Dashboard TZ | Page 504 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

MagicGirl dashboard 221 replies

Help with dashboard style indicator 2 replies

Dashboard indicators 4 replies

Dashboard Indicator Required 1 reply

Dashboard/Grid in lower indictor window 6 replies

Attachments: Dashboard TZ
Exit Attachments

Dashboard TZ

  • #10,061
  • May 15, 2022 2:57am May 15, 2022 2:57am
Quoting braintheboss
DD at min possible ( only -114p ). Net pips not decreased and lot size is quite bigger then profit had a huge increase. Anyway in future builds I will sacrifice some DD for get better net pips performance {image}
lover the DD
if system profitable with low DD, investors will line up.
WaBaS Where are Buyers and Sellers
  • #10,062
  • May 15, 2022 2:58am May 15, 2022 2:58am
TBH, its too good to be true, that a single automated system works for all pairs.
however your results sounds ok

lets see the forward test results
WaBaS Where are Buyers and Sellers
  • #10,063
  • May 15, 2022 3:27am May 15, 2022 3:27am
Quoting mr.brown
TBH, its too good to be true, that a single automated system works for all pairs. however your results sounds ok lets see the forward test results
I understand it can look tricky report but really it have logic. Evidently numbers without explanation can look smoke. I don't want write a technical post with explanation because most of members in this forum are not system builders and then is a useless information for them. But i will do a small explanation why DD looks awesome.

When you use a basket you have an effect in DD peaks. Current trades soft new trades effect. Its an effect everybody knows but nobody can measure because mt4 haven't tools for that. Knowing this is easy control how much DD you want have. Now explanation about why i told i will sacrifice DD for get more pips and how this affects profit:

1) Current report uses long term trades. This trades can be many weeks opened ( if they are profitable ). All this profit covers new trades DD. This is easy to see in month reports. Effect of closed trades too. You have like a batch of closes every month. This batch are previous month trades. When these trades are closed, all soft effect disappear and DD start to be present. You can see this in last month.

2) Second way uses short term trades. Soft effect is low because trades live not allow have multiple trades opened at same time. You get more closes and more pips but DD is bigger ( x5/x10 tmes ). Then if you don't want pass DD limit then you have use smaller lot.

Last thing is really not all pairs are profitable at current stage. Taking last report only 20 pairs are profitable. But this is because rules are still in a early stage. You have think strategy read with precision flat/trend moments and as is a basis behavior then will work well in all pairs.

As last comment this system base have precision as main goal. With finished rules it will be almost perfect system due it have low number of trades and high avg profit per trade with low DD. But this is a step by step path. We let time do the job
Try don't lose pants never...
  • #10,064
  • Edited 4:57am May 15, 2022 3:51am | Edited 4:57am
Another thing to take account is I'm using 20% DD limit. But that limit really is lower because I'm using same pip value for all pairs when i really use same lot size for all pairs in accounts tests.

This can look strange but when you are managing risk you need 2 things:

1) Fixed values
2) Break limits tolerance

Using same pip value for calculate risk while you use same lot in tests do the job. I added to reports another fixed risk pip value. Its an average of real pair pips value. Its around 80% of fixed pips value. This aproximation is quite closer to real but still higher that precise basket values.

Changing DD limit you can control risk easily. But i take in account how many trades make peak DD too. Then recommend lot size its linked mainly to this value
Try don't lose pants never...
  • #10,065
  • May 15, 2022 6:40am May 15, 2022 6:40am
Quote
'm using same pip value for all pairs when i really use same lot size for all pairs in accounts tests.
do you mean when lot size is 0.02, then pipValue will be 0.2 whatever the pair?
each day 10%,,,, https://www.forexfactory.com/thread/1371538
  • #10,066
  • May 15, 2022 7:16am May 15, 2022 7:16am
Quoting Macd-rsi
{quote} do you mean when lot size is 0.02, then pipValue will be 0.2 whatever the pair?
For have same equal pip value in all pairs you need set different lot size in every pair for compensate difference between pairs.

Report tool calculate profit and limits using this approach. But in my real tests I use same lot size because i haven't a tool for manage different pairs lot size. Then real profit is lower than report tool. That is reason i added 80% calculation too ( avg pip value). For have an aproximation of real profit
Try don't lose pants never...
  • #10,067
  • May 15, 2022 10:39am May 15, 2022 10:39am
Quoting braintheboss
{quote} For have same equal pip value in all pairs you need set different lot size in every pair for compensate difference between pairs. Report tool calculate profit and limits using this approach. But in my real tests I use same lot size because i haven't a tool for manage different pairs lot size. Then real profit is lower than report tool. That is reason i added 80% calculation too ( avg pip value). For have an aproximation of real profit
assume i take long position on all famous 28 pairs, each pair 1 lot,
does pipValue in this case is exactly 280 USD, or below 280 USD or above 280 USD?

just math issue
each day 10%,,,, https://www.forexfactory.com/thread/1371538
  • #10,068
  • May 15, 2022 12:34pm May 15, 2022 12:34pm
Quoting Macd-rsi
{quote} assume i take long position on all famous 28 pairs, each pair 1 lot, does pipValue in this case is exactly 280 USD, or below 280 USD or above 280 USD? just math issue
In real market it will be less 280. In my report tool it will be 280 because assume every pair have lot size for have 10$ per pip. Report with 80% it will be less 280 but higher than real market profit
Try don't lose pants never...
  • #10,069
  • May 17, 2022 11:34am May 17, 2022 11:34am
I think idea is clear. Get full trends and get full range areas in flat moments. Still far to work as expected
Attached Image (click to enlarge)
Click to Enlarge Name: 2022-05-17.png Size: 34 KB
Try don't lose pants never...
  • #10,070
  • May 18, 2022 12:12pm May 18, 2022 12:12pm
Max DD peak in current test is 600€, then its slightly more 10% using 5K account ( my reference balance )
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 32 KB
Closes today ( last build sacrificing DD for get shorter length trades ).
Attached Image (click to enlarge)
Click to Enlarge Name: download.png Size: 15 KB

As now we are entering in summer, its a good test for check how performs in flat markets
Try don't lose pants never...
  • #10,071
  • May 20, 2022 4:55am May 20, 2022 4:55am
Quoting braintheboss
Max DD peak in current test is 600€, then its slightly more 10% using 5K account ( my reference balance ) {image}Closes today ( last build sacrificing DD for get shorter length trades ). {image} As now we are entering in summer, its a good test for check how performs in flat markets
hi, what is your trading timeframe and how long have you been back tested? if you dont mind
WaBaS Where are Buyers and Sellers
  • #10,072
  • Edited 7:00am May 20, 2022 5:14am | Edited 7:00am
Quoting mr.brown
{quote} hi, what is your trading timeframe and how long have you been back tested? if you dont mind
H1. Current system is TE length ( around 1 month ). But I'm replacing builds every 2/3 days with newer.

Then main goal is test system cycle tolerance while I'm finishing system rules. As every new build fix some algorithm mistakes and maybe add new problems, its good for check what happens when you don't let system warm and new build trades not match with previous build cycle. This is worst situation for system. If its profitable then letting run long time it will fit with reports performance. Then i will get confirmation all reports are safe.

New build is running since yesterday morning. It have a rule with overlapping problem. You can check CHFJPY trades taken yesterday. I had 3 in 3 H1 consecutive bars. This should not happen. Anyway not look this kind of mistakes affects so much. And these are good news because means system have good tolerance to trading errors.

Trading mistakes tolerance is very important in auto systems because they are more tend to have longer losers in a row and worst avg per trade than manual trading.
Try don't lose pants never...
  • #10,073
  • May 20, 2022 7:08am May 20, 2022 7:08am
Quoting braintheboss
{quote} H1. Current system is TE length ( around 1 month ). But I'm replacing builds every 2/3 days with newer. Then main goal is test system cycle tolerance while I'm finishing system rules. As every new build fix some algorithm mistakes and maybe add new problems, its good for check what happens when you don't let system warm and new build trades not match with previous build cycle. This is worst situation for system. If its profitable then letting run long time it will fit with reports performance. Then i will get confirmation all reports are...
i mean the length of backtest if you had any?
WaBaS Where are Buyers and Sellers
  • #10,074
  • May 20, 2022 7:18am May 20, 2022 7:18am
Quoting mr.brown
{quote} i mean the length of backtest if you had any?
2000 H1 bars ( around 4 months ). I only check 2000 bars because mt4 don't handle well many bars...
Try don't lose pants never...
  • #10,075
  • May 20, 2022 10:06am May 20, 2022 10:06am
Quoting braintheboss
{quote} 2000 H1 bars ( around 4 months ). I only check 2000 bars because mt4 don't handle well many bars...
I see. forward test is important.
one thing. while you think you are improving the code you might ruin it. experienced
change one thing at a time

my2cents
WaBaS Where are Buyers and Sellers
  • #10,076
  • Edited 11:11am May 20, 2022 10:32am | Edited 11:11am
Quoting mr.brown
{quote} I see. forward test is important. one thing. while you think you are improving the code you might ruin it. experienced change one thing at a time my2cents
At this moment I'm only trying have a code works in all pairs without have use complex rules. My idea is find one and let it run in real making some money while i take my time for finish system.

But checking reports of different builds I saw if I want use simple rules then ideal way is use different build rules in every pair using always best. Its a fast way for get something working without loss much time. I can't pass an edge in any build using only one build rules. When i fix some pairs, others performing well go to very bad.

Anyway, as rules are modular, i think i can use this mixed approach. Release build it will have an AI then it will be all rules code rewritten. And that is a few months task. I think is more convenient use something making some profit even it have flaws that wait perfect release build and waste a lot of months without get any money.

We will see...
Try don't lose pants never...
  • #10,077
  • May 26, 2022 4:04am May 26, 2022 4:04am
As now real time test have a full month I can compare with backtest reports reliability. Look real time performance fits with reports.

TE shows why with a low win ratio is having outstanding performance. Closed trades are 60% profit ( as backtest report make calculations, 5k account, 3k profit)
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Screenshot_20220525-145138.png Size: 514 KB
Click to Enlarge Name: Screenshot_20220526-105905.png Size: 180 KB
Try don't lose pants never...
  • #10,078
  • May 26, 2022 11:29am May 26, 2022 11:29am
We think you change one line in a rule and you want check results. If you haven't backtest you need wait 4 months. If you have backtest less one second. Im saving my life and my nervers

Before
Attached Image (click to enlarge)
Click to Enlarge Name: Captura de pantalla 2022-05-26 182740.jpg Size: 16 KB

After
Attached Image (click to enlarge)
Click to Enlarge Name: Captura de pantalla 2022-05-26 182711.jpg Size: 17 KB
Try don't lose pants never...
  • #10,079
  • Jun 3, 2022 12:42pm Jun 3, 2022 12:42pm
First build with dynamic parameters. Both are stock algorithm without filters. Really filters are useless since rules are better now.

Most significative report difference is all pairs are profitable comparing without dynamic parameters ( 60% ). Perfomance is a big jump too ( pips and profit )

No dynamic parameters

Attached Image (click to enlarge)
Click to Enlarge Name: Captura de pantalla 2022-06-03 193410.jpg Size: 87 KB

Only 1 dynamic parameter ( its main levels parameter then it have huge effect in trades )
Attached Image (click to enlarge)
Click to Enlarge Name: Captura de pantalla 2022-06-03 193439.jpg Size: 92 KB
Try don't lose pants never...
  • #10,080
  • Jun 3, 2022 1:05pm Jun 3, 2022 1:05pm
A pair example. Only improvement is win pips, but total have a big improvement.

Dynamic parameters is similar as when you use metatrader backtester with parameters range. But its very fast comparing MT. It allows AI as pattern search too
Attached Image(s) (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 2 KB
Click to Enlarge Name: screenshot.png Size: 2 KB
Try don't lose pants never...
0 traders viewing now