BlowFish

Of course its up to you but why trade using a method that relies on volume with probably the only instrument that does not have volume reported?

Thats odd AAA the code you posted should run once a day! It should not slow things down. Multicharts rasterisation had always been snotbag slow, something I have tired of telling them about. Apparently they have speeded it up in version 5.0. It frustrated me no end that they did not put more effort into the fundamental things... receiving data from a provider....storing and managing it.....displaying it. They are getting (finally).

You will discover the answer to that as you read through the threads! Don't be fooled by what people tell you, there is no volume reported on forex, People happily (and successfully apparently) use tick volume as a proxy for volume. Be aware it is only actually a proxy for order book activity so you get ticks without trades taking place. For example If I best bid then cancel it this will cause ticks without a trade taking place. So in short why pick an instrument where there is no volume reported when there hundreds of thousands where there is?

This is pretty much the same question you asked elsewhere. Probablly better that you have all your stuff in one thread. DTN's IQfeed. Esignal. CQG (pricey). Or you might make do with a brokers feed.

Jerry uses Ensign Software. There are studies written for tradestation/multicharts. (I very much doubt they will run on OEC's platform) Ninja has VWAP and SD bands ....not sure about PvP. One of the guys here wrote routines for Amibroker and I think (though couldn't swear to it) that you might be OK with Sierra charts. oh it s Jperl btw

Just about all historical data will have time price volume. Different charting apps are better or worse at managing data....most are proprietary in how they do it. There is also tendency nowadays for applications to get data on demand and simply 'cache' it to disk. You may be better off managing your own data and putting it into a SQL database. I would have guessed that you should be able to find some free/inexpensive middleware to sit between the data providers api and your database. If there is no requirement to do it real time you could simple export data from a suitable charting package (usually comma delimited) and massage that. Or some data managers/charting apps will act as DDE servers (if my memory is not playing tricks). I guess it depends how frequently you want to get data and whether you need real time too.

If they are fungible (they surely must be?) then it has to increase liquidity and participation from the smaller guys. (Having said the many have turned to spread betting the FTSE it seems). Back to the original discussion does any one fine the DAX pretty schizoid? Here's what I mean you can trade it for a while (maybe months) and it plays pretty 'technical' respecting levels moving in a reasonably ordered fashion. When it trends and corrects, very tradeable. At others it just turns 'maniac' shooting all over the place, you have the massive spikes we talked of earlier it will congest in a tight narrow band and then just 'plunge' or surge, corrections are either non existent or it will actually correct 105% in seconds (catching stops). I dunno how to put it other than schizoid. Maybe its just me?

In a rising channel you would want to see rising volume as price moved from inside line (right) to outside line (left). You would want to see diminishing volume outside to inside. Put another way rising volume on impulses diminishing on corrections. At the outside line you might expect a volume surge (climax) however that might push price through, or turn it.(Channels frequently double in size). I used to like entering on those as if you are lucky you pretty much nail the trend change to the tick. Nowadays I try to be more sensible . At the inside you might expect 'no supply' or volume to tail right off. If you get a surge through at that point there is a good chance the trend line (inside line) is breaking. It might signify the start of a trend change or It also might drift through if the 'slope' of the trend is diminishing or as price goes sideways for a bit. The easy answer to your question however is 45,632. I have to say from the nature of your (many) questions you are still looking for simple answers where as you really just need to do the work.

At the end of the day that's all that matters! If one method 'works' for you then chances are so will the others. To be quite honest despite seeing potential value in these sorts of indicators I find that they end up confusing me as often as not and I end up going back to plain old price and volume.

not having a go at you but those two statements are not synonymous . If you are happy to use uptick/downtick as an aproximation that is great. Having run lots of these types of studies alongside each other I am inclined to agree with you:thumbs up:. However there is a difference between uptick/downtick and volume @ bid and ask they do produce similar but different studies. Actually watching the differences is interesting in its own right. I hope you'll forgive me being pedantic but there is a lot of needless in this thread despite the task actually being pretty straightforward. For that reason it is important to be clear and accurate (which is why I posted the correction.....not to :boxing:)

No you can not. You also need all the bid and ask changes as tick streams too. I do routinely build stuff from 1 tick files but you need to make assumptions about the order book and inside bid and ask unless you have recorded the order book (previously in real time) yourself. Neoticker is the only package I am aware of that does this (and that needs a kludge). Apparently ninja version 7 will have this feature too. The problem is not the trades (the tick stream) it's the order book changes. How do you know a tick is at the bid or the ask? You can not know though you can guess if its an uptick its at ask and a downtick is at bid. Best bid and best ask areavailable real time but not historically and there is the problem.

Sounds like you may have a badly written indicator aaa. You could remove them and add them back one at a time to discover which.

Resistance is seldom at a 'line' it is usually an area x to y. Of course x and y are somewhat 'fuzzy'. This is before you even start to determine what constitutes a break out. if you are generous on your zone then you can probably be tight on your break out.

You might find Jperls threads trading with market statistics intresting. This is discussed there. He rather interestingly uses the skew of the distribution to determine trend. You can get by just reading the first post in each thread if they look a bit daunting . Time based profiles and TPO's have no real statistical significance but may (or may not) be OK as an heuristic method. having said that I think they probably have become 'self fulfilling' which can add further value (if you will pardon the pun!)

There is no way to do it on historical data. It has to run live or run live and save data into its own file/database. You could of course make an approximation based on assumptions, this can work well. Rather than repeat myself http://www.traderslaboratory.com/forums/56/volume-splitter-5824-3.html#post63931 btw dont use [1] use if newvalue - oldvalue > block do something. oldvalue = newvalue

I agree it is easy but you don't need a one tick chart, any time frame will do. You just compare the total volume 'now' with the total volume last tick. Just take any one of the numerous delta/buy sell pressure indicators and they do this. Dunno where all the confusion is coming from it just needs a single conditional to compare size before its accumulated. Don't have much time right now (seem mad busy for some reason) or I'd post it.

The way most applications get volume for this print is by subtracting the volume last tick from volume this tick.

One thing I noticed with the DAX that there seems to be some pretty big block trades that are presumably negotiated off the exchange and then reported as a block. This seems particularly prevalent as you approach expiration. That skews things too and is essentially liquidity that is not accessible.

You are gonna have to look at the manual and or examples as it is not easy language as most know it.

Gotta smile it might not be a design flaw but its surely a design limitation

It's still relatively thin. At times (even prime time) the order book just goes beserk. I haven't traded it for a while but 5 points slippage can easily occur. I am trying to remember the worse slippage I had....can't off hand. Do you trade it monad? It can be brutal and whoever advised the inexperienced to stay away was spot on.

Or use volume profiles and slice and dice how you like.

Look at NinjaTrader through AMP or Mirus. Free to papertrade.

Whilst I am pretty competent at EL the strategy stuff (buying and selling) I have not really used. There are thousands of bits of code scattered round the inter web that you will probably be able to learn from. Tams also kindly posted some resources somewhere on Traders Lab. something like Var: StartingPrice(0); if date > date [1] then StartingPrice = close; if c <= StartingPrice-35 then ........buy stuff... If the syntax of IG is different you might have to mess with stuff.

Douglas is a good read. Also check Jperls threads "trading with market statistics" in the market profile section here. It's excellent.