Indicators development for T101's not very simple method | Page 119 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

SIMPLE EA for a simple method needed!! 2 replies

Simple System By T101 - My Thought 18 replies

Continuation of T101 method discussions 24 replies

My first attempt at T101's basket method... 2 replies

Attachments: Indicators development for T101's not very simple method
Exit Attachments

Indicators development for T101's not very simple method

  • #2,361
  • Jun 11, 2010 5:53am Jun 11, 2010 5:53am
@CHARVO and ROBERT
The TrendMagic gets you into a trend and keeps you there till stppoed out by cross to the other side, this can cut
into your profit deeply. However, when you use MACD for exit it gets you out @ the peak of retracement and you
come in again @ better price. see the attached pix.

@ROBERT
Quote
the current rules must all line up on the same bar

Not quite so, sometimes it does on the same bar, but most of the time MACD deviates, BI_T101 crosses, Price goes above/below
TM and I have to wait hours for SAMIR to give SHORT/LONG STRONG before I pull the trigger, or sometimes I will wait for price
to close/open above/below TM like it happened this week. PATIENCE is the name of the game.
John 3:16 - From THIS love no power on earth can me sever.
  • #2,362
  • Jun 11, 2010 5:55am Jun 11, 2010 5:55am
@ALL

I MUST SAY THIS; the crux of this strategy is timing your entry correctly else many times you will sell and price will go
and knock you out -560 to get you off its back and still end up selling massively and leaves you bleeding and wondering what
was wrong, its THE ENTRY

That is where TrendMagic becomes very invaluable to me, it shows and draws S/R for me and picks good entry point and reentry
into a trend.

Put TrendMagic and MACD on a clean 4HRS TF chart, zoom out a bit and eyeball the chart

The case is that we all want 10 basket trades in 1 week which is good but to me I like 1HR TF and 4HR TF and may be 2 to 3 baskets
in a week using 1HR TF. I wait for the 4 rules to align and if backed up by divergence then I confidently drag my script on, knowing that
5% of my account is at stake, and that is why I take caution thru patience

kos
John 3:16 - From THIS love no power on earth can me sever.
  • #2,363
  • Jun 11, 2010 7:45am Jun 11, 2010 7:45am
I think there is danger in placing too many restrictions on what has to be done, in order to conform to a given set of rules or a specified approach. It stifles innovation, no matter what the subject at hand. A horizontal, free-floating, bring your ideas to the table, approach, is more likely to result in viable solutions.
If one says it must be done a certain way, because it was always done that way in the past, or a special person has always done it that way, it closes many doors.

Best regards, Glenn
  • #2,364
  • Jun 11, 2010 8:23am Jun 11, 2010 8:23am
Robert,

Rule 4 is actually which is positive (e.g. trend), Buys or Sells, for the Week or Day, depending on timeframe used.

Regards,

Don


Quoting MrPip
The current rules must all line up on the same bar.
If the first rule is met but later is not met then the rules are all checked again from rule 1.

Current rule check order

Rule 1 is for Trend Direction.
Rule 2 is Macd
Rule 3 is BI_Red or BI_Blue
Rule 4 is sum of red and blue.

Robert
  • #2,365
  • Jun 11, 2010 8:26am Jun 11, 2010 8:26am
So you're using _6 indicator that Robert (MrPip) posted? Just put it in indicators folder and latest EA in EA folder, and should be good to go. Of course, there are things that need clarified by Kos still, and testing/tweaking that need to be done still IMHO.

Regards,

Don


Quoting forexmes
Thanks for your help, it worked, just changed the version of BI_T101_OTR_V1_5.ex4 to BI_T101_OTR_V1_6.ex4 in the folder containing the indicators.
Indeed in the "experts" tab in my metatrader I had a message saying that BI_T101_OTR_V1_6.ex4 couldn't be added to the GBPJPY chart.
So I didn't have to change the code, just change the indicator.

Thanks for your help, it's now running.
  • #2,366
  • Jun 11, 2010 8:53am Jun 11, 2010 8:53am
Quoting mas toro
yes i had too.
so i change

with
and it's work.
going buy right now.

thanks

note.
i use this trailing for profit lock.
maybe somebody will get idea from this ea (!T101_EA_Trailing).


can yot explain how the ea work????
  • #2,367
  • Jun 11, 2010 9:39am Jun 11, 2010 9:39am
Robert,

Exactly.

Stay in for awhile through retracements if we can determine tops and bottoms. The TrendMagic/SupertTrend trend indicators do a good job of that, although there are alternatives (NonLagMA_v7, Kaufman or another type of adaptive moving average, maybe Advanced ADX which looks perfect for stop and reverse trading, maybe Waddah Attar Trend and Advanced ADX together as a combo).

I have already mentioned making sure the bar's high and low is higher than last bar's if buying, or lower if selling. But we could also make sure that the CURRENT bar is above TrendMagic, not just last bar. Those two would help in keeping us out of some unprofitable trades.

The main thing to me is we now an EA "shell" we can work with. One that is trading properly for the strategy we've been given. That's all that can be asked of an EA. It will only ever be as good as it's strategy performs - at best. But you have to have a good foundation for any EA, which I believe we definitely now have - structure. I do need to probably add more code to show what's going to want to be changed and what will need to be left alone. About the only thing in trading functions I see that would need changed would be pairs traded.

There are MANY alternatives. Honestly, we still haven't talked about Core 5 pairs. We do have Universal pairs (IBFX) and non-IBFX pairs. We really deal with "groups" by giving options for reversal of pairs that begin with USD and the MIDS. Then there's Core 5 we haven't talked about. Then zznbrm came up with a couple more alternative baskets.

I have also made in indicator based on an email I got concerning "Currency Rings" (trade 3 currencies the same direction - choose 1 set from 14 sets of currency rings. indicator is still being worked on).

See, there is NO WAY that a broker or anyone else can stop anything with so many variations of baskets to trade. And that's just the pairs to trade, not even getting into possible strategies which all can be applied no matter what pairset we decide to trade. The trading functions would be set to trade the pairset chosen.

See why I wanted to get a "Shell" EA that correctly traded first? Because now we have the foundation to build the house, with as many rooms as we want.

Regards,

Don


Quoting MrPip
I was thinking about finding a better lock profit with exit signal when the trend changes. If SuperTrend does not repaint then the buy trade from earlier would still be in play until SuperTrend reverses. The same appears to be true for Trend Magic. With a less tight trail this could result in moves like today being obtained in a single basket.

This would result in a more stop and reverse type of trading.
Keep in the long backet until long trend reverses. You would then expect the next basket to be sells.

Robert
  • #2,368
  • Jun 11, 2010 9:44am Jun 11, 2010 9:44am
Yes, exactly, by using the posted indicator, the problem has been solved.

Thanks all

Quoting disbellj
So you're using _6 indicator that Robert (MrPip) posted? Just put it in indicators folder and latest EA in EA folder, and should be good to go. Of course, there are things that need clarified by Kos still, and testing/tweaking that need to be done still IMHO.

Regards,

Don
  • #2,369
  • Jun 11, 2010 9:49am Jun 11, 2010 9:49am
Robert,

When the IA is put on demo at very first tick of week (impossible without automation, but just for sake of illustration), we put on 7 buys, and 7 sells. When we buy a pair, we pay the spread first during opening the buy, and when we closebuy, we don't pay spread, just the bid. When we sell a pair, we pay the bid price, and closesell at the ask price (so pay spread on exit).

Now, if we want to make an indicator or EA and use historical data instead of making and waiting on an actual IA demo account to "build" using real-time prices, we then have an indicator that can be used on the real account or in EA on real account and have done away with need for IA.

But we have a problem. We don't have historical Ask prices. This is why spread must be used at least for the buys entry, so we add spread to bid open price to make a faux ask price for open of buys in the basket. The close of the sell could use ask price, but I decided to treat it the same in code and just add spread to bid. open of sell and close of buy are just bid, no spread. But basically I coded it as I did to overcome the problem of not having historical Ask prices, or historical spread either, and I thought MarketInfo returns current spread.

Regards,

Don


Quoting MrPip
I found some code in the EA to use spread.
The problem is that using MarketInfo will only return the default spread, not the current spread. Current spread would need to use Ask - Bid for each pair.

What is UseSpread suppose to do here?

Robert
  • #2,370
  • Jun 11, 2010 10:02am Jun 11, 2010 10:02am
trade4fun,

when you say "sumblue changed into 0", you're also saying resistance or support was hit on the weekly IA or Daily IA. That's what's effectively happening. I've seen it happen so many times over the weeks. A trend goes out from 0 on BuysVsSells and comes back to 0 after hitting support or resistance (depends on if sells are above 0 or buys LOL, but the "resistance" points moving out from 0 seem to me - and I could be totally wrong here - seem to be around every 250 pips), then comes back down to 0, which itself is support/resistance, and takes off again hitting same resistance as the last trends (even if the last trends were Sells/SellAll and this trend is Buys/BuyAll LOL). I've also noticed one broker's great entry is another broker's getting in on the end of the move and losing or just having small profit. Not good IMHO. So we really need to know the broker Kos uses, or at least the GMT offset of that broker if we want to make the "signals" the same between brokers. The other 3 indicators are going to be very close. The 4th condition is the one that varies so much between brokers because of different week and day opening times. But you are correct, you are now getting a clearer picture, and the EA is using a system which could be used as an indicator if you wanted to do the trades manually. But the point is: If you get that close to automating something, why not just automate it if possible, right? But yes, many people want alerts for their strategies/systems. Alerts are one step away from complete automation IMHO. Worth the extra step I'd say.

Regards,

Don


Quoting trade4fun
goodmorning all,

i am sorry i have not read your posts above yet, because i have to run out for work. so i just add my results for the 24 hour here.

after 24 hours this is the result so far. some very intresting stuff happened.

at the fifth trade, 1 signal changed. sumblue changed into 0.

after this happened the trade became a loser, and trade stopped when the stoploss was hit.

after this trade it did indeed stopped doing direct buys after the fifth trade was closed and EA waited 1 hour before opening a new 1.

however this new...
  • #2,371
  • Jun 11, 2010 10:31am Jun 11, 2010 10:31am
Kos,

I don't know about your timeframe alignment. Hopefully you have seen where I talk about different brokers having different week and day start times, and how this affects when we get into a trade. On one broker you'll get good entry. On another broker, you'll get in very late in the trend (at support/resistance - exactly the WRONG time to get in to the trade), and lose or may get small profit if profit-locking. We need to know your broker, or at least the GMT offset your broker uses. For instance, IBFX is GMT 0. I think this indicator works I made last year, which is attached and should help you find your broker's GMT offset. This way, I can work on making that code work so that your "trend" is my "trend" if we use the same timeframe. And my broker having different open would not affect us, since we would work toward using a universal opening time for week or day (set in EA in GMT time).

Also, I see that you've wrote a Sell on chart you attached. You see the bar that Sell is on? on the previous bar we have a long bar down that close below TM. Right now on H4 chart we have about the same scenario, yet I've had one loser and am in another losing trade. First losing trade could have avoided if would make rules that says this bar's high must be higher than last bar's high. I lost, but since everything as far as known and clearly specified rules look to have been met, the buys trade was taken. vSecond losing trade could have avoided by specifying that not only last bar had to close above TM, but price right now must be above TM.

I hope you see why being precise is necessary. Nothing can be taken for granted. You might say "of course I wouldn't get in there", but if it's not specified in the rules, it will not be taken into account when EA decides to put on trade or not.

Also, I don't see where you've taken into account profit-locking whatsoever, and have pointed out trades based solely on MACD exit. I will test this way without profit-locking, although you have said you use profit-locking before. The EA doesn't trade with discretion. It does exactly what it is told to do (if it's working right). Gaps in strategy show up.

Instead of posting good historical "possible" trades, I ask you to now test the EA and tell us whether it is trading correctly based on what you would do. I bet if you do that, you will come up with other rules like the 2 I made above, which you'll realize you unconsciously use, yet haven't "pointed out" to us before.

Regards,

Don


Quoting Kosomolate
I dont know if this will brighten things up a bit

4HR TF GJ

kos
Attached File(s)
File Type: mq4 GMT_Test.mq4   6 KB | 420 downloads
  • #2,372
  • Jun 11, 2010 10:40am Jun 11, 2010 10:40am
Are you using T101_EA_byDon_JustKos_06-09-10d that Robert uploaded? I have it opening trades after it closes with profit-locking, but those trades don't "close automatically" (sounds like a problem that was encountered in earlier version). They only close by SL or profit-locking (usually first step profit-locking) so far with this version on this market for last couple days on H4 chart.

Please use the latest version, and we all need to chill and enjoy the weekend. Let it run next week, and tell us if it's trading like you would, or not. I've noted some things you'll need to look at, and may want clearer "rules for the dumb robot" to go into the EA.

Regards,

Don


Quoting Kosomolate
o

@DON
when I got knocked out by MACD I wait for retracement back to TrendMagic which can take hours and when price bounces
off TM I reentre many times at a better price and of course IF all conditions for entry remain intact.
When the knock out is by profit locker, I check to see if conditions are still OK and reentre again. Thats because
I know that profit locking gap by julius is too tight. I f you have 150 lock 70. It will take the average move of
5 pips each of the 14 pairs to close you out @ 70. though it has saved me many...
  • #2,373
  • Edited 11:53am Jun 11, 2010 11:06am | Edited 11:53am
winsteadglenn,

In due time, I will agree with you.

For now, doing Kos' strategy, it's his, and only he can say if the EA is doing it right or not. I gave the EA, and either we stay on path of what we're doing with Kos' strategy, or I will remove myself from the discussion.

If you want to code your strategy, get the latest posted EA, change the name, put in your indis and rules, and post it if you want, or don't if you don't want. If you really know what you're doing, go ahead and rock our world!

When you say "A horizontal, free-floating, bring your ideas to the table, approach, is more likely to result in viable solutions", I say it will bring chaos which you nor any of us are ready for and maybe I'm wrong and you have a database or something that can keep track of many EAs, test results, feedback, comments, etc. without "losing it", but I don't, and really am seeing how this goes before I really am going to commit to anything else.

Regards,

Don


Quoting winsteadglenn
I think there is danger in placing too many restrictions on what has to be done, in order to conform to a given set of rules or a specified approach. It stifles innovation, no matter what the subject at hand. A horizontal, free-floating, bring your ideas to the table, approach, is more likely to result in viable solutions.
If one says it must be done a certain way, because it was always done that way in the past, or a special person has always done it that way, it closes many doors.

Best regards, Glenn
  • #2,374
  • Jun 11, 2010 11:15am Jun 11, 2010 11:15am
Kos,

The "small" candles on H4 are much bigger than 560 pips on the basket. The bottom line is profit-locking numbers either were: never used by Julius (they aren't in his profit-locking EA, only posted by him, so is in our EA). I have some revised numbers I am thinking about using. But if it retraces more than a SMALL amount on even one H4 bar, it gets taken out by SL. So unless you get "perfect entry" (on a bar that has it's high equal to it's open, or either not much of a high-open gap, then you'll probably get stopped out on the same bar and like I had like 4 trades on the same bar???? that's not an entry problem near as much as a profit-locking being too tight problem).

Regards,

Don


Quoting Kosomolate
@ALL

I MUST SAY THIS; the crux of this strategy is timing your entry correctly else many times you will sell and price will go
and knock you out -560 to get you off its back and still end up selling massively and leaves you bleeding and wondering what
was wrong, its THE ENTRY

That is where TrendMagic becomes very invaluable to me, it shows and draws S/R for me and picks good entry point and reentry
into a trend.

Put TrendMagic and MACD on a clean 4HRS TF chart, zoom out a bit and eyeball the chart

The case is that we all...
  • #2,375
  • Jun 11, 2010 12:54pm Jun 11, 2010 12:54pm
Good evening.

hours 24 till 32 update of the D version.

BUG FOUND

what happens is that the trade was still going on while suddenly all 4 signals changed to sell. however since my trade did not yet finnish the new sell signal was not opened until 2 hours later when the trade actually hit stoploss. then it opened up a new trade for sell. ofcourse 2 hours later is way to late. so that trade is also still in the loss now.

however i think it is a good idea to stop this EA now on my chart. it has run 32 hours and i dont think it is a good idea to further test the Ea under these settings. also i would like to test ART and Don their indicators so i dont have to check history data for today by hand. but i cant check the indicators unless i restart the metatrader.

regards

trade4fun
  • #2,376
  • Jun 11, 2010 2:35pm Jun 11, 2010 2:35pm
Here's the currency ring indicator I was talking about, and PDF attached. Like I said, it's not finished. There's also Average Daily Range that would be taken into account per the PDF, but not implemented into indicator. Just something to play with on the side, to show it don't matter which pairs you want to trade, pips are available.

Regards,

Don
Attached File(s)
File Type: pdf currencyring.pdf   54 KB | 1,237 downloads
File Type: mq4 CurrencyRing_Don.mq4   5 KB | 634 downloads
  • #2,377
  • Edited 3:07pm Jun 11, 2010 2:47pm | Edited 3:07pm
trade4fun,

This is why during checking entry signals, I had code to exit BuyAll if got signal for SellAll (what Robert was talking about I had exit code and entry code together).

So I agree it is a bug. I coded it to take care of that problem. In fact, I'll say that I JUST got a SellAll signal, and my trades were not closed and are still open too.

BUT, I look at sum_blue and it's at like 690, while sum_red is at like -650. So there is no way I should have even gotten a SellAll signal like I did. So I don't know if this is another bug or the same bug. I do know that neither the MACD, nor the BI, nor the Weekly trend was showing short. The only condition that was good for a short was TM. Edit: My bad. I was creating a new EA from this one, and had alerts on still, although some other stuff had been replaced, so when I placed it on anther chart I got the "Kos" signal (which had not been renamed yet), so maybe this is the SellAll signal I got. Anyways, the account lost more than it gained on testing this EA version.

I too, am going to stop testing this version and will pick up next week maybe.

Maybe Robert can look at the code and see what's up, or maybe he's thinking like I am and thinking about picking up last EA I posted, and just changing the CloseAll() code and nothing else, to see if that would have fixed it to trade like is intended.

Hope everyone has a good weekend.

Regards,

Don


Quoting trade4fun
Good evening.

hours 24 till 32 update of the D version.

BUG FOUND

what happens is that the trade was still going on while suddenly all 4 signals changed to sell. however since my trade did not yet finnish the new sell signal was not opened until 2 hours later when the trade actually hit stoploss. then it opened up a new trade for sell. ofcourse 2 hours later is way to late. so that trade is also still in the loss now.

however i think it is a good idea to stop this EA now on my chart. it has run 32 hours and i dont think it is...
  • #2,378
  • Jun 11, 2010 3:02pm Jun 11, 2010 3:02pm
Quoting Kosomolate
@CHARVO and ROBERT

Not quite so, sometimes it does on the same bar, but most of the time MACD deviates, BI_T101 crosses, Price goes above/below
TM and I have to wait hours for SAMIR to give SHORT/LONG STRONG before I pull the trigger, or sometimes I will wait for price
to close/open above/below TM like it happened this week. PATIENCE is the name of the game.
BI_T101 crosses is not used by the current EA. It only checks red above blue or blue above red. There is no check for a fresh cross.

The same is true for any trend check using TrendMagic, SuperTrend or LSMA. There is no check for a new trend change, only the current trend.

Robert
  • #2,379
  • Jun 11, 2010 3:29pm Jun 11, 2010 3:29pm
thanks Don for the answer and explanation. hopefully this can be code fixed for next week.

about Kos his patient rule. so far this week i think we can all agree on that if you add the Ea on the chart and it will open the first trade, it is always a winner. the problems that occur are after it. that it keeps opening new buys direct after trade is finnished. and even if some are winners. in the end 1 big lose wipe out your profit.

the first trade has the strongest signal and the most way to be profitable. after that not so. so if we have to wait for a clear new signal after first trade. then why not?

i rather have 1 or 2 trades a day that are profitable. then 6 trades a day where 1 wipes out your profit.

also if i am not mistaken this is also what kos means with rule of this is to be patient.

i hope Don and Robert have time and are able to fix some bugs for the Ea so we can fresh test it next week.

have a good weekend all!
  • #2,380
  • Jun 11, 2010 3:58pm Jun 11, 2010 3:58pm
I added some code to the init function that shows the trade type to be taken for each pair based on current direction of EMA(14) with result printed to the expert log. While not perfect it does show a possible filter. The idea is to try to trade each pair in its current direction based on some indicator instead of forcing the buy all or sell all rule. I realize that there are reversal inputs for some pairs but that might not always work.

Here are the results of the filter on M15.
The EA has a buy all except for rule 4.

USDCHF : SELL with Trade_USDCHF_Opposite = True so BUY
EURJPY : BUY
AUDJPY : BUY
CHFJPY : BUY
GBPCHF : SELL with Trade_GBPCHF_Opposite = False so SELL
EURGBP : BUY with Trade_EURGBP_Opposite = True so SELL
GBPUSD : BUY
USDCAD : SELL with Trade_USDCAD_Opposite = True so BUY
GBPJPY : BUY
EURCHF : BUY with Trade_EURCHF_Opposite = False so BUY
EURUSD : BUY
USDJPY : BUY
AUDUSD : BUY
CADJPY : BUY

This will result in 12 buys and 2 sells.
Trading opposite for 3 pairs could potentially give 3 losses in the backet that might otherwise be profitable trades.

Any other ideas for this filter are welcome.

Robert
0 traders viewing now