Cariocas, I agree. I don't think it would be wise to trade just off MD, or MD divergences. I'd really like to hear a contrary opinion on this. My thinking is basically to use really fine grained charts (such as attached 5 tick - IB ticks - chart of DAX) with both market delta and order book delta to determine entry. If the market is not off to the races (as DAX was earlier), I think this gives also gives you time to get out if you don't like what you're seeing . This is not to underrate the importance of longer term charts, MP, pivots, VWAP and whatever else. I think this is viable for DAX and ER2 and maybe SGXNK. Maybe less so for ES.
Hi D' I <B>finally</B>... after reading and reading found some info on the Max top and Max bottom bands for the vWap... Grey1 never says what day to use... so it is assumed by you and me that one is using today's high - low / 2 (+/- vWap) but here is a web page that talks about using the TRADING DAY BEFORE and ALSO today's high low... anyway thought you might be interested... since how does one actually use today's high and low in the first half an hour?... it does not really compute for that... http://www.tradingsimulation.com/resources/esignal_vwap.html cj... HAVE STOP <img src=http://www.enflow.com/p.gif> WILL TRADE
dcraig, You won't get one from me. If you have a decent trading system/method MD can add significant value in terms of timing entries/exits. However, it is no substitute for a system/method. Having said that I imagine that in certain markets a skilled scalper could just trade from MD with a DOM. I don't know anybody doing it - but I think it's possible. bolter
heck, a really skilled scalper can trade just off of the time sales. not even a chart. i've seen it done. now, THAT is somebody in tune with order flow. i mean, really, every chart and indicator you have (apart from trin, tick, etc. which are overall market guages) are derived from the information in the prints (time/sales). so, a really skilled scalper can just watch the prints and mentally calculate (and feel) bid/ask volume, overall volume, support and resistance, etc.
DCraig.......put a "3 Period" smoothed MA on your "order book delta" and you may find some really nice entries at delta/price divergences. The "hook" in the order book MA at good delta divergence plots should put you in a great position for entries.....just a thought.
No doubt very true and probably requiring exceptional abilities of concentration - which I certainly don't possess. One objective I have in posting on this thread is thinking out loud to help my understanding of tape reading and the order book. Hopefully as my knowledge increases it will be encapsulated in my software. As you very rightly say, everything else (except breadth data) is derived from the mechanism of the auction market. How many books on trading or 'gurus' focus on this ? As we go futher and further away from this we start getting into what increasingly becomes flim flam - searches for the better oscillator or better MA or fibs or whatever - not because there is any plausible connection with the mechanism of the market but because they are claimed to better fit the time series data. A lot of it is just data mining - in many cases done very poorly.
I would love to hear specifically about using the delta for entries. Where exactly is the trigger? Here is what I'm doing with it and what I am observing. This chart is today on the DJ Euro Stoxx50. I have previously discussed about my reasons for timeframe length in my earlier post in this thread I won't bring it up here again. I would love to hear if people are using smaller timeframes with these charts and finding it useful and how it helps them trigger an entry. Usually when I see big volume prints like that at the highs then I will immediately view it as a a short-term R level that needs to be overcome. In this case, you can see the prints where light as it tested that level. As it was rejected, you can see the POCs moving towards the lower portion of the bar. Even as it pulled back off the lows the POC remains in the lower portion. All good clues on which way the market is trying to go. When I trade pullbacks on these levels, my order is there resting at the volume cluster. Easy to see where my idea was wrong. This case wouldn't have been an entry for me, but was the first sign of a trend change. Anyone care to share specifics about entry triggering with this tool? This was what I thought the tool would be useful for when I started using it, but I have had a dickens of a time making that work for me as a reliable trigger on short timeframes. Kind regards, MK
Just a quick follow up to my question about entries. So far I'm hearing only about divergences. Is this what people are mainly looking for with regard to entry timing? If so, then I find that can work at times for bigger trend changes. But what about timing of pullbacks unless they are more drawn out? I haven't found divergences show up very much for pullbacks. They tend to be more common around the turns, usually after a few false divergences. Maybe my timeframe is too big. Any insights are warmly welcomed. MK
Chart of bund with both vwap h/l bands for today and today-1 I have applied a 5 bar lag to the bands as it stops the chasing effect on todays. I have only just started looking at them but when todays cuts ydays band seems like another use.
Hi MK, Your stuff is great. I got the same feeling from reading the thread... I have to re-read for sure. But all I hear is divergence. I am at work right now so can't confirm the parameters I played with. I have been looking at consecutive delta bars in the same direction for trend filter (detect trend change). This delta persistence idea might be a little filmsy right now like every other crap. I like to use Fib number for my time frame (yes I am a cult member) and volume chart to compare apples with apples. I "think" I used 6765 volume ES chart and look for 4 consective bullish or 4 consecutive bearish delta as a possible trigger. I also use a delta filter < 10 since I found the small guy is more responsive and better at times and then I would look for a delta bar in the same direction (current bar or later) with a filter > 9 (basically same as without filter) to pull the trigger. This maybe just a mindless useless datamining exercise but anyone is welcome to try and input ideas. Please don't use real money. Thanks, William