Disliked{quote} Im not using templates. Yes some pairs are ok others aren't. {image} {image} {image} {image}Ignored
(binned per thread starter’s request) MM (Money Maker) Detective
- #5,424
- Jul 31, 2019 10:45pm Jul 31, 2019 10:45pm
- Joined Jun 2012 | Status: Trader | 1,106 Posts
Some people say I'm lucky, but the more I practice the luckier I get
- #5,425
- Jul 31, 2019 10:48pm Jul 31, 2019 10:48pm
Everything's crystal-clear , totally respecting the original strategy and keeping it simple enough. ANS: @Viktory -- with all due respect, I disagree. Not only is it not crystal clear it also is not "simple enough". These only add complexity. I came across this saying that I have taped at the top of my monitor as a reminder: "Sometimes we fall in love with the complexities and justify their necessity"
Possibility to open charts from the Dashboard. ANS: NO WAY!! Too many open charts cause MT4 to become sluggish. Use a dashboard symbol changer instead.
For signal computations, just check the close price of every M1 candle, instead of every tick, in groups of 7 pairs after 7 at a time. ANS: The close price of EURUSD-M1 right now is 1.10741 -- and what am I suppose to do with that? @Viktory -- it's not about getting prices. And groups of '7' at a time? How did you come up with '7'? Do you know for a fact that '7' won't bog things down? What about '5'? What about "14'? Sorry, this one is ridiculous. Have you ever coded an indicator or EA -- and you make suggestions like this? Do you have any idea about MT4's single-threaded execution? Implications of using OnTick vs OnTimer? Come on -- this is why I can't spend time reading long, detailed lists of suggestions like this.
{quote} @Simontw -- the Dashboard and AllPairs EA run slow for me, as well. It's because with each tick (or timer) they loop through all pairs and compute for each one. If you are trading 28 pairs then its the equivalent to having 28 charts open. Try opening 28 charts on MT4... In the next release -- later today -- I've changed the way the EAs loop through all the pairs. Instead of doing all 28 at once it does 2 at a time. It takes around 30 seconds to touch all pairs. Doing it this way the EAs seem less likely to bog down MT4.
---
Flexible position size according to X $; example: 0.01 lots per every 1000 USD in the account. ANS: You know how much money you have in your account. The lot is configurable from the input settings. If you have $1000 then you can leave the default at 0.01. If you have $10000 you can set it to 0.1.
Automatic percentage equity/balance-based position size derived from the SL (dynamic lots changing according to the selected %risk and SL pips-distance from the entry), also considering the pip-value of each pair; example: for a $10,000 USD account, if I chose 1% and there’s a 100 pips SL for EURUSD, then the EA automatically calculates the corresponding 0.10 lots per position. ANS: Are you kidding? I've read this 5x and still don't understand. Plus -- where's the proof?
We could have a setting called "Equity (or Balance) % SL risk per trade", so if you place "1%" with a $10,000 USD account, and there’s an hypothetical 100 pips SL for EURUSD, automatically calculated by the setting "StopLoss fraction of ADR", then the EA automatically calculates the corresponding 0.10 lots per position; so if the 100 pips SL gets hit, the account loses $100 USD = 1%... That way, as the account gets bigger (using Balance or Equity), also the lot-sizing; and vice-versa. A pretty convenient automation, because we wouldn't need to be constantly calculating the lot-sizing ourselves. So it actually contributes to "risk-management", or "micro-management", whatever.
And something like that, would also allow us to do back-tests with compounding grown accounts by automatic lot sizing, as well as better gauging approximate possible Drawdowns for longer-term periods.
---
Finally, just to more settings for proper Money-Management automation, perfectly justified by my screenshots: https://www.forexfactory.com/showthr...7#post12412507 ANS: You upload 10 charts and announce "I think it's pretty evident:"? Come on, you can't be serious.
And again, here you have my previous examples of various cases with too many cluttered DOTS (entry signals) very close together: https://www.forexfactory.com/showthread.php?p=12412507#post12412507
If the EA traded like that, without limiting the MaxOpenTradesPerPair, we would have periods with very strong over-trading (too many open trades around the same level), hence increasing the risk-exposure of the account. So it would help us to prevent any single pair to fall into over-exposure.
MaxOpenTradesGlobal and MaxOpenTradesPerPair. ANS: MaxOpenTradesPerPair is already implemented. MaxOpenTradesGlobal is just yet-another-setting (YAS) that makes you feel in control -- where is proof that it would improve performance?
---
MinDistanceBetweenOrders (for the next farther one), with the option to either choose Pips or %ADR. ANS: Why limit it to just Pips or %ADR? Why not per NN time interval? Or every 3rd or 5th signal? Or, better yet, let's use those Entry signals to compute a fractal and use that to thin-out the number of Entries. I can think of even more. Thinning-out the number of signals can be achieved more simply by judicious use of TF, # pairs, increasing the 'Percet ADR' band offset from '100' to '120', increasing MinPercentAdrCup.
As you can clearly see on my attached screenshots, "MinDistanceBetweenOrders" (farther-away = higher for sells & lower for buys) helps to have better quality entries for continuous trades on the same pair, and also contributes to risk-management. That's the only proof we need man, and it's in fact something very simple!
Ok, I can give you a simpler alternative: for every pair, if a trade is already open, don't trigger the following if the entry signal isn't at least at the same price level of the previous, or farther-away by "X%ADR" (being a dynamic variable depending on the ever-changing ADR, as your "SL/TP fraction of ADR" setting).
---
Come on, all's a no-brainer! Mate, I'm just trying to help further improving this awesome tool... Cheers!
- #5,429
- Jul 31, 2019 11:01pm Jul 31, 2019 11:01pm
- Joined Jun 2012 | Status: Trader | 1,106 Posts
{quote} @robots4me – It’s indeed very difficult to reason some things with you. {quote} Why not? I don't mean that the Dashboard automatically opening the charts, but the user, one by one appart, with the touch of a click, if he/she wants (I can send you a video example of this). {quote} According to what you previously said: {quote} I just thought about dividing the 28 pairs in chunks of 7 at a time, instead of 2. And for improving the performance of the Dashboard and AllPairsEA, I mean that instead of computing with each tick, they could do it,...
Just to say I don't think Dot = Trigger, you must have an Arrow not a Dot, but apart from that I like your charts and the message you want to get across
Regards
RT
Some people say I'm lucky, but the more I practice the luckier I get
- #5,430
- Jul 31, 2019 11:04pm Jul 31, 2019 11:04pm
Yes, yes and YES!!
Partial Solution = We could have a setting called "Equity (or Balance) % SL risk per trade", so if you place "1%" with a $10,000 USD account, and there’s an hypothetical 100 pips SL for EURUSD, automatically calculated by the setting "StopLoss fraction of ADR", then the EA automatically calculates the corresponding 0.10 lots per position; so if the 100 pips SL gets hit, the account loses $100 USD = 1%... That way, as the account gets bigger (using Balance or Equity), also the lot-sizing; and vice-versa. A pretty convenient automation, because we wouldn't need to be constantly calculating the lot-sizing ourselves. So it actually contributes to "risk-management", or "micro-management", whatever. And something like that, would also allow us to do back-tests with compounding grown accounts by automatic lot sizing, as well as better gauging approximate possible Drawdowns for longer-term periods.
Solution = "MinDistanceBetweenOrders" (aside from “MaxOpenTradesPerPair").
Yes, exactly that! 
With those DOTS I mean the places where entries would have triggered. And the Dot-Arrow thing maybe also explains my doubts expressed withpink and grey letters on my charts.
Regards!
When looking at the 5th Test (Screenshot 5) with more data the profit level went down, but the graph shows more steps in an upward trend, this again shows that if the lot sizes could have been increased (based on the logic contained within my PDF) with a controlled distance between trades (should the parameter trigger condition have been met), then the growth to the account during this upward period would have been greatly improved.
I don't allow my trading EA's to just open 10 trades because the setting is set at 10, there needs to be more to it than that, if trades are going to open at the same level (within a few pips of each other) then you may as well have 1 large trade which is equal to the 10 small ones.
{quote} Hi Viktory, Just to say I don't think Dot = Trigger, you must have an Arrow not a Dot, but apart from that I like your charts and the message you want to get across
With those DOTS I mean the places where entries would have triggered. And the Dot-Arrow thing maybe also explains my doubts expressed with
Regards!
- #5,432
- Jul 31, 2019 11:21pm Jul 31, 2019 11:21pm
- Joined Dec 2017 | Status: Trader | 4,378 Posts
{quote} Hi R4M, I have done a rough and ready BT (just the way you like it) no Birt's Tick data just the very basic Strategy Tester settings. The first test (screenshot 1) was done using your exact dates, I was hoping to see the same results you showed, but during that period I have 3545 Bars tested while you only record 3112, my test only shows 1.76 Profit Factor, with a very large loss due the the EA opening 10 trades and the SL being at 1.8 ADR with a TP of just 0.2 ADR really well thought out R:R my Rookie friend. The second test (screenshot...
However, I can see that your settings are not correct. Did you see the back test results from post #5410? (https://www.forexfactory.com/showthread.php?p=12414751#post12414751)
I can see that you are using a time frame of 10080 for an H1 chart -- I suspect that is less than optimal. Also your SL and TP are different than the defaults.
I'm glad you are using OHLC data. Here's the next step...
From within Strategy Tester:
1. Select 'Expert properties' and then the 'Reset' button to reset all settings to the defaults.
2. Check the check boxes next to 'StopLoss fraction of ADR' and 'TakeProft fraction of ADR'.
3. The 'Start', 'Step', and 'Stop' for those should be 0.1, 0.1, and 2.0, respectively.
4. For 'Model' use 'Open prices...'
5. For 'Symbol' use GBPNZD
6. For dates use something like: From: 2018.02.01 To: 2019.07.01
7. For Period use H4
8. For Spread use '2'
Check the checkbox for Optimization and then press 'Start'.
When the optimization is completed, you should see something like this under the 'Optimization Results' tab.
Sort the list by 'Profit Factor' and select the best one. Then rerun Strategy Tester normally.
- #5,433
- Jul 31, 2019 11:23pm Jul 31, 2019 11:23pm
- Joined Jun 2012 | Status: Trader | 1,106 Posts
Some people say I'm lucky, but the more I practice the luckier I get
- #5,434
- Jul 31, 2019 11:27pm Jul 31, 2019 11:27pm
{quote} I have an EA that opens trades based upon % of Equity, this does allow the account to growth while ever the Equity in increasing, but if you have a losing period, then the lot sizes are scaled down which results in a longer time period for the account to recover. The best way around this is if the account loses Equity then the lot sizes return to default risk (do not get smaller) and only increase when the account size increases, I call this "High Water Mark", which means the account must recover to its highest Equity level before the lots...
After "MaxOpenTradesPerPair" (already implemented), "MinDistanceBetweenOrders" and "Equity/Balance % SL risk per trade" (in that priority order) are much more important money/risk-management options —and without changing the original strategy at all—, than for example plain unnecessary filters like "ConsecutiveCandles" and "Fractals".
- #5,435
- Jul 31, 2019 11:49pm Jul 31, 2019 11:49pm
- Joined Jun 2012 | Status: Trader | 1,106 Posts
{quote} @Robot Trader -- thanks for digging into Strategy Tester and performing some more back tests. However, I can see that your settings are not correct. Did you see the back test results from post #5410? (https://www.forexfactory.com/showthread.php?p=12414751#post12414751) I can see that you are using a time frame of 10080 for an H1 chart -- I suspect that is less than optimal. Also your SL and TP are different than the defaults. I'm glad you are using OHLC data. Here's the next step... From within Strategy Tester: 1. Select 'Expert...
I did replicate your settings within your earlier post, they were in fact H1 and 10080. you need to find your glasses
I have attached my H4 BT with your new setting, but I had to use the better quality data, to get it to do a full period BT
EDIT: I ran the BT again with my TP/SL settings
Some people say I'm lucky, but the more I practice the luckier I get
- #5,437
- Aug 1, 2019 12:04am Aug 1, 2019 12:04am
- Joined Sep 2013 | Status: Trader | 1,209 Posts
The best money management strategy is to apply a risk assessment strategy for each individual trade opportunity. Learnt that from poker.
- #5,439
- Aug 1, 2019 12:10am Aug 1, 2019 12:10am
- Joined Dec 2017 | Status: Trader | 4,378 Posts
{quote} Hi R4M, I did replicate your settings within your earlier post, they were in fact H1 and 10080. you need to find your glasses
The first result uses SL/TP of 1.8 / 0.4 and the second one is 0.5 / 1.5, right? So -- I think I may have mentioned this maybe only once before -- settings make a difference.
Anyways -- it looks like we are getting similar results. Is that fair to say?
- #5,440
- Aug 1, 2019 12:55am Aug 1, 2019 12:55am
- Joined Jun 2012 | Status: Trader | 1,106 Posts
{quote} @Robot Trader -- the print is sort of small, but it looks like the two results differ in their SL/TP settings -- is that correct? The first result uses SL/TP of 1.8 / 0.4 and the second one is 0.5 / 1.5, right? So -- I think I may have mentioned this maybe only once before -- settings make a difference. Anyways -- it looks like we are getting similar results. Is that fair to say?
You read the settings correctly, but I could not replicate your H1 Profit Factor from your earlier post, which you are trying to tell me was H4 .
Regarding SMALL you need to invest in a bigger monitor(s) DELL 30" x 3 works really well for me
Regards
RT
Some people say I'm lucky, but the more I practice the luckier I get