Macd Stochastic Fractal MA5/25 Daily Marylin

[phil_trade]

Hello Phil,
I also run a demo for about 3 month and the equity is still at about the starting point. Sometime you win sometime you lose.
I think there is something in the ea that is different with your piece of software. The question is what is the difference.
I am no coder at all so i really do not know.
I keep my demo account running (alpari-uk). (sorry no myfxbook, but it is more or less similar to the bud49 or trader689). Also Niepce got stuck in break even, he wrote sometime ago.

Sometime you win sometime you lose. that's right... but did you let the win run ?

€/$ for 2009/2012 test, win/lose trade ratio is about 50% but lose are short and win very long...so at the end you should got this :

New Benchmark 7 devises 2009 2013 +150K$.PNG

Hi phil,

I was wondering how you can get those backtest results? As far as I know, you said earlier in the thread that you tested with Forex Tester 2, and unless I'm wrong, you can't test with Marilyn's Support and Resistance indicator... right?? This particular program only seems to be able to test with its own indicators, so MA, MACD and Sto aren't a problem, just the S&R indi. It also doesn't test EAs. Please correct me if I'm wrong.

Also, I have been running on a few demos myself and before my computer crash (along with all relevant passwords), they were mostly doing fine. I was running forward tests on FxPro and ICMarkets because I wanted to test them with only 5 daily candles. With initial parameters, one demo was doing over 100% after 2 months (with 1 lot per 25k$). The swings back and forth are huge with original parameters, but I am one of those people here who think your EA should get more love. If I can find out how to backtest this in something other than Empty4, I'd be willing to try out a few ideas, but mostly I think it's already fine the way it is, as this is such a personal matter. You might like more filters for entry or exit than me, and really, it comes down to the individual and their own objectives.

Cheers from Montréal!

Hi

just the S&R indi. It also doesn't test EAs. Please correct me if I'm wrong.
I have written a transcription in Delphi of S/R indicator

It also doesn't test EAs
Yes, I wrote the EA with Delphi XE2 and then translate it to Windev/Empty4.

And I have control with FT, that in new last month ( april/May) that FT give the same result as my live
Windev/Empty4.

cheers "cousin"

Philippe

[TonkaTuff]

Hi all,

I've had a demo account running since 30 May 13. No Manual intervention.

EA: Marylin 0.36
Symbols: AUDJPY, AUDUSD, CHFJPY, EURJPY, EURUSD, GBPUSD, USDCHF, XAUUSD
Set:Spread 1.5x Broker, Lotsize 0.25
Set files as posted by:niepce (thanks) here: http://www.stevehopwoodforex.com/phpBB3 ... &start=200

There have been some great trades, but nothing has closed in profit. Can anyone suggest any variable changes we could make?

Would there be any value in imp,implementing Chandlier Stops to take profit?

ChandelierStops_v1.mq4

Below is iExposure, Empty4 Statement and a shot of each pair that has traded with P4L_OrderHistory script show the Trade History.

iExposure

mm_0.36 13 jun 13.png

Empty4 Statement

MM 0.36 as at 13 Jun 13.JPG

CHFJPY

chfjpy_d1-13-jul-13.png

EURJPY

eurjpy_d1-13-jul-13.png

EURUSD

eurusd_d1-13-jul-13.png

GBPUSD

gbpusd_d1-13-jul-13.png

USDCHF

usdchf_d1-13-jul-13.png

XAUUSD

xauusd_d1-13-jul-13.png

[phil_trade]

"TonkaTuff"]Hi all,

I've had a demo account running since 30 May 13. No Manual intervention.

EA: Marylin 0.36
Symbols: AUDJPY, AUDUSD, CHFJPY, EURJPY, EURUSD, GBPUSD, USDCHF, XAUUSD
Set:Spread 1.5x Broker, Lotsize 0.25
Set files as posted by:niepce (thanks) here: http://www.stevehopwoodforex.com/phpBB3 ... &start=200

There have been some great trades, but nothing has closed in profit. Can anyone suggest any variable changes we could make?

Would there be any value in imp,implementing Chandlier Stops to take profit?

Hi Great results !

There have been some great trades, but nothing has closed in profit
Marylin will close them when trend end will be detected. Sometime, trade are close to SL = OPENPRICE cause trend not strong enough.

Would there be any value in imp,implementing Chandlier Stops to take profit?
Could you explain how to use it ?
Have you already test this ?
it could be a good added option.

Thks

Philippe

[TonkaTuff]

MM definitely picks the trends nicely.

The MQ4 for Chandelier is attached above. Here's a definition of I found.

Chandelier trailing stop: Invented by Chuck LeBeau. He calls it the Chandelier because the stop loss
hangs off the most recent high in an uptrend. Not sure what he calls in during a down trend – 'bar stool' perhaps? For a buy trade (reverse everything for a sell trade):
• Each time the market makes a new high, the stop loss is moved by this price
minus Atr * ChandelierAtrMultiplier.
• Because Atr can change quickly with increasing/decreasing market volatility,
this becomes a highly adaptable stop that moves lower during volatile
periods and higher during stable ones.
• This is a trend-trading stop loss, so the ChandelierAtrMultiplier will decrease
as the trade becomes more profitable, initially reducing risk and eventually
locking in more of the profits:
▪ when the profit reaches a multiple of 4 Atr's, ChandelierAtrMultiplier will
reduce to 2.
▪ when the profit reaches a multiple of 6 Atr's, ChandelierAtrMultiplier will
reduce to 1.

Would it be an idea to setup a standard set of exits that could be incorporated in any strategy. Maybe as discrete pieces of code so developers could incorporate as relevant. Or as a standard code block, so as to develop a standard code stream. Maybe start with a list such as (based on Steve's list from Tiger):

• 1. Set SL & TP in pips: Pretty standard stuff
  2. ATR SL & TP: Use multiples of Atr to set a 'hard' SL & TP.
  3. RSI exit: Close a buy/sell trade when RSI changes to over bought/sold respectively.
  4. ADX exit: Use ADX value set to close a trade.
  5. Exit on change of trend to range or vice versa: Is this what MM laready does.
  6. Chandelier trailing stop: As above.
  7. SAR Parabolic TS: Use the SAR Parabolic to set the trailing stop.

Cheers,
TT

[phil_trade]

MM definitely picks the trends nicely.

The MQ4 for Chandelier is attached above. Here's a definition of I found.

Chandelier trailing stop: Invented by Chuck LeBeau. He calls it the Chandelier because the stop loss
hangs off the most recent high in an uptrend. Not sure what he calls in during a down trend – 'bar stool' perhaps? For a buy trade (reverse everything for a sell trade):
• Each time the market makes a new high, the stop loss is moved by this price
minus Atr * ChandelierAtrMultiplier.
• Because Atr can change quickly with increasing/decreasing market volatility,
this becomes a highly adaptable stop that moves lower during volatile
periods and higher during stable ones.
• This is a trend-trading stop loss, so the ChandelierAtrMultiplier will decrease
as the trade becomes more profitable, initially reducing risk and eventually
locking in more of the profits:
▪ when the profit reaches a multiple of 4 Atr's, ChandelierAtrMultiplier will
reduce to 2.
▪ when the profit reaches a multiple of 6 Atr's, ChandelierAtrMultiplier will
reduce to 1.

Would it be an idea to setup a standard set of exits that could be incorporated in any strategy. Maybe as discrete pieces of code so developers could incorporate as relevant. Or as a standard code block, so as to develop a standard code stream. Maybe start with a list such as (based on Steve's list from Tiger):

• 1. Set SL & TP in pips: Pretty standard stuff
  2. ATR SL & TP: Use multiples of Atr to set a 'hard' SL & TP.
  3. RSI exit: Close a buy/sell trade when RSI changes to over bought/sold respectively.
  4. ADX exit: Use ADX value set to close a trade.
  5. Exit on change of trend to range or vice versa: Is this what MM laready does.
  6. Chandelier trailing stop: As above.
  7. SAR Parabolic TS: Use the SAR Parabolic to set the trailing stop.

Cheers,
TT

Except MQLChandelier, I have tested all SL technics.

The main concern with SL, is to avoid to break trend to early. Otherwise you expose Marylin to bad trades during retracement and avoid good added trades at the end of retracement.

So I only have implement :
SL = OPEN when Stoch OBOS if delta pips minimum
SL = Fractal
SL virtual to limit max loss

TP is detected at the end of trend by MACD.

I don't think SL is a problem for Marylin. I would prefer to avoid bad trade without limited good trade entry. Problem is that same conditions occurs for both type of trade. Need something more to separate them but what ? volume ? any idea ?

Philippe

[TonkaTuff]

I understand what you mean regarding SL.
What do you think about Chandelier as a TP strategy?

Not being a great believer in back testing, I could set up 3 to 4 more instances of MQ4 to trial different settings. What I need is some educated guesses as to what settings to use. What does everyone think would be good to try?

Thanks,
TT

[phil_trade]

New Brain Storming request !

I never success to filter bad trade and good trade : they always look like to me similar at the beginning !
The only way to manage them it's to close them after a while.

For all of us who have trades since 2 weeks : please analyse each bad trade/good trade ! If you can see or add something to distinguish them.... please let me know !

€/$ example :

EURUSD GOOD TRADE and BAD TRADE look like similar to filter signals module.PNG

any idea ?

Philippe

[phil_trade]

I understand what you mean regarding SL.
What do you think about Chandelier as a TP strategy?

Not being a great believer in back testing, I could set up 3 to 4 more instances of MQ4 to trial different settings. What I need is some educated guesses as to what settings to use. What does everyone think would be good to try?

Thanks,
TT

What do you think about Chandelier as a TP strategy? Before testing, I need to find or translate it in Delphi for FT2. I will have a look.

[spotdespot]

New Brain Storming request !

I never success to filter bad trade and good trade : they always look like to me similar at the beginning !
The only way to manage them it's to close them after a while.

For all of us who have trades since 2 weeks : please analyse each bad trade/good trade ! If you can see or add something to distinguish them.... please let me know !

€/$ example :

EURUSD GOOD TRADE and BAD TRADE look like similar to filter signals module.PNG

any idea ?

Philippe

Hi Philippe,

I have had a look at the last couple of weeks good and bad trades and added an indicator I have been using recently as a manual trading aide - the LSMA in Color with a 63 period. See the charts below. "Bad" (ended in loss) trade lines in RED, good (no loss) trade lines in GREEN

Basically the idea is simple and is to add a filter to the existing logic that says only take a long where the LSMA is green and vice versa for red.

It doesn't filter ALL the bad trades but it does filter a fair few and just as importantly it did not prevent the taking of ANY of the positive trades. (EDIT: Having said this it would have filtered out your "Good" EU trade. Maybe just worth some thought anyway )

This is a freely available indi (attached).

I hope this helps.

Cheers,
Dave.

[phil_trade]

Hi Dave

I will have a look on LSMA. May be we can adjust Marylin strategy to integrate it.

Thks

Philippe