Better ADL | Page 2 | Metals Mine

Options

Go to
Permalinks
Print Thread

Similar Threads

A better Editor for mql is available. 100x better than MetaEditor 20 replies

looking for Advance Decline Indicator ADL 13 replies

Broker needs to do better, much better 4 replies

Better momentum and better Sine wave indicator 0 replies

Is Better still better? 1 reply

Attachments: Better ADL
Exit Attachments

Better ADL

  • #21
  • Sep 14, 2015 7:39am Sep 14, 2015 7:39am
Quoting kprsa
{quote} Hey NT, Thanks for starting the thread and sharing your ideas in the first place! If you don't mind, I would invest a bit of time to code your ideas for ADL for MT4 (as far as I can understand them) and share the indicators here? Cheers, k
Absolutely k, go for it! I was hoping that you or one of the other great MT4 coders would convert it because, as you know, I only program NinjaTrader stuff, which is a fat lot of use to most people on here.

The start of the week is really busy for me so I won't be able to post any code until later, after FOMC probably. But if you code something from what I've specified so far then by all means go for it. Currently I'm wondering what to do about gap bars. Should their spread include the gap or not? I suppose they should because of all the orders that were filled "in the gap", and the volume that is created as a result. But there aren't many gaps in FX anyway, it's more of an issue for gappy instruments like the Nikkei.

Thanks and good luck.
You reap what you sow.
  • #22
  • Sep 14, 2015 8:37am Sep 14, 2015 8:37am
One question about the formula. The two bars below will be given the same ratio -1/6. Is it wanted?
Attached Image
No greed. No fear. Just maths.
  • #23
  • Sep 14, 2015 5:10pm Sep 14, 2015 5:10pm
Quoting NorthTrader
{quote} Thanks, it must be a good indi if you've been using it all this time. I read on that thread you linked to that the indi uses no price information whatsover, only volume. Wow! Mine uses price and volume, and kprsa's uses only price, so it's going to be great comparing them all on the same chart one day to see which one gives the best signals I'll PM you when I need the source thanks, but it won't be just yet.
it uses price also, here is a example from the code so you can see how it works, i think its simular to your idie,

for(i=count;i>=0;i--)
{
if(i==Bars-1) OBVbuff[i] = iVolume(NULL,TF,i);
else
{
closeNow = iClose(NULL,TF,i);
closePrev = iClose(NULL,TF,i+1);
if(closeNow==closePrev) OBVbuff[i] = OBVbuff[i+1];
else
{
candleHeight = iHigh(NULL,TF,i)-iLow(NULL,TF,i);
if(candleHeight>0.0)
{
if(closeNow<closePrev)
{
val = (iOpen(NULL,TF,i)-closeNow)/candleHeight;
OBVbuff[i] = OBVbuff[i+1]-iVolume(NULL,TF,i)*val;
}
else
{
val = (closeNow-iOpen(NULL,TF,i))/candleHeight;
OBVbuff[i] = OBVbuff[i+1]+iVolume(NULL,TF,i)*val;
}
}
else
OBVbuff[i] = OBVbuff[i+1];
}
}
}
  • #24
  • Edited 6:44pm Sep 14, 2015 6:00pm | Edited 6:44pm
Quoting PipMeUp
One question about the formula. The two bars below will be given the same ratio -1/6. Is it wanted? {image}
Yes, -1/6 is fine for both of these bars. In the context of VSA, these kind of bars can be very meaningful, depending on their volume and the "background volume" (volume of nearby bars) at the time. But in the context of a simple indi like ADL or BADL (mine), I think it's OK to assume that the wicks represent a balance of buying and selling pressure, and the bit that's left (the body) represents the unbalance. So -1/6, which is a fairly small number despite the huge volume that might have occurred within the wicks, is a good compromise.

Thanks for the question PipMeUp, and great graphics!

Edit: crossed out ADL above, because the original ADL produces a ratio of +2/3 for the first bar and -1 for the second, which isn't ignoring the wicks. Pretty weird that the ratios are different though for inverse bars like this. Again I think that my calculation, producing an initial ratio of -1/6 for both, is better in this case.
You reap what you sow.
  • #25
  • Sep 14, 2015 7:00pm Sep 14, 2015 7:00pm
Quoting slicken
it uses price also, here is a example from the code so you can see how it works, i think its simular to your idie
Thanks slicken, and yes you're right. It calcuates an intial ratio in the same way as mine: (Close - Open) / (High - Low). I must have misread about it not using price, or maybe it was modifed after that. It doesn't apply the mean spread of the previous n bars though, unless there's more code after that, so it's still different from mine.
You reap what you sow.
  • #26
  • Edited 2:51pm Sep 15, 2015 2:38pm | Edited 2:51pm
Quoting kprsa
{quote} The first one I have coded (with binary additions of the ADL) before, now I made also the non-binary one (adding the height of the zone). It looks to me that the first one is kinda better for trading -> by divergences, like standard ADL. The second one does not distinguish itself from the price that much... Anyways, here they are. {file} {file}
I've taken a look at the 1st indicator, I love it, it looks like it'd be good for filtering out range, I want to test some idea's with it but for some strange reason the indicator buffer returns no value when I try to get it for an EA. Any help will be appreciated.

I use an EA builder, which gets the values from the indicator buffers, my coding skills are nothing to write home about.
  • #27
  • Sep 15, 2015 2:57pm Sep 15, 2015 2:57pm
Quoting sinqua
{quote} I've taken a look at the 1st indicator, I love it, it looks like it'd be good for filtering out range, I want to test some idea's with it but for some strange reason the indicator buffer returns no value when I try to get it for an EA. Any help will be appreciated. I use an EA builder, which gets the values from the indicator buffers, my coding skills are nothing to write home about.
Hi sinqua, can you just write the iCustom() call line that your ea uses? Then I can try to detect the problem.
Happy you find it useful.
Cheers,
k
  • #28
  • Sep 15, 2015 4:35pm Sep 15, 2015 4:35pm
Quoting kprsa
{quote} share the indicators here
Version 1 of Better ADL as proposed by NorthTrader:

1. No fractionalization -> I am not sure how to achieve that in MT4 or if it is possible at all. The idea is to calculate the ADL on low TF bars (~1 min) and show it on higher tfs (for example 1 h, 1 day, etc).

2. Could contain bugs due to my misunderstanding of NT's idea or plain old lack of skill. So let me know if there are some.

Also good to compare with NT's upcoming screenshots to see if there would be significant differences because of 1) and 2). Also, if you know how to deal with 1), feel free to resolve it, and share your solution.

Cheers,
k


Attached File(s)
File Type: mq4 BADL_v1.mq4   3 KB | 854 downloads
  • #29
  • Sep 15, 2015 5:02pm Sep 15, 2015 5:02pm
Quoting kprsa
{quote} Hi sinqua, can you just write the iCustom() call line that your ea uses? Then I can try to detect the problem. Happy you find it useful. Cheers, k
Here you go:

double _iCustom_6_ObjText() {
double retval=iCustom(CurrentSymbol(),CurrentTimeframe(),"ADL_transient" ,0,0+IndicatorMoreShift());
//if (false && retval>=EMPTY_VALUE) {retval=0;}
SetLastIndicatorData(retval,CurrentSymbol(),CurrentTimeframe(),0+IndicatorMoreShift());
  • #30
  • Sep 16, 2015 12:45am Sep 16, 2015 12:45am
Here is my initial version of the BADL indicator for NinjaTrader, including source code. Sorry for the delay, it was caused by:

1. Making the mean spread calculation more efficient (subtracting from and adding to a running total rather than looping through the previous n bars every time).
2. Adding funtionality to handle gap bars, if required. Includes hidden gaps.
3. Commenting the code so it can be understood by MT coders more easily.

There are now three parameters that can be set by the user:

Bar calculation period
Period of each bar (in minutes) on which the calculation is performed. 0 = same as chart bars.
Default = 1.

Note that you can't set the period to less than 1 minute because my broker doesn't provide historical second or tick data to NinjaTrader. If you want it to work with tick data then please modify the code yourself and post it here.

Spread period
Number of previous bars to use for the mean spread calculation. Minimum = 1.
Default = 21.

Include gap spreads
True = use previous bar's Close for current bar's Open (include gap spreads).
False = always use current bar's Open (ignore gap spreads).
Default = False.


Please note that it's designed to work in "Calculate on bar close" mode, like the original ADL. In other words, it doesn't refesh the current bar's value as the bar is being built (tick by tick). I've noticed that it does refresh tick by tick if "Calculate on bar close" is set to false, but I haven't tested it in this mode much so please treat with caution.

Ran out of time now, so screen shots of it in action will have to wait until tomorrow (hopefully). If you have any questions about the code, please ask.
Attached File(s)
File Type: zip BADL_V1.zip   3 KB | 749 downloads
You reap what you sow.
  • #31
  • Sep 16, 2015 9:42pm Sep 16, 2015 9:42pm
EUR.USD Daily

Here is a chart of the CME 6E contract (EUR.USD) showing the original ADL and two versions of the Better ADL (BADL). The bottom red line is original ADL. The middle green line is BADL calulating on the daily bars. The top blue line is BADL calculating on 1 minute bars within the daily bars. Mean spread period is set to 21 bars - this is just a guess and may not be the optimal period. Include gap spreads is set to false - if set to true the line is almost the same because there aren't many gaps in EUR.USD.

I've drawn a few yellow trend lines on the chart to show the difference between the three ADLs and to highlight divergences between the ADLs and price.

As you can see, the blue BADL, which works on the 1 min bars, seems to produce the best signals in terms of price trend confirmation. If you'd acted on the divergences between price and green BADL or red ADL at these pivot points then you'd have been wrong. And both BADLs confirm that the overall trend is up over this wide range, whereas the original ADL seems to suggest that it's been heading down since the middle of June. Personally I'm still holding a small long on EU and the BADL gives me confidence that there's still more upside to come. Let's see!

I'll post more charts as time permits.
Attached Image (click to enlarge)
Click to Enlarge Name: 6E 12-15 (Daily) 14_02_2015 - 18_09_2015.png Size: 77 KB
You reap what you sow.
  • #32
  • Sep 17, 2015 12:35am Sep 17, 2015 12:35am
EUR.USD Daily Reversal

So does the BADL ever diverge from price to predict a reversal? Here is the same 6E daily chart but shifted to the left to show the start of the reversal from 1.05.

You can see that the first yellow line is much steeper on the price panel than on the BADL panel. I suppose it could be argued, with the benefit of hindsight, that this indicated that selling had run it's course and buyers were stepping in.

However, the spike upwards on the BADL (yellow ellipse) might have suggested to me at the time that the bottom was already in and the reversal had begun. There was certianly some serious buying pressure in that area. But I would have been wrong, because price fell after that another 900 pips.

It could also be argued that the spike on the BADL was a sign of things to come, and in fact the BADL didn't break the low of the spike until price broke 1.10. So there probably was some accumulation going on there.

Anyway, no indicator is perfect and of course false signals will occur from time to time, especially when discretion is required to interpret them (as in the case of divergence). But so far so good on the Daily I think

Next up: lower timeframes.
Attached Image (click to enlarge)
Click to Enlarge Name: 6E 12-15 (Daily) 14_10_2014 - 20_05_2015.png Size: 68 KB
You reap what you sow.
  • #33
  • Sep 17, 2015 8:12pm Sep 17, 2015 8:12pm
Quoting NorthTrader
EUR.USD Daily Reversal So does the BADL ever diverge from price to predict a reversal? Here is the same 6E daily chart but shifted to the left to show the start of the reversal from 1.05. You can see that the first yellow line is much steeper on the price panel than on the BADL panel. I suppose it could be argued, with the benefit of hindsight, that this indicated that selling had run it's course and buyers were stepping in. However, the spike upwards on the BADL (yellow ellipse) might have suggested to me at the time that the bottom was already...
But the Daily (green line) didnt give the false call.
Maybe 1 minute is too much of a contrast to the daily?
Possibly try a 30 min bars within the Daily.

BTW great contribution NT!
What's it all about? It's all about money.
  • #34
  • Sep 17, 2015 9:08pm Sep 17, 2015 9:08pm
EUR.USD M240

The upmove following FOMC doesn't look so convincing on this scale (BADL blue line), so I'm tightening my stop to 1.1385 (spot price) just in case.
Attached Image (click to enlarge)
Click to Enlarge Name: 6E 12-15 (240 Min) 18_09_2015.png Size: 78 KB
You reap what you sow.
  • #35
  • Sep 17, 2015 9:19pm Sep 17, 2015 9:19pm
Quoting Innate
{quote} But the Daily (green line) didnt give the false call. Maybe 1 minute is too much of a contrast to the daily? Possibly try a 30 min bars within the Daily. BTW great contribution NT!
Thanks Innate. Yes, I need to play around with the parameters, maybe this weekend, in order to find the best combination. But I can't help thinking that smaller is better for the bar calculation period, even on the daily
You reap what you sow.
  • #36
  • Edited 9:47pm Sep 17, 2015 9:33pm | Edited 9:47pm
Nikkei Emini M240

Time for equities to rise? (yellow arrow)
Attached Image (click to enlarge)
Click to Enlarge Name: N225M ##-## (240 Min) 18_09_2015.png Size: 70 KB

By the way, the blue line is Include gap spreads = true (probably better for the Nikkei), and the green line is Include gap spreads = false. Actually it doesn't seem to make a lot of difference.
You reap what you sow.
  • #37
  • Sep 21, 2015 6:51am Sep 21, 2015 6:51am
EUR.USD M15

When you get to lower timeframes, the differences between Better ADL and Original ADL become less obvious. And I've found that a great way to use it, apart from looking for relatively rare divergences, is to gauge the relative strength of a move against its subsequent retrace. This works because the ADL is basically a momentum indicator.

For instance, let's assume that you're looking to enter when price nears the 50-period EMA (white line on the price panel). I'm using the blue BADL (top) for guidance. The green BADL and red ADL are only there for comparison. See chart below.

A. This is a classic "strong move / weak retrace", and is confirmed by the BADL. A great place to go short.

B. There isn't a convincing reason to go short again here according to the BADL. Wait for more confirmation that shorts are still in play.

C. This is the confirmation, so try to enter on the retrace. Easier said that done! Could well be a "no fill".

D. Not a good place to short due to the shift up on the BADL.

E. A good place to long due to the strong move up and relatively weak retrace, both on price and the BADL.

F. Not compelling, wait for confirmation.

Is it easy trading like this? Not really. A great deal of discretion is required, and a fair bit of luck. But as indicators go, I feel I could do a lot worse

Next I want to try to code kprsa's Mid TZ (no volume) indicator in NT7 and see how it compares with the BADL. It would be nice to not have to rely on volume data. But how effective are the signals? I'll post some charts when I've finished.
Attached Image (click to enlarge)
Click to Enlarge Name: 6E 12-15 (15 Min) 17_09_2015.png Size: 92 KB
You reap what you sow.
  • #38
  • Sep 22, 2015 5:44am Sep 22, 2015 5:44am
ADL transient

So, here is the same chart as the last post but with kprsa's ADL_transient indicator added. I've removed one of the BADLs for clarity. The lines are:

Blue (top): BADL calculating on 1 minute bars
Red (middle): original ADL
Green (bottom): ADL_transient calculating on 1 minute bars, with +1 for Up Mid TZ and -1 for Down Mid TZ (both h=1)

Attached Image (click to enlarge)
Click to Enlarge Name: 6E 12-15 (15 Min) 17_09_2015 TADL.png Size: 84 KB


And here it is in action on larger timeframes:

Attached Image (click to enlarge)
Click to Enlarge Name: 6E 12-15 (240 Min) 22_09_2015 TADL.png Size: 89 KB


Attached Image (click to enlarge)
Click to Enlarge Name: 6E 12-15 (Daily) 08_01_2015 - 23_09_2015 TADL.png Size: 78 KB


What do you think? Is this something that can be developed further? I think it's pretty effective considering it doesn't use volume at all. Thanks k

(I'll post the NT7 version of the indi soon)
You reap what you sow.
  • #39
  • Sep 23, 2015 12:20am Sep 23, 2015 12:20am
Transient Zone ADL

Here is the NinjaTrader version of kprsa's indicator. I've renamed it TZADL (Transient Zone ADL):
Attached File(s)
File Type: zip TZADL_V1.zip   4 KB | 749 downloads

There are three parameters that can be set by the user:

Bar calculation period
Period of each bar (in minutes) on which the calculation is performed. 0 = same as chart bars.
Default = 1.

TZ Width
Width of transient zone (in bars) each side of the mid-bar. Values > 1 will measure stronger TZs, but will increase lag.
Default = 1.

Use TZ height
True = add or subtract the TZ's height (in pips) to the running total.
False = always add or subtract 1 for each TZ (ignore TZ height).
Default = False.


Please note:
a) like the BADL and original ADL, it's designed to work in "Calculate on bar close" mode.
b) a TZ must be at least two pips high in order to contain at least one transient price, so one-pip TZs are ignored.

I'll post a couple of charts in the next post.
You reap what you sow.
  • #40
  • Sep 23, 2015 12:25am Sep 23, 2015 12:25am
On these charts:

Blue (top): BADL calculating on 1 minute bars
Green (middle): TZADL calculating on 1 minute bars, using +1 for Mid Up TZ and -1 for Mid Down TZ
Red (bottom): TZADL calculating on 1 minute bars, using TZ heights (in pips)

Both TZADLs use TZ width = 1 (h=1).

It seems to me that the bottom TZADL matches the BADL more than the middle one.

EUR.USD M15
Attached Image (click to enlarge)
Click to Enlarge Name: 6E 12-15 (15 Min) 17_09_2015 TZADL.png Size: 74 KB

EUR.USD Daily
Attached Image (click to enlarge)
Click to Enlarge Name: 6E 12-15 (Daily) 09_01_2015 - 24_09_2015 TZADL.png Size: 76 KB
You reap what you sow.
0 traders viewing now