Hi Did anyone effectively adopt usage of delta imbalance when trading ES? I was asked to write some test algorithms based on bid/ask imbalance data. I did research the concepts and even some vendor videos ie marketdelta etc in order to try to see if they work and really help to determine a reversal. I did run comparisons on 5,15,D charts in order to see any dependencies. I was not able to highlight anything definitive. The only information it really gives you is how many hits were at BID and ASK ie someone may hit 100 contracts at BID and they will get consumed, but it does not mean that supply for ASK is over....and a downtrend will start or continue... However, it does not look to provide a validation of that supply from one end or another has exhausted itself. Is this just a 50/50 thing or I am missing a clue here?
There are 2 ways to use cumulative delta... One is cumulative delta summary for any time period(say like from start of pit session or start of day). This keeps a running summary of the contracts traded at the bid versus at the ask. The other way is this is to use delta per bar. Delta per bar is the contracts traded at the bid -those trade at the ask. And you can use any bar you chose.. time based bars, range bars, tick bars, etc. In either case , you are looking for divergence .. and both can be helpful if you know what to look for and have a good accurate data feed. Many data providers do not provide accurate bid vs ask data , so the data will be meaningless. Either of those methods are confirmation tools though and should only be used for confimation at a predetermined place of entry based on mastery of another method.. whether it be price action, volume profile, or whatever. A common mistake is to use delta volume as a primary tradng tool and it is not suited for that at all.. I have attached an example of cumulative delta summary on ES just for your reference. Hope this helps. Wolf
Thanks for the input. I have ATAS trading platform at hand with AMP market feed. I am using it to do the visual research for futher algorithmic implementations via API. Any feedback on their feed and maybe any feedback on a reputable data feed providers to use?
Have you ever thought you are trying to find something no one else or not many are doing and you trying to re-invent the wheel? There is a reason why better traders trade the way they do as opposed to those who don't, do something simple, concentrate on money management 99% and trade it the way you wrote the rules, that is the toughest thing to do and best automated. CQG might still have best data around, but when you know how to trade, you find that data not as important or how many jumps between you and the exchange, speed of PC, fast cable speeds is nice, why worry of a nano seconds.
I do not trade myself. I work for a company as a coder/research person. I do what they tell me to do/research and try ) I have also heard that best traders seem to use simple stuff. I have not met one yet, so I can't validate
Just wondering if these quant market researchers use the conventional tools such as Ninjatrader or Tradestation to run the tests. Any ideas?