I've skyped a bit with some DOM analysis traders (Futurescalper of Futurescalper.com & Viper of the recently vanished viperspeedtrader.com, as well as a nice guy from nobsdaytrading.com), and concerning my first two sources, what they propose is that price follows size on the side of the DOM ladder. If this is already understood here by the users in this thread, I apologize in advance. The counter-intuitive idea here is partially based on the theory that there are players in the futures market that almost always execute their sell orders on the offer side and their buy orders on the bid side, where as most participants like us are doing the opposite: buying on the offer and selling to the bid--we are almost always giving up a few ticks to get in the market-whereas these more powerful players are almost always making those ticks rapidly as price moves up and down. So, these "rulers of the DOM" as I call them are supposedly able to take control of a few levels of the offer lets say, AFTER they have already accumulated long positions at the bid...then when they are ready, they'll begin moving the offers up and the majority of market participants see this as bullish and begin to buy from these rulers of the DOM--helping them unload their longs at a profit. This way these players are almost always making the spread between the bid/offer-even if this spread is only 1-2 ticks, over 1000s of contracts minute-by-minute those ticks add up for these guys.
So, what we see is that the offer side begins to get larger. Naively, most people think these are sellers preparing for the market to move the downside. But, according to what I've been observing, these are buyers looking to unload their longs at an advantageous higher price & might convert to shorts as the market moves in their favor higher. Vice versa for shorts, we'll see the bid side get fatter, most folks interpret this as buyers getting ready for the market to go up, when it is more likely that these powerful players have already accumulated short positions, take control of a few levels of the bid side, begin pulling these bids lower and the market sees the DOM move a tick or two lower and begins to sell into the bid helping these more powerful players take profit on their shorts.
I haven't used this DOM bands indicator, but I think that the ratio part of the band that is showing the dominant side of the DOM is being interpreted incorrectly, where I believe it should be pointing above the zero line if the offer side is getting larger and below the zero line if the bid side of DOM is getting larger. If this is already understood and it is me who is misinterpreting the screen shots, I apologize.
Sometimes, the DOM bids/offers seem to appear/disappear rapidly and many say that this spoofing proves that DOM analysis is nonsense. However, if you calculate for increased/decreased variance in DOM levels--how often they are returning to appear full after disappearing less and less, this could be showing a hidden persistance among the players looking to control that side of the DOM. So, the ratio of total bids to total offers is a lower tech way to track this persistance, but I feel that it is decent as a confirming indicator. I feel that once one begins to play with this tricky counter logic of these powerful players executing the majority of their orders on the receiving end & and not as clicking on the offer to buy or bid to sell, then the DOM analysis I believe becomes more useful. So, in this view, these players are almost always receiving our orders, and then moving the DOM side up or down by controlling the various levels of DOM price--kind of like a dangling worm on a hook and then the rest of the market jumps in and auto-trading systems get triggered and these players make profit.
Often just before a true bottom is put in place, these DOM controllers will fatten the bid side and push the bids lower getting sellers to jump down and sell to the bids...but secretly what is happening is these DOM controllers are getting more buy orders in at the best prices on the bid and will soon take the market higher by quickly fattening the offer, the sellers who were attracted to the fatter bids moving lower will have their stops run---conveniently triggering a series of buy orders which will be executing at the offer where the DOM controllers have already placed their sell orders, thus making a killing. This scenario will show the bid side getting fat as a bottom is being noisely made--then often very rapidly the bid size of DOM will become less and the offer will rapidly get larger.
The simple way to understand this theory is that price is magnetized to where size is on the DOM, offer side larger = high prices expected; bid side high = lower prices expected. Another way to look at this is that the idea that the larger amount of orders resting on the bid/offer are initially targets and not entries...then after targets are hit, entries will be entered in preparation for the reverse move. So there's a conversion taking place during the extent of the move from unloading at a profit to reloading for counter move. This is probably tricky to track exactly. Love to hear your ideas, and sorry in advance for my excited run on sentence vibe going on here.
Have some beers and watch Futurescalper's videos on youtube, search "Futurescalper" and he kind of alludes this this stuff. He's plotting the size as an indicator on his charts rising with offers being larger, and falling with bids being larger...often the move is predicted by a few moments with his lines. His analysis is much more complex-beyond my current understanding and involves sophisticated ways of disciphering spoofing on the DOM and real professional intention. I haven't used his software personally, only skyped and watched the videos available in youtube.
The following 11 users say Thank You to bgtrader for this post:
Favorite Futures: CL & 6e, looking at ES, ZB and AU again.
Posts: 2,100 since Nov 2009
Thanks: 1,104 given,
Zondor and I debated how to display the data; The attached show how the sum of the DOM rows display.
Have posted these before, but I marked them up a little better. I used to know the ES would move when the bid and ask were less than 5000 cars summed on the DOM. Once it got to way unbalanced it would stall and possibly reverse, like in the double top example.
R.I.P. Andy Zektzer (ZTR), 1960-2010.
Please visit this thread for more information.
Last edited by ZTR; July 25th, 2010 at 04:47 AM.
Reason: spelling clarity
The following user says Thank You to ZTR for this post: