Jay

Why I LOVE CUMULATIVE DELTA and think profiles are SHIT! Written by Bartholomew Fatus Catus Back in 2019 I made a video talking smack on cumulative delta. I'm sure it's still up buried in my live stream tab. It had a cult following back then and quite frankly it didn't make sense to me. The divergences people used it for back then were horribly unreliable. It's one of those indicators where sometimes it goes up all day and price goes down all day. Or they work with each other going the same direction, or the divergences failed or worked in the most random ways. Basically a shit indicator like all the indicators. At least this is what I thought. See I used time and sales, a DOM, and volume profile. Even VWAP, which is another story. I had big pee pee energy well before order flow was cool. In fact I spent close to a year trading like ugh... ZN ZF ZB UB TN ES??? I think? 6 symbols all at once DOM only. No charts. I use to raw dog it hard, before going to charts. If you never tried raw dogging just a DOM - Well it's worth the experience. Not for the sake of you might be consistent from it. But more so from the perspective you will be forced to focus on just that. I also ran profiles on all my DOMs. In fact I even use to hand write TPO in the note section. Now at some point jigsaw my DOM package - Has auction vista, which is a version of bookmap with a heat map on resting liquidity. When I discovered this and started trading ES only, I found it to be? I wouldn't say powerful. But it was good. It was price action on the X axis VS just Y axis. So this means I don't have to remember some of the movement unlike just staring at a DOM. Staring at a DOM sure does help you internalize ranges and shit however. BUT!!! That memorization of price action, does get blurred. With some form of chart you are now going from a 1 dimensional space to a 2 dimensional space. Price only goes up and down afterwards. Not fucking sideways. See trading pure price action via a DOM means you're always on the hard right edge. But for whatever reason those rotations still fucking mattered. Past price action still mattered! I could see at prior rotations or stutters near high of day - context forming, or a story. Pure hard right edge trading is not about the moment NOW! There's real pressure accumulating, and when I saw resting liquidity on the heat map I could see the testing and probing of these areas. Now I had started on technical analysis like any trader back in 2016. John Grady was the only order flow guy out there and, I had went raw DOM like him. I had for some reason ran CVD which is short for - cumulative volume delta - on my auction vista the entire time. Why!? Don't know. I really tried using it and it just sucked. Somewhere after COVID Merritt Black left SMB capital and started Apteros. I remember when he wrote a blog on how the DOM was "archaic" or old fashion. I still have him specifically saying "archaic" seared into my mind. I was obviously offended by this. I had followed the content he had been on and liked him. But calling the DOM "archaic" really left a sour taste in my mouth. Not because he called it that, but because he mentioned how the fucking CVD was better and more modern... There's a few types of traders who have cult like mentalities, and if you or anyone attacks their "thing" the hand of God will come upon you. Those traders are: -DOM boys -Cumulative Delta boys -Auction market theory boys AKA Volume profile -VWAP girls -MACD kids At least that's how it use to be. Times have changed, and no one gives a fuck about MACD anymore.. Now it's GEX. Needless to say my precious old fashion DOM was attacked, which was not cool. I did nothing about this however. Weirdly enough I kept the fucking thing on my auction vista chart. I ended up doing the 2nd ever round of intensive with Apteros. I traded alongside and just listened. For what ever reason he could fucking read it. He could make sense of this piece of shit line, that for some stupid fucking reason I still had on my auction vista! Needless to say, there seemed to be magic here. Many years later when I was hardcore on my reclaim and slip rants with price action, one day in the discord I was trading and for some reason pulled it up. There had been some die hard enthusiasts on the CVD around me who just swore by it - and I fucking noticed it was reclaiming as well. It was almost like the price action. So I attempted aligning the reclaims on price and delta and at times it would work really well on momentous pops. A year or two later, I was doing a lot of time and sales work. I ran like 50 different power meters on 50 different time and sales. All for their individual sizes. 1 lots 2 lost so on and so fourth. I just had always noticed 11 lots would always seem to trade in the same area and bail on the tape. To me that was one guy. Especially on these odd lots as you went up in more size. I had been going down a conspiracy on wash trading. I felt a lot of futures volume was fucked up. I believe this is around the time I made my order flow is broken videos. So I attempted to vibe code some shit to remove the wash trading. Spending a full week failing. My dumbass realized that buy orders minus sell orders is the fucking delta and it already removes the wash trading. So, now I have a second perspective on CVD. Side note here. Everything in trading has multiple perspectives. Rainbows for example require the viewer to be at a specific angle in the right place to see one VS someone else at a different angle. The rainbow alone is a symbol that can represent different things as well. For most it could be a symbol for the LGBT community, for others it could be a beautiful phenomenon of mother nature, for others a symbol from God in the book of genesis after the flood with Noah's ark - or spiritual. A tool such as CVD takes on different meanings. It can have all 3 defined meanings like the analogy of the rainbow. For me at this point in time, the CVD represents this reclaim vacuum, and removal of wash trading, or raw outstanding order flow. The order flow I felt was what moved the market, as my conspiracy theories believed most orders where washed at the time to bait in volume. Also HFTs working spreads can stop out break even a lot because the code doesn't see correlated movement elsewhere - Thus this is a form of a wash. As I continue to write on CVD I want to let you all know that this line of thinking had comes with some flaws I now see. Just because volume is matched and there is no buyers or seller imbalance seen on a delta doesn't mean that matched volume was purely washed. Meaning that there can be open interest in the matched buying and selling. Much like you and I trade with each other both opening a position. We both have open interest at the same price, but the delta is zero. So at some point a push in either direction can lead to one of us stopping. Some months later after understanding that delta does remove matching, which seems obvious, and in a way I knew that before hand - however as I continued to learn, there was one source that pivoted my line of thinking forever. I read "A Complete Guide to Volume Price Analysis" by Anna Coulling She taught a lot of volume histogram in there. Which I had turned back toward and gave up on after a few months. Will until more recent. But this was some time ago. This is beside the point, but context we need for later. The main point was her chapter on market makers. They obviously play the spread. But they also do position trade. She even mentions she swears she herself is not a conspiracy theorist on this happening and gives older references. One being Wyckoff. The market makers have to balance inventory. This is the moment I had a pivotal AH HA! moment. The wash trading on the delta is imbalance. Meaning more sellers or buyers came in once you remove the matched orders or at the time I thought washed trades. This imbalance needs to be rebalanced. The market makers must always rebalance their books. This is why we see the CVD reclaim like price action. I also have a personal friend - whose father traded since 1985. He always told me they needed to balance the book. When you watch CVD you can see massive delta spikes. These often can come out near or around the same prices with spikes almost similar in size on the CVD. This made me realize it was similar to my time and sales tests where I tracked odd lots like 11 lots and 31 lots all coming in and out near the same prices, and most likely this was the same trader or algo, as some of these lots just were too odd and hit very fast in and out in the same area at quick speeds at times. Surely the probabilities that some of this volume or price action was the same entity. The CVD did the same thing. What this told me between tape and delta is large traders or most of the trading activity is like retail. Large traders take losses and they all seem to be trading some what intra-day or shorter time framed. The CVD could have 1000 lots spike. This would cause a big sweep on price. Clearly or most likely one trader market order 1000, but due to the lack of liquidity he swept many different bids or offers taking all that was there. But often at times you could see the price put pressure on his entry area OR the delta would maybe chop for a while - but then come back to that initial delta spike of 1000 and push on it, and you can see that same 1000 delta spike react the opposite way which can spike price down - or sometimes someone seems to know there will be a flood of volume - and runs an iceberg and reloads INTO that delta rebalance effect - WHICH FILLS his passive order! This IS a rebalance effect. You can see it all the fucking time to this day. Not always panning out exactly the same way. But predictable enough it's a contextual clue, and it can be leaned into. So we know market makers have to balance books, and we know delta causes one sided offset meaning one sides volume has either more buyers or sellers. We know tape on individual odd lots seems to come in and out at roughly the same areas at roughly shorter intervals - tipping off how short term alot of trading is - and the delta also does this even on super large sweeps like 1000 lots. Therefore these are larger traders and even they bail or get stopped, and also have shorter time frame participants. Now keep in mind when the delta spikes like this, someone else is on the other side of that trade. It's zero sum. Much like how I mentioned if you and I open interest against each other. I'm long, you're short - one of us must lose. This explains why larger traders seem to lose at times when I see the delta spike in and then back out at a worse price or sometimes the same price. Someone is always on the other side of that spike. Large traders need to squeeze other large traders or enough large volume to make profits. This is also why when I see odd lots of 11 or 31 individual tape orders come in, they seem to come out pretty quick at times. Someone is on the other side of that 11 or 31 lot. I had often forgot that someone is always on the other side, and just looked at these things as one participant - forgetting it's two sided. The community also spreads misinformation on smart money and institutions especially. So many institutional sized traders go under and lose. Leopold Aschenbrenner is an example of a large trader who loses. Just because someone can trade hundreds of millions in capital doesn't mean ALL of them are always right and know what's going on. Large whales lose. If they didn't their ROI would be as high as a shitcoin pump year after year. But most fail to outperform the SP500 year over year. They're no different than us and how we trade. Don't get me wrong. There is highly sophisticated players. Paul Scott on chat with traders EP 328 - mentioned in the pit era trading copper - if large players could squeeze another large player they almost would always try. Hence why we can see large delta spikes clearly lose at times, seeing how someone is on both sides and one guy must of got squeezed. The GME squeeze is a prime example of a large player getting fucked, and how that lead to massive profits for some retail traders like Roaring Kitty. Areas where a mass of stops and institutional sized stops or the need to bail, is by far the most lucrative areas to squeeze, and much like us who rather have bigger one way moves to ride - other whales want this as well. Why churn tight shit? Sure you can run spreads in options on this. But a squeeze produces money faster than time decay spread trades, and when those lose, they always lose more than they can ever make. See. It's this perspective. I see the CVD in a way others don't. This is also coupled with much more deeper theory and other unique prescriptive shifts. Now let me explain why I don't use profiles anymore. Nor do I like delta profiles. After so many profiles you start to realize bell curves are ranges on price. The price action on very identical profile shapes - whether D, P, B, d, shaped - The problem is similar profile shapes have vastly different price action. A textbook D shape can have the price do a model range where each up and down leg hits the profile high and low. Or we can have it get stuck on the edges or on one side of the POC forever before going to the other side once, then back up. Or we can have it get stuck on the top only, then bottom only - then dump. Or we can have it do the most sloppy unpredictable price action in the D shape profile. Or sometimes we have an uptrend that channels up and makes a perfect D shape bell curve. I got to trade price action no matter what. And leaning into any standard profile had been met with problems as the price action matters more than the shape of a curve. Curves or profiles are often associated with auction market theory. But in my own 10 years of trading experience, only ranges were true structure. True auction market theory. The edges are discount and middle is fair value, much like profile is taught. But profiles again as I explained how one D shape without price action can look very similar - once you look at the price action especially on a 30 second chart - you can see how vastly different the price action is, which I have to still get right! Side Note - A 30 second chart is the smallest time frame I can fit on my screen that shows enough noise, but I can still see the entire RTH session on one screen. Higher time frames tune out noise that you need to know for more precise timing. Too low of a time frame removes broader context. So context to noise ratio is important. How low can you go but still see the big picture? This does often require more time framed charts, but at the cost of splitting mental bandwidth on processing "more" Now with this all being said I realized I could see the profiles without seeing the profiles. And trust me when I tell you - I have significant experience using profiles as the most long running tool in my career that I eventually retired. Removing unnecessary shit on the screen prevents my eyes from losing focus, on shit that quite frankly I don't need to see because I can see it without it being there. The less I need to look at, the more free my mental bandwidth is for processing information - thus making decisions faster. The copper trader I mentioned talked about how they traded ranges in the pit. Dealers often executed clients in the middle and bought and sold for themselves on the edges. This is true auction market theory or true structure. A "D shape profile" not being complimented with nice ranging becomes useless from my experience. Even leaning on low nodes and high nodes. Leaning on this stuff became more unreliable as the years went on as well. I also believed in the wash trading at the time and resorted to delta profiles. The issue with these is the delta only moves up and down on the specific price trades. Where as CVD had clearly shown me larger traders with spikes that do not happen at the same prices. Much like watching 11 and 31 lots entering and bailing on different prices in short periods. Therefore a delta profile has a major flaw. Traders who often enter an area like you and I and everyone else - often exit somewhere else. The issue with both standard volume profile and delta profile, is this: volume is collecting and stacking creating these shapes. But the trading is most often more short term among traders. Therefore they're entering and exiting in vastly different places. Yet the issue with the profile is it's just collecting information or data that is no longer relative. It's often filled with stale transactions which create these shapes, and for me I want to attempt to figure out who is doing what where and why, and when they might get squeezed. Knowing who has to bail and where they are wrong is edge, because there WILL BE A MOVE! Triangulating this is difficult, but can be done. However the staleness of a profile showing all transactions if far from a solid filter. It's more like a hoarders house. The more shit in the house the harder it is to find what ever it is you're looking for. Now obviously delta profile removes this right? You can see the specific prices actually move up and down on buying and selling unlike a profile that stacks it and never removes the stale shit. Delta profiles are flawed. They only move up and down on one specific price like 90.00 for example. But I mentioned the time and sales odd lots and the CVD showing clearly same traders bailing in and out at different prices on very similar volume, but at more random "times" - "times" being a reason volume and delta OVER TIME matters more as this adds a 2 dimensional view. X and Y axis. Not just the Y axis that volume profile, delta profile, and just a pure DOM are tied too. Thus as I wrote earlier on the DOM, it makes "remembering" what price had done, become harder, as a time based chart helps us remember the precise ebb and flow of the price action we seen on the DOM. A delta profile can not pin point with precision at what time - as it's always random - When a larger position bails or the rebalancing effect - a CVD can show. These CVD spikes and even time and sales larger odd lots can come out in seconds, minutes or hours. You can see specific CVD spikes come out hours later and even anticipate it. Now I'm passionate about everything I'm speaking about here because it IS MORE THAN THEORY! It fucking works for me! A one dimensional tool fixated on Y axis only requires you to memorize these moments of volume events - thus clogging mental bandwidth and ultimately leading to you forgetting. The standard delta profile has another major flaw. The delta coming in and out only at the specific price it trades at. This means any large CVD sweep on multiple prices where a large trader actually swept the book on say a 1000 lot, now gets cut across multiple prices. It is not only getting cut across multiple prices, but merging with other delta volume orders for the last several minutes which become hours. Now we no longer can pick out this larger trader, who is clearly one entity as these mass sweeps are shown on the tape. And depending on how you construct your tape you can see that it indeed came from one source at times. But now it's cut across multiple prices and merged with stale flow even on a delta profile. So how can you fucking even know where that one specific trader is, or even know he exists to begin with - or even if you know about him - pick him out of the profile that is stacking? If traders enter and exit at different prices, the delta profile does not remove delta from older prices as it only reacts to the price at hand. Thus stale orders are still stacked in certain prices giving you the illusion, there is still some form of open interest here. I had remedied this. I built a delta profile that removes stale orders and responds more like a CVD which I no longer use but Chronik my coaching student swears by. In the end that remedied profile still stacks orders on a Y axis only cluster fucks shit from hours ago and cutting up any larger trader who swept across multiple prices make it near impossible to pick them out. It's rare I see pure Y axis only DOM traders. Most every order flow trader uses some form of X Y axis charts. Whether it's time based or some other shit. I can go on about time based charts and why. But that's another post. You use charts because you can't remember that pivot at 55.70 - 2 weeks ago on just a vertical DOM. There's no fucking history there. Pure price action on a DOM is like a daily candle bar. It's one fucking candle that showed the open high low close. No day trader uses a singular candle to make a decision without some form of historic transactions. So if we tend to all use charts to see actual historical price action, then why the fuck would you pigeon hole yourself by looking at only Y axis or vertical volume only!? It seems like common sense at this point. But again much like me trying to vibe code something to remove wash trading and later realizing it was just fucking delta for the formula. The same under your nose shit was always here. It takes a perspective shift to see this. This is why I started going all in on volume histogram recently. If I came to this conclusion some time ago on delta, then why the fuck am I not doing this to my volume totals!? Will it's because I thought CVD showed where market makers have to re-balance. But as I know, the ES arbs the SPX price which is the official ticker of the SP500 which has ZERO order flow! It's a math formula based off the top 503 stocks weighted. Google has two tickers for class A and C shares. Hence why 503 stocks. These stocks by index weight which is a total of 100% divided across the 503 stocks and weighted more to the top performers are what moves the "hypothetical price" on SPX. No one stock can move the SPX as they can only have an effect based off their weight. So you often need a majority rules to move it all like voting but some votes have more weight. Every SP500 spin off arbs this price movement. This means volume elsewhere does move what we trade! However there's still market makers and participants in the ES futures. This is still zero sum. Someone is still on the other side of the trades that come through. Inventory can still become imbalanced and market makers still need to balance their ES books. They can cross product hedge all they want on shit like SPX, but at some point the positions on the ES book must re-balance as inventory opened in ES must close ES - therefore ES delta and volume alone still has value and directional cues without trying to pull in volume from stocks and SPX to attempt to gain the full picture - now there's still value here. But the ES book still has participants that must bail if they get pressed on hard enough like you and I, and market makers still need to at some point rebalance their ES book in ES. If they can rebalance their book in ES with ES - They will! Why let that become an ongoing problem and drag it out and make it more and more imbalanced by just cross market hedging!? This is why CVD imbalances can often rebalance intra-day over and over. This rebalancing effect is the CVD not only slamming back delta spikes, but also if it drifts or chops too far one way, it rebalances the other way. There most certainly are days where the CVD completely inverses from price all day long. More than likely there is either cross product hedging or often what I see - a rebalance taking place from a CPI delta spike several days ago. The "rebalance effect" has it's intra-day short term trader rebalancing every session to some degree - but can often get diverged from the overwhelming amount of one sided participants or even a lack of order flow in ES - seeing how ES arbs SPX by market makers pulling bids and offers to stay with the SPX - yet by exchange rules obligated to leave some liquidity instead of pulling it all - Thus forced into an imbalance based off this arb effect and a lack of third party market orders driving the ES with SPX. These phenomenons are not understood by the trading community, and makes CVD look broken on certain days with clear opposite diverging from price all day, or even a lack of syncing with ES price. Any standalone symbol with it's own order flow and no options chain meaning it's order book is not fractured. Should have a CVD for the most part match it's price. TO BE CONTINUED My recorded trading sessions and course modules build around this post at fatcatsofwallstreet.com/ - where I go over deeper theories on this and trade my actual concepts on CVD explained here with a degree of success leaning into "Fat Cat Theory" Please heart the post!