StepsOfGlory

**updated**

I'm not offering this EA anymore due to lack of interest, and so I don't really have time to keep it updated. Feel free to PM me if you have any questions.

Here is a screen shot of a current euro backtest using the settings file from above. This is what your backtest should look like if you installed the strategy correctly.

I noticed not much interest in this strategy (though I did see 50 or so downloads)... so I went ahead and posted an updated version.

The new version is smaller (code wise), and runs considerably faster. For those of you who want to run some optimizations to come up with your own variations (aka your own strategies) with this thing, you'll be happy to know that it runs many times faster in the optimizer now than the old version.

Oh, and now that it's smaller, it's just a single zip file now rather than two. So its easier to install.

You guys can play around with it now a little faster than before, and if you find any good settings, do share. This is a large EA (over 3mb in size) with a lot of settings and possibilities. Enjoy.

Well I just skimmed through the thread and didn't find any explanation of how this ea trades so that may be the reason.

Can you give us an explanation of how this ea trades your settings?

Here's my 11-year graph for E/U with OOTB settings.

My question is the same as iffex's. How about a PDF that explains in detail the trading principles, as well as describing the individual input controls?

Ah I don't have a PDF to share at the moment...

But here is the optimizer settings file. I'll add it to the top post. Download it from the first post. Just load up the strategy tester, and then load this file into it, and run the optimization on hourly time frame, open prices only.

This should cause it to start cranking out tuns of different strategy settings. The test will probably take 10 hours to run fully. But you'll end up with a lot of good strategies to pick from.

As for how the code all works under the hood, wow even I don't know it all totally anymore, and I wrote it lol. It's really just designed for the optimizer. See, I got frustrated how I would find a good strategy and then the market would change behavior and my strategy wouldn't work anymore.

So I designed this strategy to be able to be "optimized" where it wouldn't just tweak the settings, but where, in the optimizer, it would actually find totally different strategies altogether. In other words, it could be massively adapted to different market behaviors.

So it doesn't really have any one particular strategy it trades. It can use the stochastic all kinds of ways... or, depending on the optimization you picked, it may not use the stochastic at all. It can use linear regression many ways, or not at all. It can use moving averages many ways, or not at all. It can use price action bunches of ways (I think it always uses price action lol). But depending on what optimization settings you are using, it may find any number of indicators to use or not use. So it's hard for me to say exact what this strategy does, because it's designed to adapt to changing market behaviors, using the optimizer.

One thing I will say though, is this strategy has a "steps" feature which is the biggie. It basically spread out its entries over a range of pips, rather than just in one spot. This feature is controlled by the "MaxSteps1to7" variable. I recommend a value of 3, personally. That typically gives you a nice boost in profit, while keeping drawdown nice and low. Setting it to 5 is the absolutely highest I will trade... 5 will give you more pips, but larger drawdowns. 7 is just downright dangerous, but it's there. And actually, you could set it to any number you want. None of these variables have any restrictions, except maybe the lot size variable.

Hope this helps. Any specific questions, I'll be happy to answer more.

P.S. I'm going to add the strategy optimizer settings file to the top post...

looks very nice....but this forum is one of sharing and discussion....so unless u either explain the logic of how ur EA works or share the code (mq4), do not expect much interest in this thread....:-)

very interesting. thanks for sharing.

Please do make a PDF to help explain the various parameters and so on. Your OOTB robot gets great results, even when the starting equity is only 3000. But your "optimized" settings are blowing accounts all the way up to 25000! Shouldn't it be the other way around?

It would help to know what all the settings are capable of ahead of time.

Not sure what you mean by "optimized settings"? The optimized ".set" file I uploaded... supplies a range of settings, to be scanned in the optimizer. I assume you mean that you ran the optimizer and it gave you a list of different strategies, and you found one that gave more profit than my default settings in the .txt files?

If so then yes, you are correct, there are other strategies that it'll find that will give a lot more pips. But those tend to be more dangerous. Because the reason it's giving you more pips, is probably because it is trading more aggressively. The aggressiveness is in the "MaxSteps0to7" variable, which was probably higher in the strategy that gained more pips and that's part of why it gained more. What you want to look for though is strategies that..

a) have plenty of trades made... there is safety in numbers of trades.
b) have very low draw down.
c) show a smooth looking account balance growth on the graph.
c) doesn't have a MaxSteps0to7 value higher than 5.

Oh, I forgot to say, to use the Optimizer .set file, you need to put a check in the checkbox "Optimization" when you run it in the strategy tester. Otherwise it will just do a regular backtest. I'll add that to my instructions above.

Will do... as I can hehe.

Let's start with the external (optimizable) variables...

 

1. extern int Load_File_Settings0to1 = 1;
   This variable tells the strategy whether or not to load the settings from the appropriate .txt file (such as EUR.txt). It only works if there is a file available to load, for the market you are working with. It can have a value of 0 to 1. 0 is false, 1 is true. If you set it to false, it will use whatever settings you type in directly in the Expert Properties -> Inputs tab.
2. extern int Expert_ID = 1234;
   This variable does basically nothing, except it is used to generate the Magic Numbers, which help to keep trades from this strategy separate from signals made by other strategies.
3. extern bool auto_adjust_lots = false;
   This variable tells the strategy to automatically increase the lot size. I'm not totally sure off the top of my head if it works. I'm sure a backtest will quickly reveal if it works.
4. extern int UseMAExtries0to1 = 0;
   This variable tells the strategy whether or not to use the large moving averages on entering trades. The large moving average has more settings below, and it is the moving average that helps trade the trend. This variable can be 0 to 1, where 0 is false, 1 is true.
5. extern int UseMASteps0to1 = 1;
   This tells the strategy whether or not to increase the value of the moving average each time the market makes a "step", which I'll explain below. This variable can be 0 to 1, where 0 is false, 1 is true.
6. extern int UseMAPeriods4000orANY = 4000;
   This sets the value (the period) of the large moving average that the strategy uses to determine the trend. This variable can be any value you want.
7. extern int UseLinReg0to1 = 1;
   This variable determines whether the strategy will use the linear regression line or not. The linear regression line has more settings below, however, if this variable is set to 0, then the other settings below will (i think) not be used at all. Can be 0 to 1, where 0 is false, 1 is true.
8. extern int UseLinRegPeriods15orANY = 315;
   This sets the value (the period) of the linear regression line, which the strategy uses in about a dozen different ways (assuming UseLinReg0to1 is set to 1). I'll try to list all the ways the linreg is used in a sec. This variable can be any value, except zero, I guess.
9. extern int UseLinRegReverseTouch0to3 = 3;
   Ok here is a complex variable. It can be 4 different values, from 0 to 3 (aka, it can be 0 or 1 or 2 or 3, which is 4 different possibilities). Setting it to each value does some wildly different thing, so let me give a run down... these all will apply to SELL orders. The opposit would be true for buy orders...

1. Setting it to 0 means: The strategy will not require the current high (High[0] to be greater than the current value of the linear regression), to enter new sell orders where it previously was flat or where it was previously only buying.
2. Setting it to 1 or less.. (aka <= 1), causes it to not check for the Low[0] to be lower than the linear regression, when exiting all sells (I think for taking profit).
3. Setting it to 3 or more (aka >= 3) causes the strategy to neither not open new "steps" trades, nor exit trades for a loss, when the High[0] is <= the linear regression line.
4. Setting it to 2, turns "reverse touch" off completely.
5. Note: all the above applies to SELL orders only. It's all the other way around (aka Low[0] <= instead of High[0] >=) for buys.
6. P.S. this all hurts by head to think about, so if it hurts yours too, welcome to the club. lol. But you asked me to explain it, and I tried to warn ya that it was a lot to talk about...

1. extern int UseLinRegIncrement0to1 = 1;
   This variable determines if the linear regression value will increase during each "step" it makes. You can set this from 0 to 1, where 0 is false and 1 is true.
2. extern int HuggSteps0to3 = 2;
   Here is another complex variable. Hug steps is used during hedging. Basically what it means is that, if it's in a buy trade, and the buy has already made several steps, and then it enters a sell trade, should the sell start at step number one? Well, depending on this variable, it's possible the new sell hedge could "hug" to the buy, by starting at a higher step value. Bleh this is hard to explain... Here is the meanings of this variable for SELL orders...

1. Setting this to 0 means: It will not hug to the buys.
2. Setting this to any value above zero means that the new sell will start at a step value of... (NumberOfBuySteps - HuggSteps0to3 + 1). So if the buy was at step 3, and HuggSteps0to3 was set to 1, then the new sell would open at step... (3 - 1 + 1) which is 3. Thus, the buy is at 3, and the new sell would also open at 3, I think.

1. extern int MaxSteps1to7 = 5;
   This is the most powerful variable of them all. It determines how many steps it will make to spread out trades over a range of pips. It starts out all new trades at step 1 (unless it "hugged" to trades that were already opened in the opposite direction, as I explained above). Step 1 is always a trade of 1 lot. Each additional step adds lots in the amount of (CurrentLots * 2). Thus, increasing MaxSteps1to7 to a high value could result in a massive number of lots being opened at higher steps, which would be a martingale situation. But lower values such as a maximum number of steps of 3 (which is what I use), is safer and, if the rest of the strategy settings are good, it can significantly improve performance.
2. extern int StepIncrement0to1 = 0;
   This variable determines if the size of the distance between each step should increase with each new step. This value can be 0 to 1, where 0 is false, and 1 is true.
3. extern int Profit50toANY = 300;
   This is pretty easy: your take profit value. If it's set to 300, then it will take profit at 300 pips. The take profit is calculated toward the trades that were opened on the CURRENT STEP. So if the current step is 3, then the lots that were opened at step 3 are the ones who will get to have a take profit value of 300... but the lots opened at steps 1 and 2 will not be considered where the take profit is concerned (they will close when the lots at step 3 close, if step 3 is the last step). Basically, set this value to your desired profit for each trade.
4. extern int BaseStepSize50toANY = 50;
   This variable determines the number of pips between each step. Setting this to 50 will mean that it will add more lots each 50 pips the market moves against you, until it reaches MaxSteps0to7. Making this value larger will often provide more safety, as this gives you a larger range of area where you will be profitable, but a larger area also slows the strategy down, thus it would make fewer trades. So it's a balance. The optimizer will figure out what works best.
5. extern int BaseTradeLotSize = 1;
   This variable sets the base number of lots.I think I kind of explained this in my installation instructions. But the size of the lots is calculated in the code like this...
   
   double BLS = MarketInfo(Symbol(), MODE_LOTSTEP);
   double checkit = MarketInfo(Symbol(), MODE_MINLOT);
   if (checkit > BLS) BLS = checkit;
   
   Print("Using base lot size: " + BLS);
   
   LotSizes[0] = (BLS * 1) * BaseTradeLotSize; //<-- This is step 1.
   LotSizes[1] = (BLS * 2) * BaseTradeLotSize;
   LotSizes[2] = (BLS * 6) * BaseTradeLotSize; //<-- this is step 3, which I like to use.
   LotSizes[3] = (BLS * 18) * BaseTradeLotSize;
   LotSizes[4] = (BLS * 54) * BaseTradeLotSize; //<-- This is step 5. Pretty nasty.

Anyway hope the above info helps some to explain the settings that you can use in the optimizer, and what those settings do.

PS: I just noticed that I it looks like I had removed the code for adding lots above step 5. So setting MaxSteps0to7 to "6" or "7" probably won't work. Yeah, that would be just too dangerous. So it looks like I had removed that code completely. I forgot that I did that. Sorry all. It only works up to step 5. But you don't need to be blowing your accounts anyway, so it's all as well.

Yes, that would have been helpful.

So would a PDF explaining the guts of this thing. I see a bunch of "0-7" type of settings possible, but the problem is that we have no idea what the algorithym is behind the numbers. 0-7 of what, exactly?

MA? Which MA? EMA? WMA?

See my point?

I could say the same thing for the rest of the param input sections as well.

I'm not asking you to teach me how to trade, I'm asking to be made clear on the bits and pieces of your robot.

Just to let everyone know what they're getting into with optimization... I've been running it for just over 9 hours now (on a duo core computer with 4GB RAM), and the countdown timer still read 21hrs and some change. So this is becoming a total time investment of nearly 32hrs altogether.

I'm not implying you shouldn't try it yourself. Not at all. I'm just pointing out what to expect if you do.

Aye, that's what I had to do to find the settings for the EUR.txt, NZD.txt and JPY.txt files. It tokes a while. But as it is working, it shows you what it has found so far, and you can open another copy of MT4 and play around with those variations and see how they look. Helps pass the time. :-P

Well, so far it has run about 750 completed passes. A quick check of the progress report shows that only about 10 strategies have a DD of <30%. So only about one test in 70 is showing results that would be considered "safe" by an experienced trader.

I'm going to go ahead and let this finish all the way to the end, but I'd like to know what you consider to be "safe" DD numbers for how you trade? Do you go above 30% with your settings?

Well the EUR.txt settings file I have, which I found in the optimizer about a year ago, shows a maximal drawdown at the moment of 16% and a relative DD of 31%.

JPY currently shows a maximal DD of 38%, and relative of 48%. And it looks pretty smooth to my "eyes" when looking at the graph.

NZD shows maximal DD of 8% and relative DD of 20%. It does look pretty smooth.

Those are all with the settings files I included above, which I got out of the optimizer I think November of last year.

I don't think that a DD of "30%" means you will DD 30% of your account balance. Becauase all you'd have to do is lower your lots, to avoid DD'ing so much of your account balance (or have more funds in your account + keep your lots the same size). So I assume it means it can lose 30% of the strategies total profit. I'm really not an expert on how MT4 calculates DD. But I'm assuming that a 30% dd means that it gives back 33 pips for every 100 pips made? Which, if that's the case, then that's not bad at all...

Right, if a trader could predict that his highest give-back was 33 pips out of every 100 then we would all be millionaires within weeks.

FWIW, some of the returns I'm seeing are coming back with DD of well over 70 and 80%. But even a give-back of just 51% means that you are a loser in the long run.

If any of my posts sound ignorant it's because this is the first time I've run an optimized EA. So there are still some things I'm getting used to, and so I might come up with some questions that you take for granted.

Here's one now:

You have included 3 settings files. So... does this mean that those settings are the ones being used automatically when I run that pair, or is there some step along the way where I must specifically select that file to be attached to the tester?

Just 18 hours to go! LOL...

i'm really missing some explanation of the strategy your EA uses. Can you please post it along with maybe the mq4 file so i can see the code too?
And are you optimizing the EA over a couple of years, or is the EA self-optimizing/adapting ?

It's automatic, as long as the Load_File_Settings0to1 is set to 1 (it is by default). However, it NEVER loads those files if you are running an optimization, as obviously that would cause it to not optimize anything lol. But if you just run a single backtest, and that variable is set to 1, then it loads the settings in that file. Otherwise if you want to manually type in your own settings, and not load the file, then you'll want to set that variable to 0.