 18 AM series, which decays up toward 7680, and re-read it after the close for the overnight. answer at 34:02
so the positive gamma gets more and more localized/narrow to the strike? So charm flow is effective until we increase right to that precise strike?
Correct, gamma narrows onto the strike into expiration, so charm can keep pushing until price is at the strike. answer at 39:49
I understand that charm has time compnonent where dealers have to sell/buy a caertain amount every 5mins. With gamme, I understand it is price dependent so every dollar movement leads to dealer repositioning but is that done immediately after each point movement or is it clumped up together after certain time period/ price moved?
It varies by desk: a mean reverting market gets small frequent clips, a trending market gets wider tolerance, and nobody hedges linearly per dollar anymore. answer at 41:12
Dan (obviously contingent on where we close tomorrow) do you still hold the bearish bias for rest of month?
Yes, still bearish for the rest of the month and through October. answer at 39:36
do u think we will get any opportunities for overnight slight dump of 20 to 30 points to load up call spreads for tomorrow?
He would not bank on an overnight dip to get into call spreads. answer at 49:35
@Tarheels5 term structure for your calendar is not currently favorable?
Not covered on the recording.
Is the 8-21 DTE negative charm the primary reason you see the initiation of a pullback in that timeframe?
Not covered on the recording.
Pick a concept to keep only the transcript paragraphs that cover it.
Cleaned from the Zoom recording. Timestamps open the player.
There's a Zoom link. What I'm going to do here is share my screen, which is my view of Zoom. So you can see me on camera if I do something, or on the screen share.
So, gamma drift, it could be one of two things. I was looking at the charm rendering from 12, maybe. The most recent update I had... maybe I was in RTM when I was sharing that. But the most recent update is why I put together the webpage. So use the webpage as your guide. We don't have gradient charts selectable, it's too much compute. There are 5 factorial constituent types, and we have a tremendous number of expirations. We also group some expirations, because there are certain types of contracts that are not traditional SPX contracts that we embed in the profile, since they actually have more traction on hedging than the listed stuff.
It's just way too much to figure out how to do a deselection or a custom gradient. It would be wonderful, but I would basically have to do that custom on my own. It's too hard to serve to everybody independently.
Okay, cool, so you have the Zoom, you can join us here and ask questions in the chat or in the Q&A in the Zoom. The main point I want to articulate is that this is very much mechanical. You see we keep pressing higher, but we have gotten stuck a little bit. It's strange, when I have these positions that are set up in a certain way, like in the morning meeting, for example... let me share my VS3D screen really quick.
Okay, so if you see here, I know it's going to be a little difficult to follow, and I can't really zoom in too much. But you should have the same data, generally. I'm going to go into Market Maker Position, and go back to what it looked like in the morning.
For today's 0DTE. Today's 0DTE is interesting. We have a position that had a strange charm-up feature with not a lot of containment in the middle. I'm always uncomfortable when the range is not symmetrical at all. So we had a test at 7610, and then at some point a little bit of a boundary at 7650, which I'll go through really quick.
This is today's position, not September, not the big one, we'll talk about that next. But look at the blue dots.
We're trying to get a profile out that anchors these. It's difficult, because we actually need to do a little bit of that deselection. The point is, the profile that I think has the most traction, the one I would pay the most premium price for, is to take everything as it was.
The easiest way to say this is: just don't include 0DTE trades. Just literally don't include 0DTE volume. So once the 0DTE position is there, it stays there, it doesn't move.
Which I know seems strange, but the point is that it's touched by different groups. They don't trade it the same way. A September contract even wouldn't... there are groups that won't trade tomorrow, they'll only trade today's inventory turnover. That's the whole point with the blue dots. We're working on doing that so it's just done for you in the product.
It's hard to see because of all the trading today, it's crazy. But if you see where the blue dots are, there's this test level at 7610.
And that's the range boundary. The next range boundary of any consequence is really 7650. The problem is this whole thing was decay up, not only because of this, but also because of September, and then the only real balance point that makes any sense... it can't really be 7615, because it's charm up in the background.
7640 to 7645 are the only locations inside the morning framework that have any pin potential. So when I look at this, I'm thinking that's the most likely place to land today. But even when we were selling off to 7610, if you were to ask me where I placed my bets... I got a little excited, I used that GIF from the chat.
It was mostly this: I don't think we sell off. I don't think we hold. Even if we flirt with it, and we go back and test it, the whole point is you can trade this point-to-point.
We went and touched 7611, and maybe 7613. Both times we bounced. It's really hard to fight opex charm. So that's why I put this together for you too. You've noticed I'm not just sharing on X. I want you to get better insight than what I put everywhere else.
This is stuff you can't get anywhere. This is stuff fund managers are asking me for. You need to visualize and think through how the position decays over time. Sorry, one second. Give me one moment.
Okay, you good? Sorry. Apologies. Yeah, I appreciate it. I'm glad this is good for you. It drove me nuts. I didn't get to trade it as well as I wanted to, and I actually have the thought that I almost want to see if somebody has been having a really good time with my insight, if they've been nailing the trading, and just ask them... maybe we should set up a trading desk. But that's in my head. We'll get there.
What I was trying to articulate is that this is really important to understand. The charm stuff is hard to articulate because of how it changes over time. We have our 0DTE charm, and we have options expiring next week that contribute to the charm gradient. This one is the one you want to surface, kind of like what you're asking: you want to see the greeks without today's expiry?
We're going to have that as well. That's an easy, standard one. But the whole point of this one is to see how this expiration drives things.
So I built it so you could do it this way. You can see gamma, charm, vanna, delta, and vega, any of them. Obviously you should understand, I would never, even in my wildest dreams, think about the greek called vega when I'm talking about one-day options. That's not really thinking.
It's really about just charm and gamma at this point, and maybe there's edge with delta. If you've looked at delta and found something there, in some sense it's just a representation of charm.
But you know I don't think about this one too much. It's not to say there's no edge, it's just not the way my brain works. I think in terms of dynamic stuff. The point here is, if you're looking for a pin, for gamma drift, use this. I'll update it after the close as well, so that we have accurate numbers.
That's for the overnight range. The whole point is there's going to be a place where a consolidation is implied. Charm will inflect, meaning when you cross the place where charm flips from bullish to bearish, that's where you expect things to settle down. This is the last one we had. Again, I'll update it at the close, but if you look right now, what it's saying is Thursday at 2:50 PM ET, so this is contemporaneous.
It's bullish right now. It's actually getting more bullish. So this is the maximum bullish part of the charm profile at the moment. If you look at where we're trading right now, we're churning around 7645, 7650. There's still this dynamic with the 0DTE position. There's been a lot of intraday flow. Maybe that has mattered enough today, I don't know.
It's a balancing act. But what I look at most importantly is that this has to happen. We know there's some kind of bullish behavior coming. I also staged it because sometimes firms have big positions in the expiring series, so if, for example, market makers and market makers plus firm accounts look totally different, that might make you feel like there isn't as much edge, because you can't know for sure who's going to do what, or how they're going to hedge it. Maybe into expiration they're both going to hedge it, because it's important, it's expiring. In this case, though, it aligns. So whether you're looking at market makers or market makers plus firms, they both have delta to buy into tomorrow morning.
Where does it resolve? 7695 right now, but you're going to see gamma drift, to your point, as you do this. When I was doing the content I was using this more. But it has to be taken with a grain of salt, because as time passes this changes, and that's really why I built it for you.
So, scroll through? I would imagine, and this is just my guess, I'm going to go back to market makers, because that's the VS3D angle. I guess I should do this first just to see. 7713, interesting.
But again, that's right now. The whole point is that right now is not necessarily the way to think about this. You have to visualize it through time. So when I'm doing the morning meeting framework, for example, the morning framework, I'm doing this in my head. I should probably just do it like this. Let's see. Going through... this is only going hour by hour, but you're going to wind up seeing this should be around 7680.
I would think it's going to cross here. Let's see if that's true. And you can do this on your own.
Very close. That might actually just be it, I don't know. So we're going to keep going, keep going, and you can see here on the screen, sorry, I'm pointing at my own screen... exit here. This is the timestamp against which we're benchmarking, so we're going to get one more, and after that it will be gone.
Just based on the way it was done. I'll do it more granularly, but to be clear, at the end, the reason I talk so much about how charm works is that in those last moments your guess is as good as mine for what the delta should be at the straddle.
Or on the different options. It's mostly about understanding how inventory works. You're getting boundaries. You have a boundary at 7675, where if you go above it, charm flips aggressively. It's local pinning at the end. So part of it is imagining my view about where a likely pin would be.
That would probably be a little different if, for example, we overshot to something like 7670, which might have been my expectation when I wrote that. That's all.
Part of what I've observed, and this is from my time market making too, is that I would have moments... I remember I once traded 6,000 futures at once for a cash spread, a big block trade, where I had been waiting for this guy forever, and we needed it. It was December of 2018. A big put spread roll, 18,000 Dec puts against something else, and we really needed to sell skew.
And I'll never forget, we sold off about $6, and then it disappeared. It came right back. I moved the index a small blip in the sea of things. That flow was then just crushed. Hold on a sec. Hold on.
And just scroll down. So what I was saying is, when I think back to my career, there were so many times where a big order was just swallowed. It's almost like somebody was waiting for it, or knew that it had nothing to do with any kind of program.
But then there are also so many times where I had to buy 200 minis, and I tried to get cute with them and stage them over icebergs, or do them every so often.
And especially in a lower liquidity environment, somebody would just move it against me. I couldn't buy delta. If I sat my minis on the bid, we'd just move up. If I bought on any observable frequency, if it wasn't completely randomized, it would just be moved against me.
And then as soon as I stopped having a need, it just normalized. It drove me nuts. So I think there's some of that going on. When you see drift, people don't realize how much of the trading volume we actually observe is just games being played, or arbitrage conditions being held. For example, when SPY diverges from ES, somebody is going to snap trade it together.
That produces volume, that produces these big volume numbers in the index, but so much of it is just maintaining the glue. And you have to realize that the net imbalance effect is against a much smaller book, I think. And there's something about observing a program.
I don't want to speculate about what I don't know, but it seems like what makes the most sense is that...
Yeah. If there's a one-and-done splash, someone must interpret that as somebody being liquidated. But if there's an iceberg order... I've seen this before in vol, too, and I thought it was so stupid. Take the IB Whale, for example.
He would just execute his orders on a PAR station. It would literally be a 10,000 lot put spread, and it would get routed electronically the same way yours or mine do to Interactive Brokers. But it would get announced by this guy, KSI, on the floor. A broker would just bark it out. And it would be with 50 lots, 300 lots, 10,000 lots.
The guy wouldn't go to Goldman as a prime broker, for whatever reason, because he thought they would take advantage of it, or that he wouldn't get a good price. And I would see this all the time, where instead he'd break up his order and put it on the complex order book and do it in small stages, and never get filled. You have an order, a tied-up risk reversal, and if you quoted at the beginning for 10,000, the whole pit would take it down within $1, maybe 50 cents of that starting price. Instead, he'd put it on the complex order book and signal his direction, and that's the big part, signaling his direction, and then all of a sudden.
It can't get filled. If you bid one half on 500 and it trades, and you bid one half on 500 behind, the next time it trades is $1.70. And if you step it up to $1.70, it's probably not going to trade. And then I've seen it go to $5. So that was a really hard problem to solve in market making, where the way an order was executed meant more than its volume, if that makes sense. The impact didn't align with the volume by itself. You get my point. It's hard, it's crazy.
That was always a tough thing, because you had to hold in mind how much really traded. So we'd have these conversations all the time where we'd say, that was only a 1500 lot that traded. One of the biggest things we've tried to do is actually measure and model how much of a parameter traded in order to move the parameter, if that makes sense. If we're going to have our 2-month skew up
on a 75th percentile fit, how much actually traded in that direction to move it? Because what would happen is you'd get these iceberg orders that would put it in that direction. You might move your skew $2 when the actual amount that traded in the market would only justify a historical fit of 75 cents. So you need to be aware that it's time to revert. It's hard.
For large orders, to me it always made sense to just crash the market, just hit it. You're better off doing a market order sometimes.
Or go to a prime, where they're going to stop you out. The way it works mostly, and you might know this, is you call around and you get competitive bids. BNP might make you a market, Goldman Sachs, Bank of America, Nomura, they're all going to give you a price, and then one is going to commit to a better price than the other. Then that one has to go.
That one would have to go to the floor and put it up. That's why Goldman, sorry, not Goldman, JP Morgan, moved to SME. Makes sense. So... Super Friday.
You have 6K contracts. Yeah, it is weird sometimes, the dotted line versus the full one. So the question on what's on my screen right here is, what's the dotted line versus the full one?
If you see here, dashed. I don't like the way Claude phrases this and I have to correct it, but the series at 2:50 PM is a reference point to show you how it's changing, is my guess.
But I have to be a little more precise, maybe, because you'll notice the cross never changes. Even though the shape changes, the cross never changes. It must be a fixed reference, and then it joins.
Are there different ways of making a trade change? There certainly are. Jane Street sometimes would take down orders like these 10,000 lot put spreads, where it was bizarre to watch it happen, and nothing would move.
Not an input, not a mini, not anything. They just had the room. That's a huge balance sheet, a huge ability to tolerate that and to do whatever they needed to do with it to make money in their context, whatever that might be. All jokes aside about what they do. But not every firm is like that. That's the exception, not the rule. A lot of people who dismiss what we do claim that's the rule. It's not the rule.
Reference the previous line, yeah, I think that's what it is. My perception is that the way charm is a constant bleed means the impact is stronger. It's also a persistent pressure, and so the whole morning framework idea is that if you're observing a range.
And the range is defined by a series of participants buying and selling, and you know one of them is relatively large, again, because it's directional. When you see this volume, you don't actually know how much of it is directionally biased. So much of it is just locking things together.
So it's difficult to get the right denominator to measure the impact, for one, but also remember it's a persistent pressure inside a range. It's a continuous counterbalance. So if you have positive charm, or sorry, supportive charm,
you're also going to have upside gamma. So there's buying that's biased, and if the market moves with it, that move runs into selling from gamma hedging, and then it gets bought back later. So it's a channeling mechanism. We would grind along and recycle the hedge in a way that was tough for us as market makers to avoid. We try not to do that, but that's really how it works.
No, this is based on right now. Well, not right now, but 2:50 PM. So the question is, is this based on yesterday's position? There's some material here that lets you compare yesterday's position.
See the close, and the day's trades, that's about the FOMC day. If you want to explore how things change in the position, feel free, have at it. Again, this is not final form. I'd like to make it more quickly discernible, but the point right now is just to get good information to you guys who want to go down rabbit holes.
This is live as of 2:50 PM. So when I updated this, that was what it was at the time. And I'll do it again into the close, so that we can understand what to expect in the overnight, because that will be important.
I'm going to take some questions quickly from the Zoom here. We have a couple things going on. First question is, Matt, could you mute? I think that's already taken care of. Does charm flow completely overpower... this is a hard one. Oh, there's Matt again. Is that Matt? And he's gone again. So does charm...
Is the morning meeting lost? I thought it was found. If it's lost, I'm sorry. It was because it was started by a different participant.
RTM in the background? Turn that down. You guys hear Steve Irwin in the background? That's our team.
Good questions in the VS Pro Zoom. Does charm flow completely overpower positive gamma? If the gamma is positive while charm flow is dealer hedging, wouldn't they cancel each other out? Yes, but it's that feature where you just have to imagine the position as a risk reversal. A dealer short put and a dealer long call, and then think about how that changes at a certain time, at expiration.
The position is going to have virtually no greeks in the middle anymore, because the option influence will be shrinking around just the strikes. So if you're $30 away at the beginning of the day on a 0DTE contract, it's gamma neutral in the middle.
And maybe even around the strikes, if it's a clean offset. Because on a big straddle price, both options will be like straddles, and so their gamma will be distributed widely around the strike, and they'll offset each other, so the gradient won't reflect much. But as you go to expiration, it's going to be such that you have features more like what we see on the gamma profile right now.
Where it's very distinct. This is a night and day difference. So if you look here, for example, I'm going to just use a random one. But that's a big difference. 12 billion versus minus 3 billion, only 20 points away, and that's a function of the influence of the option getting smaller and smaller around its strike price. It has no delta. It has to be near the money to actually have that gamma.
So if you've been listening lately to a lot of what I've been saying, I've been drilling this point more and more, that the type of position that we have is a big deal. Because, for example, if we have a position that is very much like these risk reversals, where it's across the board in tight ranges, it's short the downside and long the upside.
That, in and of itself, at the money, even if it's flat options or flat vega, will create positive gamma. Why? Because options that are cheap, like a lower volatility, which is usually the case in the SPX if they're upside call options, have additional gamma. Gamma scales with IV.
So then the options that are to the downside, even if they're vega neutral as a position, they have less gamma. And so for the market maker, it looks like you have positive gamma when you're short one option versus long one option, but they're a good vol spread. That's bullshit.
It doesn't work that way. You get penalized all the time if you pretend that, because what tends to happen is, obviously, the vol goes up on a down move, and down on an up move, and you really don't get to observe that, but that's difficult to model. No competitor, I think, that does this will ever produce a gradient like that. We probably will, it's just a matter of, do we want to put it in a low-tier tool to get that.
You don't need to know exactly what the real gamma is to understand the point, though, that when you see red and green on the same profile in big form, there's a reason we can just keep chewing through in one direction, because the gamma's not really real until you get to the straddle. So charm allows for that path to graduate up.
That's why I often say, the most stable markets, when you really don't want to move too much, or don't want to make trades based on that assumption, you don't want to have a lot of vanna in the profile. You don't want to have a lot of gamma, or sorry, a lot of speed in the profile, because the two of those things are shadow gamma.
That's in-depth, obviously, and you might need to chew through a little bit of that, because I see somebody else had the same question for a long time. It is difficult, but it's the point that the gamma is actually decaying alongside, getting smaller and smaller, and more cycles of the strike, and that's just pinning.
Oh, vanna trapdoor, yeah. So this is proprietary. I'm not going to share exactly what it is, but the idea is basic. This question... can you talk about vanna trapdoor, which is quiet watch pressure? So, the vanna trapdoor, VS Trapdoor, here we go, let's see, this is beautiful. Is this updated to today? Good job, Claude.
Alright, let's play this. So what I love about this is that it is a representation of when I think the position is most risky. I spent a lot of time trying to figure this out, because I would see something, and I couldn't exactly figure out what it was. It turned out it was actually a little bit closer to near data than I thought it was.
But the point is, when you look at this... hold on a sec. Okay, so what you're going to see here, as you play this, there are conditions. When a condition triggers, it casts a shade across the next 3 weeks.
When our condition is present, it is considered as trapdoor. You'll see it say trapdoor here. The point is that when our proprietary condition and the position is present, the other three are not proprietary at all, they're just market measures. I think it's momentum fading, and it's a little bit of pickup in both implied vol and the rate of change in implied vol.
But when our position condition is present, it's a real watch, like a flag. So I'll play this quickly. So here you have the SPX moving. So here's a moment where ours gets triggered, I think. You can go through this.
Quiet, quiet, quiet, and it's these moments where vol is starting to rise from low levels into a high vanna book of a specific kind, normalized a specific way. Here, it's live, obviously, but this is much riskier. When you see all the conditions present, that's pretty dangerous. Here, pretty dangerous. Usually these last about 3 weeks. Interesting.
Oftentimes, when any of them are present, we start to slow down. It's a moment where actually buying risk reversals, buying puts, selling calls might be a good trade. There's only a few times where we really push higher. Obviously, you have to be very aware of the slingshot out, because the same force that's involved in our mechanism, vanna, is active the way back up. The bounces are severe.
The watch condition, it has to do with the front 2 weeks VIX option position. You probably know what I'm talking about. When you see that shape, it turns out that shape is important, and I think I've been able to drill down into why that shape is important. When you see the VIX in a situation where it's above a floor, so it's above a dealer short, and near a dealer long to the upside.
In the first 2 weeks combined, the reason that's important is because that creates an additional vega-to-buy condition for the market maker community holding the VIX position.
And the problem with that is that that's completely independent of what implied vol is doing right now. And so most people see VRP and they think, well, we're not moving too much, 30-day out options, and they have this naive assumption that that's a good proxy. But the reality is, we don't know what the next 30 days will be like. And so to benchmark them to right now is not really intelligent in any way. And that's vexing, then, when you see this condition where the market has a default mode.
And maybe there's a rolldown most of the time, but guess what? Maybe there's a rolldown most of the time, but suddenly you have another construction in the market where there's actually vega to buy just as time passes.
And that's absorbing the rolldown, and suddenly it holds a level that's a bit higher than the market feels like is sensible. That's when, if you ever hear about the jaws opening, the jaws start to open, and usually they collapse eventually. So that's what our watch condition is, it has to do with that.
After expiration, charm will reset, the data plots will be available with coming positions. I'm not sure yet. I'll probably do it when it's meaningful, but in theory, the reason I'm doing this for this expiration, I'll do it for sub-quarterly as well.
It is because these are bigger. These are bigger. Most of the time, you can get away with the VS3D gradient, and that's good. On a day-by-day basis, it's excellent. The times when this is relevant are not all the time. It's just the way it is. Book, I would use Natenberg. Natenberg's a good book. When people ask for books for learning this, I say Natenberg, Option Theory and Pricing, or whatever it was called. Option Pricing and Volatility.
For beginners, and then in the middle, I really, really like this book by Colin Bennett, which is free if you just search for it on Google. Colin Bennett's Trading Volatility. He was in a position like Matt was in. So you're managing a vol book at a bank. So it's very pragmatic. If you learn the basics first, just the actual mechanics.
Some of the charts on where the greeks are, positive, negative, and all that good stuff, and how the spreads work.
Once you start to want to think more dynamically about how things work in real life and what people are doing in the market, the Trading Volatility book is excellent. It's one of the most pragmatic, practical, simple-to-read guides I've ever seen. And then if you want to have nightmares about what it's like to be a market maker, I recommend most of Nassim Taleb's work. Anything that he's written is probably good enough for that.
Natenberg first. And then Trading Volatility by Colin Bennett second. And then Taleb's Dynamic Hedging of Vanilla Options, I think it's called. Or just Dynamic Hedging.
So, with respect to how the data in VS3D works, as well as the vols... yeah, so it's all the same data, obviously. But you want to look to tomorrow's session, and really tomorrow's session is still pretty obvious to me, it's this charm up feature.
So what I have to do for that, remember, this position will be expired by then, so you're going to want to get rid of this. You're going to go to Custom.
You go to September 18th AM. AM is way more important in the session than PM. PM will have some background relevance, but nothing like this. This is pretty significant. Look at this shape. We are decaying up.
There's going to be a local collapsing around 7650, wherein eventually we could see a situation... sorry, I'm trying to make this bigger and make it smaller.
We could see a situation where we get a little bit of reversal off of 7650, maybe somebody takes profit, but by and large, when I look at this, I think it's crystal clear. This is broadly decay up, and it's decay up to 7680. If we overshoot, we could shoot up to 7730 and then ping-pong back down, I don't know. That would be a huge move, to be honest. Another 80 points.
But I do think that we sometimes get big moves in the overnight. I can't even tell you how many times I've doubted it, and then yeah, $50 or something is not out of the question.
We closed 7645, it's going to drive me nuts. Sorry. But I hope you see why it makes sense. The morning framework is so good. Sometimes I think I just want to quit this and trade the morning framework alone and perfect it, because I have never found anything that works so well.
And did you see me step through this? This is the same data, not this one specifically. But here...
Sorry for the screen jumping around, I know it's probably hard on your eyes, because you're not expecting it.
But look at the way this converges. I'll update this again into the close, but you should expect it to align with the position I just showed you. So if that position changes into the close, you can see that yourself. Remember, it's the AM expiration, so the AM position cannot trade after 4 PM.
There was some kind of rule change where they were going to make that different, where it could be traded in the overnight session. That's not the case yet. For some reason that didn't go through, I don't know. But the way this works is, this is gamma, and you're going to want to look at charm. The whole point with pinning is that below this level there's buying, and above this level there's selling. That, combined with gamma, is what gives you a pin in the end. Now, when you look at this curve here, think about what the gradient looks like.
Remember, it's the exposure rule. Negative exposure gives you a positive hedge contract to trade.
And so when you have negative charm exposure, this is bullish, the orange, this whole area. I should just call it orange, I guess. This is where you should expect buying pressure. And then the flip, where you would see blue above orange, would be where the line intersects zero. So if you keep going.
And this is just moving one hour out. Don't obsess over the way the shape changes. This straddle is not an easy one to model as far as how it's going to decay. But generally speaking, in this area you just want to watch where the cross is. The cross should eventually be right on that biggest long position into Friday at... well, okay, there we go, 8:20.
That should be 7680, like we just showed you in VS3D. And then 8:50, 9:20... Now it's virtually... ignore this, because this is going to be gone. It's a ghost. Now the question becomes, where are we on the open?
Where does this take us in the first place? This is the same thing, but it's the PM position. So now you can continue the exercise. Where did we open? Did we open at 7600? Do we open up there? Well, here's a guide.
If we open on the higher end of this and get a little follow-through, guess what? The day's bullish. If we open up here and don't push higher, there should be a little pressure back to about 7650. I know on the vanna webinar I said I expect 7775. You have to take this in real time and evaluate as you go. I wouldn't expect that by default, not anymore, necessarily. That was more like, if we got to 7700 or 7710 today, if we kept going... this morning we were trading 7750.
That was something that was very possible to me, but not anymore. I mean, it's still possible, but I wouldn't bet on it now the same way. So for tomorrow I have call spreads that are 7670, 7680, and basically I'm long the 7660/7680 call spread, thinking that it will pay. Makes sense.
And yeah, I still have a bearish bias for the rest of the month, more like through October. I just don't think we're going to get away with this. I don't think we're going to get off easy.
Positive gamma gets more and more narrow and localized to the strike, that's correct. Exactly correct. Eventually, you think about an option distribution, this is how I teach in mentorship, trying to refine this methodology. Think about an option distribution.
The option distribution has a few key points on it that you need to understand well. If you understand those well, you can start to compose spreads and ideas in your head with just three key variables, and then understand when they change and when you need to adjust. It's very basic.
And the straddle is the gamma option. The skews have path dependency in them, and the wing options are very vega-sensitive, implied vol-sensitive. But they all have their characteristics, and gamma is most important on the at-the-money option. When you have a lot of gamma from options that are not at the money, you have to work really hard to hedge those a certain way.
And charm flow, by definition, exhausts at the strike. A 50 delta option has no charm, it has no vanna. But at expiration, as the clock actually expires and the options settle, it will have vanna and charm at some point, because even a 51 delta option will be 100 delta, and a 49 delta option will be 0 delta. So it is there at the end, and it's right on the strike. That's why there's pinning.
On your question about gamma and how it's done, every dealer is going to have a different approach. Obviously, as a market maker, your trade-off is that you don't want to spend too much money hedging. There's the actual cost of hedging, the trading cost of executing a futures trade is non-zero, but there are also the implicit costs. So you're constantly trying to evaluate what kind of range you have. Do you have a type of market that's mean-reverting and highly erratic back and forth?
Because you hedge that gamma very small, in small clips. Funny enough, you would hedge that kind of gamma to take scalps locally, because if you didn't, you wouldn't go anywhere. Of course, when you think the market's like that and you hedge gamma that way, you actually reinforce it.
Same with, if you think the market's drifting and trending, you would hedge gamma with more tolerance, and probably have some kind of capture on a reversal logic. You do a lot of smart things way above my pay grade. But it's funny, because in both situations this is just inescapable: your read of the market and your choice to hedge it reinforces whatever kind of market it is.
But it's different for everybody. Once upon a time we had a gamma hedger. I ran a desk in 2013 that didn't go anywhere, I knew nothing at the time. But we had a gamma hedger that moved a dollar, and you hedged it linearly. That was stupid. We're way beyond that now.
Nobody does it that way now. That's why I think it's important that in RTM we look at the straddle, we look at the half straddle and the full straddle, because oftentimes those are measures people hedge. The one standard deviation move might be hedged. So I think it's smart to look at those boundaries, from the open, for example, and expect some interaction there.
Matt was at a firm, he was at UBS, Credit Suisse, some other big banks, along with Belvedere, but he would not be hedging as locally as we were at Belvedere. He would have wider ranges.
Yeah, I'm still bearish for October. I'll put more thoughts out over the weekend, but I think tomorrow what I'll probably be doing is adding some puts and put spreads, far out-of-the-money puts in October, end of October, that have some wing convexity on them, cheap things, and then stuff that expires next week, maybe Sep 25th, but higher strikes again.
The Sep 23rd worked, but only if you nailed the point-to-point. And it didn't work 10x at all, it worked at best 2.5 to 3x, and I did not get that. But again, that's also a learning experience. It doesn't make sense even to me in arrears. My own hypothesis was that we flush on the meeting.
And that we then bounce and slingshot higher, and then we crash. So why would I use the Sep 23rd put spread?
Instead, when I think the straddle's cheap, if I use downside gamma expiring on Wednesday or Thursday to capture the first part of that leg, I would have done way better. That's why I want to remind you: trading is not easy, it's difficult to get the expression exactly right, and don't hang on my words, because some of you have way more capacity to be head down in the market and thinking about exactly what you're going to do and where.
I was on a phone call, I was on a live thing, and then my son sprained his wrist. And I couldn't actually do anything at the bottom. Very frustrating.
Yeah, that's exactly how to think about this. I hope that what VolSignals continues to be... we're open to new traders, obviously, but I hope what it continues to be, and be more of, is that we get good traders here and teach them not to doubt themselves or think about what they're doing as junk. I don't want you to do that.
If you have something that works already and you're pretty decent, we're trying to give you a leg up. You're like, oh wow, now it really clicks.
It's that disorderly period, the 3:55. You guys see why, in active markets, I sometimes wish we could go back to 2024, because it was so nice being able to hold stuff until settlement. But now, imagine you traded my balance idea in the morning. We had a beautiful move for that.
7610 was my test level, and what is this, the low? 7611 something, very close, very good. 7615, very good. Not quite back to 7650, but a good move. The way I read this is from test level to test level in an active market. I never saw 7615 or 7625 as balanced, really. I only saw 7645 as balanced, so there's too much charm up.
So there was a lot of interplay. I'm not going to lie, many times I got frustrated and doubted myself. I didn't really trade this at all past this point. I got a little more exposure here, and then I just lost interest and hoped this would be the outcome.
The only point I want to make narrating this is that it's interesting, and you see it a lot: if you have a fly oriented around the balance point with the market maker's flows, it often does this. But this is why I don't want to hold an expiration. Look at it. The whole thing, I mean, what was the straddle price here? $3?
$4, and we moved $10. If you had a fly, this is what happened to me. One of my first trades was a fly trade, actually a really good one. It was a 1220, 1300 put fly. And it happened to be when Fukushima happened, whatever March that was. I had a 40 lot.
That was huge for me at the time, because I was a pretty young trader. I paid about $3 for it, and it was centered around 1260. The day before expiration, and it was an AM expiring trade, the night before, futures were trading 1257. I'll never forget thinking, wow, I am a hotshot, I'm going to buy a Ferrari tomorrow, and all of that. And we were out at a steak dinner, a super expensive steak dinner that I had of course volunteered to pay for.
And then the next morning, remember this was trading 1260, and the night before we were actually down to about 1257. The next morning we printed 1287. Which doesn't sound like a lot now, 30 points, but at the time that was a 2.5% rally overnight. And it was just, ugh, what have I done?
It was pennies at the end. Still a great trade, but it was a nightmare. You see this more and more at expiration sometimes, especially in active markets, so that's the answer if you guys ever wonder why I don't hold the whole thing until the close. Sometimes I leave runners, but often I'm afraid of exactly this.
Cool. I'm going to hop now. Matt, you can keep answering, and if you guys have any other questions, feel free to keep posting.
Can you hear me? Okay, cool. Actually, I have to hop too. Is there a specific question? I have to go pick up my son.
Alright, well, are we all good? Good session, guys. Any questions for me? I have time for one.