BlowFish

Woops way to completely balls up :crap: I only changed a couple of lines and managed to get an additional end statement and undefined variable (block) as you guys pointed out. It's got to the stage where it needs a tidy up (as well as comments) but this should at least run inputs: UpColor(darkgreen), DownColor(red), DeltaBar(1), MaxBlock(9999), MinBlock(0), ResetDeltaEachBar(0); variables: MyVol(0), Block(0), color(yellow), intrabarpersist MyCurrentBar(0), intrabarpersist VolumeAtBid(0), intrabarpersist VolumeAtAsk(0), intrabarpersist BAVolRatio(0), intrabarpersist VolTmp(0), intrabarpersist Delta (0), intrabarpersist DeltaH (0), intrabarpersist DeltaL (0), intrabarpersist DeltaO (0); if LastBarOnChart then begin MyVol = Iff(BarType < 2, Ticks, Volume); if CurrentBar > MyCurrentBar then begin VolumeAtBid = 0; VolumeAtAsk = 0; BAVolRatio = 0; VolTmp = 0; MyCurrentBar = CurrentBar; if ResetDeltaEachbar = 1 then Delta =0; DeltaO = Delta; DeltaH = Delta; DeltaL = Delta; end; Block = Myvol - VolTmp; if (Block >= MinBlock) and (Block <= MaxBlock) then if Close <= InsideBid then Delta = Delta - MyVol + VolTmp else if Close >= InsideAsk then Delta = Delta + MyVol - VolTmp ; VolTmp = MyVol ; end; DeltaH = maxlist(DeltaH, Delta); DeltaL = minlist(DeltaL, Delta); if Delta <= 0 then color = DownColor else color = UpColor; plot1(DeltaO, "DO"); Plot2(DeltaH, "DH"); Plot3(DeltaL, "DL"); plot4(Delta, "DC");

Thanks I'll check them out..... Edit. Ahh they look like a 'proper' broker that might explain it.

Sure, however the chart you posted certainly looks like the winners would offset the losers. The first red box you highlight doesn't even seem to have a cross. It looks about 50 50 win/loss with maybe a scratch and a couple of winners that are of (comparatively) greater magnitude. The reason I ask this is I wonder what your expectations are from such an approach? The characteristics of 'traditional trend following systems' (of which MA crosses and channel break outs are) is that you might have quite a few losers which are offset by the fewer but larger winners. If you start filtering you can change the parameters but you will filter good trades too.

I can't see a problem with the chart you posted. With suitable money management that looks like a highly profitable day to me?

Put simply the 'fixed spread brokers' are actually bookies. What is worse they are taking the other side of your bet and setting the price you bet at. Like the bucket shops of yesteryear. Another alternative on a real regulated exchange is the new currency futures mini contracts.

The ACD range 'formula' appeals to me (though never used it as presented). of course after a wide range day that closes at an extreme the zone is going to be way to big to trade as Fisher proposes (maybe he has rules for that I only have a passing familiarity with his work). Having said that the zone between floor pivot (HLC/3) and the mid point HL/2 sounds like a plausible balance area.

Whats on your radar this week KP?

Hi essentially the first is an oscillator. The second shows cumulative 'order flow' for the day. Normally I would add comments at the top of the indicator. In my defence it started as a quick and dirty to prove the concept I will tidy them up and then post them both in the indicator section.

Great Idea too!!! ...

To avoid confusion here is the code with a max block size and min block size rather than a filter. Much as bakrob said (I edited it just before I got to his post!!) It should work but I didn't compile it as the indicator is running live on my charts right now. inputs: UpColor(darkgreen), DownColor(red), DeltaBar(1), MaxBlock(9999), MinBlock(0), ResetDeltaEachBar(0); variables: MyVol(0), color(yellow), intrabarpersist MyCurrentBar(0), intrabarpersist VolumeAtBid(0), intrabarpersist VolumeAtAsk(0), intrabarpersist BAVolRatio(0), intrabarpersist VolTmp(0), intrabarpersist Delta (0), intrabarpersist DeltaH (0), intrabarpersist DeltaL (0), intrabarpersist DeltaO (0); if LastBarOnChart then begin MyVol = Iff(BarType < 2, Ticks, Volume); if CurrentBar > MyCurrentBar then begin VolumeAtBid = 0; VolumeAtAsk = 0; BAVolRatio = 0; VolTmp = 0; MyCurrentBar = CurrentBar; if ResetDeltaEachbar = 1 then Delta =0; DeltaO = Delta; DeltaH = Delta; DeltaL = Delta; end; if (Block >= MinBlock) and (Block <= MaxBlock) then if Close <= InsideBid then Delta = Delta - MyVol + VolTmp else if Close >= InsideAsk then Delta = Delta + MyVol - VolTmp ; VolTmp = MyVol ; end; end ; DeltaH = maxlist(DeltaH, Delta); DeltaL = minlist(DeltaL, Delta); if Delta <= 0 then color = DownColor else color = UpColor; plot1(DeltaO, "DO"); Plot2(DeltaH, "DH"); Plot3(DeltaL, "DL"); plot4(Delta, "DC");

Tasuki when I first introduced the 'fix' I mentioned as coded it was actually a small block filter. You should read a bit more carefully volume less than clock filter makes it small block filter. volume > than block filter makes it a large block filter. Anyway thanks for the charts.

Mjackson is there a MT4 broker that provides half way decent tick volume? Those that I have looked at seem only to show there own customers ticks? Thanks.

Personally I split oscillators into two types bounded or unbounded (which I think is probably the norm). As Tams says they should all be centred. You can usually make an unbounded oscillator bounded by some sort of normalisation. CCI is another example of an unbounded oscillator. Unbounded oscillators still don't reach infinity they are 'bounded' by statistics.

Did for me too, kind of went slow motion. Could still understand it though.

Good for you ....

Most surely there are times that it will be true others not. Different participants have widely different agendas and modus operandi. Hell, some participants are not even profit motivated (not as you or I would think of 'profit'). I would recommend any one who is interested to check out a decent book on Market Microstructure Harris writes 650 odd pages on the subject it is quite a deep and involved subject and hard to do justice to in a couple of forum posts. Really we would need to go through each type of participant and look at how they operate. Back to the indicator (the 2nd one which is going to need a name I guess....how about net order flow? cumulative net order flow is a bit of a mouthful) again, today, it has given some pretty good insight to the days goings on. I am tempted to incorporate it into my trading. Trouble is to do any testing is going to take an eternity as it only runs live. I'd also like to look at longer accumulations (like a week) of data too. Again a pain really with the real time restriction. It's about time charting applications save order book information and generated update ticks on all data streams.

Not at all. I say please keep posting. It is not an approach that appeals to me but let people judge the method on it's own merits. If people want to judge you (rather than your work), you seem like an honest and open chap as Brown pointed out. Brown is a candlestick fan. Nisson the self proclaimed "introducer of candlesticks to the west" is not even a trader but many would argue (including Brown I am sure) that his writing has valuable insights into candlestick charting.

That is because they are not used by Crabel (as far as I recall). XuanXue introduced another idea which looks to be based on Fischers work.

I wondered why you chose SDS as the vehicle for your trade?

Annotating chart images in MSPaint is a pain. If you don't have an app for the job (snaggit the screen capture software is pretty good) this might suit you:- http://www.sumopaint.com/web/#/home/ It is an online flash app so no junk on your PC. Just open directly from the url of the image you want to draw or annotate (or load it from your PC).

Glad you like it. I have to say it's a while since I have had it on a chart but was rather surprised by how revealing it was when I ran it yesterday. I do have an issue with it intellectually and that is the assumption that informed traders ('smart money') will buy or sell aggressively (hit the bid or ask). That's just not the case for certain types of traders or under certain market conditions. Having said that it does seem to be a decent guage of order flow.

Indeed, contribution and intent. I too favour laissez-faire approach until the intent becomes clear through what is contributed. (or put another way I agree with DB)

Absolutely! As for your second point that would seem sensible to me too. Mind you, if they (the longs) have all been sold then there won't be anything in the portfolio anyway. One way of reading this is that the shorts will be posted in the week(as they occur) and then the updated portfolio (with any new shorts) will be posted at the weekend. We shall no doubt see.

A couple of pictures of the absolute delta no splitting or filtering so kind of off topic. Very interesting though (imho).

How about using one of Ehlers 'zero lag' filters that would probably produce good results?