SWB Stealth (Martingale)

Okay, I think I've got it! The problem was the original code, although it seemed to work fine in version R1. I changed it in both versions and both seem to work fine now.

I've also added a couple of other features:

I added top_level setting, which allows you to set top_level when using tp_in_money, instead of leaving it open. In order to do this, you need to set limit_levels (right below tp_in_money) to true.

Have also added these mods to version R1, and included an update of that as well.

Have fun!

Oh, I also added original RSI option (per your request) and am calling it pre_rsi. Plus I set daily_rsi as a default, so it doesn't matter whether it says true or not ... although if you want to use pre_rsi, set it to true and it will override daily_rsi.

Thank you. I will put this version next to original and will let them ride. With original with global SL closure Im down 300USD with 2 weeks testing. So I hope I will get same positive results with your last reverse version.

Many Thanx..

OK its wierd. Two separate accounts. One with original SWB and another one with reverse version of SWB. They both have 2000 USD and they both start with same lot size 0.01 IBFX minis.. I setup them to compensate the spread. The reverse version is down by 60 USD by one single day.. I have never seen such a performance with original SWB. Original has gained 20 USD which is Average.

Reverse version has SL in money 0.30 (to compensate spread)
Original version has TP in money 0.25

I will look closely to make better setup for each chart separately..

Will be back soon.. Now its time to celebrate my wifes birthday..

Okay, have a new version here. Have added inputs for RSI on all time frames. In order to use, select which time frame(s) you want and set to true. The "median," or "mid-range," is set to 50. If RSI goes higher than this (starting at 51), it's leaning towards "buy." If it goes lower (starting at 49), it's leaning towards "sell." The further it goes above or below the median point, the stronger the trend, although it's generally considered ranging if it's less than 5 points above or below. This is how I understand it anyway. If I am mistaken, I would appreciate it if someone corrected me.

Currently there are three time frames selected by default, M15, H1 and D1, with their respective opening points above and below the median. This is no hard and set rule, just yet, in fact I could be entirely wrong. It just seems like a good place to start. The whole point is to see if we can optimize the opening points for each trade and try to minimize the drawdown.

On top of that, there are two different lot settings, two different tp settings, two different pending offset settings, which correspond to the first tier or first order, and the second tier or all orders that open after the first (when the Martingale starts). You also have a choice between pending orders (buystop and sellstop) and market orders, by specifying an offset value of less than 5 points (typically "0"). If you do, it will send a market order. Anything at or above 5 points, it sends a pending order.

There's also a pending reset (p_reset) that deletes pending orders after a specified number of points, and a reset offset (r_offset) that sets a new offset value (generally higher) after that. This too is adjustable and can be set to 0 if you wish only to open market orders. Anyway, these values can be tweaked too, in order to optimize your opening targets.

There's also tp_stealth mode which, when set to true, sets EA up similar to the original swb Martingale. Except now you have a choice between market and pending orders, as well as multiple RSI time frames ... which, by the way, is enabled to work with tier 2 trades by setting rsi_tier_2 to true.

Thanks.

ETA: Getting faulty sell signal on M15 TF ... changed code. Also added indicator.

If anyone has downloaded the EA and attempted to get it to work, I was getting a faulty signal on M15 time frame and had to change code. It should work better now. http://cdn.forexfactory.com/images/s.../yim/happy.gif

Thanks for updated new version, I will try

You're welcome.

Oh, I found another bug when I started forward testing it on my (IBFX mini) demo account. Everything seemed to work fine when I was backtesting it but, when I started forward testing it (after fixing faulty signal bug), I noticed it wasn't opening any new orders after the initial orders were placed and closed, yet it was giving a definite signal to open. I had several charts open at the same time and they were all doing the same thing.

Come to find out the original code (from original version) wasn't compatible with what I was trying do with it so, I had to modify that too. I have no idea why it wasn't working?

Well, hopefully that should take care of it, at least for now ... of course bugs are part of the process and should be expected, which is why it's good to have other folks help you test it.

If the version you downloaded says 15.0 KB (and hasn't been edited after this post), it should be okay.

Yes, only one bending order only place and execute, when it reach the profit, then only trade another bending order

So other option has not working

so please update and give the details of EA functions

thank you

bye
ajentha

No, I the way you are describing it I think it's working correctly. The problem occurs when it doesn't open a second pending order, even though it gives a clear signal to do so. If you look at the status display in upper-left corner of screen, it says RS120, followed by three numbers. The first two are the D1 and M15 time frames (I'm probably going to change this), and the third is the actual signal that tells whether to buy or sell. If it says 0, that means buy. If it says 1, that means sell. Or, if it says 10 (will probably change this also), that means it's out of range. It also says, right above this, how many open and pending orders there are. So, if both say 0, and yet there is a clear signal to buy or sell (0 or 1), this is when it should place the next pending order. Like I said before, it wasn't doing this.

Oh, so long as the price doesn't dip below the user-defined range (set to 40 by default), there should be only one order open at a time, either a pending or, open buy or sell. It's when the price dips below the range, that the Martingale type grids takes effect. Albeit this is user-definable too, as mentioned in the post above.

I hope this makes it a little more clear.

Oh, here is a picture of my latest backtest, which I started on 6-17, 2009 and ended 11-20, 2010. I started with $500 US at.05 lots, and ended with $497 US. The absolute drawdown was $50.21. The big dip at the end occurred because that's when the tested ended. (I forgot to set it to end today on 11-30.) Also, the last order was placed on 6-9, 2010 and remained "suspended" (without starting Martingale grid) for over five months. Obviously this shouldn't have happened. Oh well, looks like another bug.

Of course I was trying to get it place more orders like it did when I first set it up (before I came up with all the indicator options) so, maybe the two are related? And no, it didn't place neary as many orders (using Martingale grid), as I expected the account size to at least double, which it didn't do.

Thanks.

So, I also continue the demo, I will follow Your instruction

thank you


ajentha

Great.

I should be uploading another version pretty soon that should make it easier to work with, and possibly help spot the problem with it not placing the trades when it's supposed to, if there is one? I think part of the problem was due to the faulty M15 sell signal which was generating a lot more buy signals for those particular settings. So, it may be a matter of adjusting the settings?

We'll see ...

So,
I am waiting for your another version

Have a blessed day to U

bye
ajentha

Okay, it looks like I might have something here. Sorry it took so long. It took me the better part of a day to fix this "suspended" trade issue. It still doesn't look like it's working right, although I think it is, it's just that when a major trend occurs, and you're on the wrong side of it, it may take forever for it to signal a trade to open in the opposite direction. And, although it may look the same, it's not. Before, I noticed several opportunities where it should have opened a trade, but it didn't. Now, it's opening these trades, however, if the trend is strong enough it just doesn't provide the signal ... which is actually probably better, especially if the trend continues. In which case, so long as it does work properly, I'd say it's time to look for a better exit strategy.

I actually changed quite a few things on the EA, but there's only a couple of things you need to be aware of. The main thing is the time frame inputs need to be set to a value greater than 0 in order to work. There is no need to "preselect" them as before. Currently M15 is set to 4 and D1 set to 1 by default. And, even if both are set to 0, so long as no other time frames are selected, they will remain the default. (There has to be at least one time frame selected.) As soon as you select another time frame, however, they are deselected.

I also added a second r_offset and p_reset in order to separate tier 1 and tier 2 trades, thinking it may be causing the problem above, even though it wasn't. I decided to keep it this way anyway, because it gives it more flexibility.

Also, if you look at RS120 in status display (upper left-hand corner) it shows status of each time frame selected ... 0 = buy, 1 = sell, 2 = out of range. For example, in case with the two default signals, it displays only these two ... plus a third signal (basically the first two combined) which signals the actual trade. So, when you see either "/ 0 / 0 / 0" or "/ 1 / 1 / 1," this is when everything lines up and you can expect it to place a trade.

It also displays an "N" when no trades are placed, a "P" when a trade is pending, a "T" when trade is open and price rises above opening price and, a fourth number that begins counting towards the specified range as price drops below opening price. And, once it reaches range, it displays an "R," and continues to do so so longs as it reaches and/or exceeds it. This is when to keep an eye out for the next trade, that is, when you see either "/ 0 / 0 / 0 / 40 ... / R" or "/ 1 / 1 / 1 / 40 ... / R."

Any questions?

Have fun!

Oh, don't forget the indicator! (It's still the same.)

Hello EC and all fellow traders!
My computer for testing was infected by a virus so I have to reinstall Win and all records and tests are gone. I will start to test next year, because I am too busy now... Looking forward to test reverse and regular version of SWB..

Cheers and Merry Christmass to all!!

Karel

We'll still be here, I guess? Or, if not, just shoot me a PM.

Oh, sorry to hear about your computer.

I updated versions 1.01 and 1.02 by the way. Come to find out I needed to add a reset after secondary pending orders were deleted (when out of range), similar to the one used on primary pending orders. I thought I could use the same one for both, but no, they are triggered at different levels.

This still doesn't address the "suspended" trades issue, however, and that still needs to be addressed.

Hi Ec! Its great to hear from You.. Im glad this little thread is still alive..
Now I need to do some preparations for Christmas. I have for now some suggestions but still I will be smarter after some records of backtesting..

Wishing to You all nice Christmas..

Hi all,
think it's time to pay more attention to how to get out of the trade when we using Martingale system.
It's obvious that when the trade levels increase on our trade, increases lot power too.That's the way we coming to comic situation when on higher levels with big lot power moving in our favour,when one pip itself cost tens (hundreds) of pounds,when the trade take days, we collect only fraction of a pip(wish is posible if you start with 0.01 lot and your TP is pound or two).
In post #1 ecTtrade says: "Am also adding another feature which allows you to scale tp_in_money target up and down, depending on the trade level..." -that would be great!
OMO elegant way to do that is piece of code added to the EA:"after level one when TP in pounds is hit, wait for number of pips(changeble from settings tab)and then close the trade",because TP increasing proportional of the trading levels.
example:let set number of pips to 1;start lot to 0.01;TP to 1 pound.
you can see what gives you 1 more pip on every level:


level 1.---1.0pound
2.---1.3
3.---1.7
4.---2.5
5.---4.1
6.---7.3
7.--11.27
8.--26.55
9.--52.15.............etc.(digits are nearly there)
As you can see that increase profitabiliti with the same amount of risk.

need to discuss
need yor opinion

hope i catch attention of the coders too

wake up!

pips to all of you!

peter

ecTrade-
I am new to this forum, but I have been trading forex for a number of years now. I found this SWB while doing research on grid trading and I find it remarkable, right before it hits the "death trade." I know there are a ton of much smarter people than me on this forum, but I think I have a simple solution to make the original SWB function the way everyone wants it to. I can only think there MUST be a reason nobody has thought of it.

It is to add a hedge trade, but not the way this hedge version does it.
Here is how it should work.

Options on the EA:
Allow_hedging = 0/1 (or true/false)
Activate_hedging_level = 4 (level at which to being the hedge trade)
Hedge_pip_lockin = 1 (amount of pips to lock in the hedge trade)
Hedge_lockin_after = 10 (amount of pips to go before the lockin is activated)
Hedge_stop_loss = 5 (amount of pips for the stop loss – just in case the hedge trade is activated, but not enough to lock in.

The idea isn’t to make money off the hedge by using a take profit or a trailing stop on the hedge trade, it is just for big loss minimization. The last hedge trade version took profit at each level.


Here is how it will work.

Example - The EA is has fired it's level 3 trade and drops another 25 pips. Instead of firing the usual 4 level martingale trade, hedge trade activates at the same lot size a normal 4 level would be. After 10 pips, the hedge lockin of 1 pip is activated, which also moves the hedge stop loss to the lockin value. The hedge trade will close 1 of 2 ways. 1) if the hedge trade makes the required pips to cover the other 3 levels (25 pips or so), then it closes all of them, just like it would do if the price did it's 25 pip retrace. 2) if the price retreats activating either the stop loss or the lockin of the hedge trade. If the price continues, the EA will close the other 3 levels like normal. If the price reverses and all of the criteria is met again, then the hedge trade will activate again. And repeat……


I think you can't lose doing this, but I can't program this in to test it. You win regardless of which way the price moves.


What do you think? Does this idea have merit?


Thanks a bunch for everything you do!