BlowFish

Not just at the tick level, sometimes the ES will look weak and the European markets whilst not surging ahead will not drop so far. Other times the reverse. Currencies, notes, and even commodities play there part too. It's a huge intricate web imho. Also a lot of the largest companies in the world that directly effect the indexes are in Europe and Asia.

It will depend on where the data is coming from. For Interactive Brokers it is TICK-LSE ($TICk is TICK-NYSE). The fact IB don't provide a DAX one (which is only 30 stocks) leads me to suspect that it is in fact an exchange provided metric.

Firstly good luck with the thread. You wont need it, nonsense isn't courted here. I have always thought that the labelling of volume sequences one of the areas that could have increased clarity. The volume segments 2R and 2B have different meanings depending on context. To make sense of any element or pair of elements in the sequence you need to look back until you discover B2B or R2R. Of course things aren't going to change, but If you labelled each segment in the sequence with a 'vector' things would be much clearer imho. So each term would have Volume annotated (D) Diminishing or (I) Increasing Direction annotated ® Down (B) Up So a down trend would be DR IR DB IR An uptrend would be DB IB DR IB The beauty of this is it allows you to look at any part of the sequence and know where you are without having to 'look back'. You could bracket (DR IR) to signify that the sequence has changed direction if you like. When I studied the material I went as far as ditching R and B in favour of (U) up and (D) down. After all the colours are a visual aid. You then have DD ID DU ID. It 'sounds' better in my head too, of course it makes communication with other practitioners difficult as you need to 'translate'. Anyway I don't want to derail things but back in the day when I was trying to figure out this approach I found this difficult, partly because I found that particular annotation counter intuitive. Most of the rest of the terminology I thought was well chosen and so, intuitive.

Indeed. Those interested in the original post could do a lot worse than search the interweb for posters called Trader Dante and James16. They have written a lot about this sort of approach.

Another is it's actually a betting platform (spread bets, CFD's, and spot fX bets) unless I am mistaken (which is certainly possible). This rather limits it's application.

Yes, thats pretty much it. You need the two inputs also.

For any interested in 'volume' its probably worth taking a look at Studies in Tape Reading by Rollo Tape (a pen name of Wycoff) or Tape reading and market Tactics by Neil. These cover 'volume sequences' but in language that most any market practitioner will understand.

Very simple Syzmon. As I mentioned 11 dot and 11's are floor pivot and S1 R1. I simply added R2 S2 (using the 'floor' pivot formula) and averaged those for 3 bars in the same way that the envelope works. I'll try and post some charts later.

Try something like..... inputs: iStartTime (0800), ResetMinutes (60); if mod( (TimeToMinutes(time)-TimeToMinutes(iStartTime)), TimeToMinutes(ResetMinutes) ) = 0 then begin reset code starts here

If you are going the full mechanical route it might be wise to run a couple (or more) systems on several instruments. This approach is more robust. There is a bit of an art (so I am told) knowing when a system needs tweaking when it needs a rest (some stop working then start again) or when it needs retiring.

There are several ideas here http://www.traderslaboratory.com/forums/f229/how-do-i-avoid-chop-chop-6104.html You might find Ehlers books interesting, he comes from an engineering background and is looking to isolate the cyclical component using signal processing techniques.

I believe you can program ninja to generate order book changes. In tradestation you get 'events' for actual ticks within the event processing you 'peak' at the current bid ask. This leads to race conditions or sequencing errors. You can minimise the effect of this in TS but not totally eliminate it.

You might want to google open interest.

Identifying trend or chop (as broad categories) might not be a bad place to start. Many automated systems either trade one or the other though some 'flip a switch' and essentially use a different approach for those different market conditions.

Could be 'flawed' but still 'work' FD. I believe zenfire is a complete feed (that of course is in respect to actual prints) not sure about the inside bid and ask info they provide, one could always shoot them an email and ask. In any case I think NT it is a better platform for this sort of analysis as it is event driven.

Indeed you might want to look at my post a couple of days ago with the DAX chart.

I am pretty sure they are reported (i.e. cross the tape) but you are right they would not show up on the order book. It puzzles me exactly how it works if it is negotiated away from the market price.

I did have a search of the CME before I posted and if there was an obvious helpful link I would have posted it (honest I did). There are quite a few links returned and I went through the first few but nothing that helpful (just talking about it in principle). There is a minimum size that is set (and sometimes adjusted) by the exchange. Eurex has the same feature too incidentally.

Using SD's actually is a no brainier (well it is to me but then I am a no brainer myself )....except they are rather slow in execution unless you implement them as a continuous algorithm. I also wonder whether it is worth eliminating outliers in certain cases. For example Globex allows you to negotiate blocks with a counter party outside the auction process (i.e. off the exchange). This is then reported as a single large block trade. There is a minimum size to use this feature (can't recall what) but it is probably worth eliminating them altogether.

I think I mentioned this earlier in the thread but you could use one of the 'fancy' low lag MA's Jurik, Hurst, Ehlers etc.

This is a function of taking a BO of the previous bars high/low. You are always entering with the immediate trend (in respect to the last 5 minute bar) and so would expect momentum to propel the BO. 6 ticks has always been a popular point based stop on the S&P though as you observe the current volatility is whats key.

As far as I know dealbook is pretty much propitiatory to GFT (they may license it to one or two others). Long story short you are unlikely to get much help from anyone but GFT or maybe on thier forums (if they have any). Traders that use it will be a very very (very) small percentage compared to more main stream platforms. Ask away anyway some one might be able to help. If you are trading spot FX a free account with any of the MT4 bookies will give you a seperate free charting platform with a better chance of finding the indicator you need. EDIT: What does deal book have in the way of scripting or programmability?

Place an 'emergency stop' when entering the market. This is not based on money management or market structure it's the 'oh crap the PC is broken' stop. It's a shame that more brokers wont hold OCO orders for stop and target that are contingent on your entry stop/limit order being filled. If your broker does provide this use it...if not pester them to implement it. A failure between broker and exchange is less likely than one between you and the broker. If you enter on stop consider placing the order as late as possible. Problems can occur when you have a resting order at the exchange that is elected whilst your PC/Internet/Phone is broken. This is less practical for limit orders as you will be further back in the queue.

If I was to tackle block size as a project I would look at some sort of statistical analysis of blocks to determine dynamically what was 'large'. Not sure how well that would work but would dynamically adjust for seasonality and may be even fast enough to adjust for open, lunch, globex etc. I think this might be surprisingly powerful, send me a cheque if you make millions from it

Thanks! I had planned to write something for ninja (event driven is certainly the way forward) when I got bored of a MMORPG that I just signed up for - Now I can play longer and just butcher this if needs be!