[volume] Delta Volume in Intraday Trading

[ScottB 10]

AuctionMarket, thanks so much for that resource, it is great.

[sdoma 10]

looking at delta is like looking at a candle chart when price goes up so does delta so i dont see a point there, as far as delta divergence or anything like that their is no point because when price goes up in an aggressive fashion the orders are pulled out of the que so the delta will be magnified not showing the real intent or the actual quantity or strength...

 

The fact that orders are pulled from the book tells you something, doesn't it?

 

Like I said before, delta only tells you something in context. If there is a high delta with high movement, that means market makers and reversion traders have pulled their limit orders for the time being. If you see a high delta with low movement, especially in the context of a level, you can take that to mean that aggressive buyers or sellers have run into a large amount of limit orders.

 

There is no tool that will give you a complete and total view of the order flow. It changes much too quickly, and it is responsive and dynamic.

[AuctionMarket_Trader 10]

The fact that orders are pulled from the book tells you something, doesn't it?

 

Like I said before, delta only tells you something in context. If there is a high delta with high movement, that means market makers and reversion traders have pulled their limit orders for the time being. If you see a high delta with low movement, especially in the context of a level, you can take that to mean that aggressive buyers or sellers have run into a large amount of limit orders.

 

There is no tool that will give you a complete and total view of the order flow. It changes much too quickly, and it is responsive and dynamic.

 

you just went back to my point that better off analyzing limit orders

[TheNegotiator 13]

AuctionMarket_Trader,

 

Perhaps rather than trying to dissect something which others use, you could share exactly how you analyse the market and what the benefits over delta volume are with the technique.

[sdoma 10]

you just went back to my point that better off analyzing limit orders

 

Well, not quite. I think that both are important, and I don't understand how someone could claim to understand order flow without considering all types of orders.

[AuctionMarket_Trader 10]

Well, not quite. I think that both are important, and I don't understand how someone could claim to understand order flow without considering all types of orders.

 

We have and done extensive research on delta but found no edge .....

[Karish 10]

Here is the part that is difficult to explain: they didn't necessarily want those first orders, they were just placeholders. They would then cancel order and see if other algos followed suit thus allowing the orders further back to move forward. They would cancel orders/add new ones until they felt they were in just the right place in the book and then let them rest to execute or get pulled if the market moved the wrong direction.

Scott

 

Scott,

If I understood you correctly, a lot of the logic you describe above depends on knowing the location in the bid/ask queue. Currently, when I send my limit order I do not know my location in the queue, how do they?

 

Many Thanks

[MightyMouse 12]

We have and done extensive research on delta but found no edge .....

 

I am sure that if you keep doing "extensive research" you will conclude that nothing gives you an edge over time.

[TheNegotiator 13]

I do think an important point to consider is whether you are using this or any other method in discretionary or automated trading. I also would point out that as system designers/testers should know, a system is only as good as the research and testing done on it. Just because one person doesn't find something particularly useful or as having quantifiable edge, doesn't mean another won't. It also depends very much on your money management, style, account size, timeframe ...............

[ScottB 10]

Karish, first, they have completely unfiltered/uncoaleseced data so they see literally every tick, every book change.

 

I recently spoke with one of the major players in the data feed market and when I asked just the right questions, it turns out their ads about how their data is the best, you get every tick, etc. weren't true.

 

Their answer basically was - oh, you mean all the data like actually every tick, every book change? When I said yes that was exactly what I meant, the response was - oh, well, (a few umms in there too), that wouldn't be practical to deliver over the internet.

 

So right there is the primary difference. High end platforms like X_Trader Pro give you a column on their DOM screen called EPIQ - estimated position in the queue.

 

There are algorithms that estimate your position. Start with how many were on the bid when you joined the bid (you want to be long). Then subtract every trade you see at that price and then with the right amount of historical data, you can build algorithms that estimate how many orders that are cancelled were in front of you, which leaves you with a decent idea of where you are.

 

If you are a big enough player, you sprinkle 1 lots in all this and when one of them executes, you know exactly where the rest of your orders are relative to that 1 lot which also helps pinpoint where your orders are.

 

I can tell you from personal experience, knowing with a fair amount of accuracy, where you are in the queue is a huge advantage. Think of all the times you were trying to decide whether to bag a trade at the bid (you were long) or hold out for your limit order at the ask to be hit. If you know you are 3rd in line, you hang in there, if you know you are 200 back, you hit the bid.

 

I hate to say it but there is a world out there the average trader never even sees.

[sdoma 10]

Think of all the times you were trying to decide whether to bag a trade at the bid (you were long) or hold out for your limit order at the ask to be hit. If you know you are 3rd in line, you hang in there, if you know you are 200 back, you hit the bid.

 

Most traders don't realize just how small their edge is, don't you agree? Even a lot of the big players have an edge that is very small, a tick a trade, but they make trades with big size. So, even if this only improved your execution by a tick 20% of the time it's a huge advantage.

 

To those of you wondering about data feeds, as yourself, "Is this data really expensive?" If not, it's not unfiltered. I hate to say that, but it's true. Paying 100 - 200 a month for data is not going to buy the highest quality stuff.

[ScottB 10]

sdoma, that is an excellent way of putting it; I hadn't thought of it exactly like that. Institutional data feeds often are in the 1,000s not 100s.

 

The prop shop I worked at had trades that made $20.00 per million traded (notional) but traded over a billion per day (notional). That one trade made between 20-50k per day but it took almost 1,000 trades per day to do it.

 

We completely agree on the edge thing. I sometimes think about writing a book titled "What trading has taught me about life". If I do, I now have a new chapter - "Do you know how small the line is between being ok and not"

[Karish 10]

Scott, Thanks

As these shop doing most of their business in the limit side of the book they play Market Maker role, is the money they are sucking out of the market is it coming out of the overall MM business ,or in addition their practice cause for unfavored execution to liquidity takers?

[sdoma 10]

We completely agree on the edge thing. I sometimes think about writing a book titled "What trading has taught me about life". If I do, I now have a new chapter - "Do you know how small the line is between being ok and not"

 

A great analogy to edge is The Old Man and the Sea. You have this big tuna, your edge, and you have a bunch of sharks taking a bite here, a bite there, and all you have left is a pile of bones. Human mistakes, psychological problems, computer and data-feed failure, sick days, all of these and many more can erode your edge until there's nothing left.

 

You can thank me for this brilliant analogy in the forward.

[ScottB 10]

sdoma, thanks here and I will thank you again in the foreword.

 

Karish, MM come in two flavors, the true market makers (think NYSE specialists of the olden days) and prop firms that are just trying to take what they can out of the market.

 

I don't have an issue with either type, we are all trying to make money, why should they be any different?

 

I think both types provide liquidity by acting as the facilitator for anyone wanting to take liquidity. If you want to take liquidity, do you really care who you get it from?

[Karish 10]

Scott, thank you very much for the very valuable information. I had my guesses about how these prop shop work and you confirm many of them.

My belief is that even if I cannot play this game myself, understanding these players behavior and intention can assist me in building my own strategies / indicators to present their activity.

 

The other group I'm trying to understand/analyze their behavior are the so called "smart money" who execute part of their strategies at limit and not for MM, HST business. We cannot see them directly looking at Volume Delta, but their footprint is there.

[sdoma 10]

Scott, Thanks

As these shop doing most of their business in the limit side of the book they play Market Maker role, is the money they are sucking out of the market is it coming out of the overall MM business ,or in addition their practice cause for unfavored execution to liquidity takers?

 

You know, I wouldn't even worry about this as a trader. I used to get upset with whoever I thought was taking money out of my pocket, but I eventually learned not to take anything personally. Also, it's impossible to know why some large trade was done that spiked you out of your position. It might have been a stop run, but, especially in the indexes, it could also be that someone legged into or out of a spread, or an options player had to hedge a position in the futures, or a basis trade, or any number of complex strategies that many traders have no inkling of.

 

Worry about two things - your edge and how well you execute. Take care of those and you won't worry about market makers.

 

My belief is that even if I cannot play this game myself, understanding these players behavior and intention can assist me in building my own strategies / indicators to present their activity.

 

Just saw this. Go back to what I said above. A lot of trade, in the indexes especially, is complex, non-directional, and done for a lot of different reasons. It's good that you will try and model for it, but remember, concentrating on the WHY in trading will eventually drive you nuts. Find something that works and get as good at using it as you possibly can. The first part is easy. The second part is hard.

Edited March 3, 2011 by sdoma

[Dcash 10]

ok noob here, but after studying delta volume for a while along with cumulative Delta, wouldn’t the delta volume divided by total volume be more telling after all 500 delta means 1 thing when there are 600 traded and another when 5000 are traded or did I miss something?

[MightyMouse 12]

ok noob here, but after studying delta volume for a while along with cumulative Delta, wouldn’t the delta volume divided by total volume be more telling after all 500 delta means 1 thing when there are 600 traded and another when 5000 are traded or did I miss something?

 

Every trader is a Noob from their first trade to their last trade. And you are right. a delta of 500 could very well mean 2 completely different things if they have very different trade volumes. Sounds like you are on the right track.

[TheNegotiator 13]

AuctionMarket_Trader, in what way do you go about analysing bids and offers? Do you think it is easy to do for a retail trader without institutional grade software, hardware, connections and data feeds?

[ScottB 10]

Negotiator, I have and still do work on both sides of the institutional/retail fence and I can say from my experience is no you cannot do that type of analysis without the feeds, etc. The most insidious part is you can easily get the illusion you can.

 

The differences are subtle in nature and virtually impossible to reconcile if you aren't aware of what to look for. For example, how would you be 100% certain a trade went off at the bid or ask or how would you ever spot algos running deeper in the book without getting every book change, not just "most or enough" of them? These are terms data providers will use if pushed to explain EXACTLY what their data streams consist of.

[TheNegotiator 13]

This is definitely what I was getting at. I guess if you do the analysis on a long enough scale and look at the queues proportionally to one another, the specific orders that may have been 'missed' or even 'skipped' will become less and less relevant. So if you are a highly skilled analysis technician, perhaps it may still be possible to do this kind of thing with retail tools available today. What do you think?

[ScottB 10]

You have actually pinpointed virtually the only way you can do it at the retail level.

 

There are good proxies for most of these things if you truly understand what you are looking at/for. If someone is doing the equivalent of throwing paint at the wall hoping to create a Rembrandt that likely won't work (you might get a Picasso though).

 

However, if you are looking for market structure to give you a probability based edge then that is much different (imo). Start with something very simple.

 

One thing I do is show a histogram of the number of consecutive bars in the same direction. That is mildly interesting but I also know (not that anything can really be known relative to the market) the approximate probability of another up bar given we have had 4 up bars in a row and we are not within N minutes of a market moving report.

 

That kind of analysis take a lot of time. One of the worst feelings in trading for me isn't losing on a trade. It is spending a week researching something like that only to find it has zero predictive value.

[TheNegotiator 13]

One of the worst feelings in trading for me isn't losing on a trade. It is spending a week researching something like that only to find it has zero predictive value.

 

I can imagine! talking of counting bars etc, do you look at stuff like demark?

[ScottB 10]

I don't want this to sound wrong but I almost never look at other people's stuff.

 

One of the reasons I am always happy to share what knowledge I have is my guess is that almoste everyone has their own ideas and likely not enough time to research them.

 

If I can point someone in the right direction or at least warn them of the potholes I have fallen in along the way then maybe some of those late nights weren't wasted after all.

 

Of course, having just said I don't look at other people's stuff much, I must now go do a little research on DeMark, lol