Follow the BIG Candle

nice thread abokwaik, i see potentials in you.
looking forward to test your ea later in the day.
regards
Austine

In this case you need to download the expert from post # 23.

@ ِAustine : Thanks .

I would like see some constructive ideas based on visual testing, along with some screen shots where the script fails, in order to analyze and try to handle in the script before going live.

Few ideas for discussion:

1. Out of your tests is it better to open on current candle (candle_period=0) or Previous candle(candle_period=1) ?
2. What if we add a break-even stop ? this will be less than trailing stop. For example Stop Loss could be 5xatr, Trailing Stop 10xatr and the new break-even stop could be 3xatr. Or we move to breakeven after partial profit taking ?
3. If we decide to open based on Previous candle, Should we take the candle SHAPE into consideration (i.e. Hammer, Engulfing ... etc.) ?

Tests in real time on a demo account. 12 pairs. Still no position.

Which timeframe?

Please note that the script is not YET able to trade multiple pairs simultaneously. So. its better for now to stick to one pair, or use multiple demo accounts/different brokers.

M30. No couple is not twice. So should not be a problem. On another account I will give another setting.

Thanks. Will demo on 4 different pairs in 4 different accounts.

Well I have old and new (509 and 600) version of MT4 and neither indicator works ...Thread owner on three pages did not post chart yet

Like your simple and easy ideas for this ea.
I will download and test it BT 99% on this coming weekend when i free.

Will follow this thread...

Thanks

Please note that this is an Expert Advisor not indicator . It is intended to trade automatically for you. File has to go into experts folder, please review earlier posts.

Charts are not of importance in such an EA becuase you can see how the Advisor opens/closes trades using the Strategy Tester (plus of course the strategy description provided in this thread).

However, I will not let you leave empty handed. Here's a chart for a trade taken by this Advisor, I hope the chart is clear. The yellow lin in the chart is EMA200 settable from EA in addition to all values shown on the chart ATR/TP/SL/TS...ets.

People good with charts please post some trade examples from your visual tests or demo tests.

Regards,

Well, don't know what to say. No trade taken for me all day long. 4 different accounts with 3 different brokers on M30. Smiley on top right corner ok.

Hello Abokwaik,

About two years ago I created a very similar EA. Interestingly, believe it or not, I called it exactly like you did, i.e. "Follow the big candle" :-)
I spent many weekends testing my sys all over again, improving the parameters, fine-tuning position size, experimenting with using ATR from a higher timeframe, entering on retracement with limit order, using dynamic SL and TP ATR-based, etc, etc.

I thought it may be helpful if I share a few things I learned in my journey. You may already know some or all of this. In this case it will be no news. Becasue our systems are so similar, some of my lessons may be relevant.

Before I share my inputs, I will honestly and upfront tell you how my work on my system ended. I dropped the idea. I have not even started trading it live. My system failed on demo. It still runs on demo on my VPS, just for fun and for me to see, but it is not making money. To be clear, I am trying to say that your system will not. I hope it will.

So here is I would suggest that you consider:

1. test it on high quality historical data. If you trade it on H1, ideally you need 10 years of history, hundreds of trades, not around 40. Where to get it? Ask your broker, try dukascopy data feed, you can also try fxcm but you need to pay for it unless you deposit $5k.
2. make sure you set the realistic spread for backtest.
3. before you trade it live, definitely forward test it on demo. ideally on a VPS, to keep connectivity. this is a kind of trend-following system with high reward : risk ratio, so you cannot afford to miss a trade.
4. see results of demo. compare to backtest results for the same period
5. if you trade it live, start with microlots, risk just $20 per trade. Let it make 20-40 trades and increas the position only if profitable.
6. do not have more than 2-4 parameters, otherwise you risk over-optimization or curve-fitting
7. read the free ebook at mechanicalforex.com , it is the best free material I found on mechanical systems
8. check the stability of the parameters. turn off genetic algo in mt4 and draw a chart of parameter setting vs. balance. If you have great results with TP 35 pips, but very poor with TP 36 pips, it is a very bad sign. Pick a parameter not the highest but in the middle of the "mountain" on the chart. Plot the two parameters map in MT4 and look for green "island", i.e. combination of parameters surrounded by other profitable combinations. This is good. A single green point is bad. Choose not the parameter with the highest profits, but the one in the middle of the green island.
9. notice how many trades per month / per year it is doing. Think if this is acceptable to you. If it does 60 trades per year, will you have patience to let it run? What was the worst losing streak in demo? 4 trades? multiply it by 2 and think it you will be able to withstand it. Adjust position size in such way that even if you lose 8 trades in a row, it will still be an acceptable loss. If position size is so adjusted, is the expected profit satisfactory?
10. check the relation between CAR (compound annual rate) and MaxDD (maximum drawdown). Good systems have CAR 2x higher than DD. Around 1.5 is still acceptable I guess. Below 1 is bad. BTW, very important, check this NOT on your backtest results posted here. Check it on 10 years of history and hundreds of trades. I am sorry to say that, but with less than a hundred trades (as I saw in attached screenshots from backtester), it may be a curve fit.
11. use static datetime Time[0] kind of code to check for new bar, do not use Volume[0]>0, it is not reliable
12. test it on data feeds from a few different brokers. ideally results should be almost the same.

I hope I have not negatively impacted your confidence in sucess. I hope this list of ideas helps you develop this idea to a success. If you follow the suggestions here, you should be safe. Good luck!

Sorry for the late reply as I was on holiday for the weekend.

Before thinking of starting the thread I searched the net for similar ideas and couldn't find any. I do not know if you have posted your work anywhere on the net, or it was for private use.

I thank you for sharing your experience, and I will surely take your comments and advice into consideration.

I invite you to stay with us on this thread if you like, and share your thoughts as we go along.

Regards,

Khalil

For the chart lovers:

This is a live demo trade (forward testing) taken by the script on M5 timeframe on Friday when I was away. This was meant only to show some charts with live trades. I do not recommend trading this on short time frames as it is not intended for Scalping.

The script was running on 5-digit broker. It made 40+ pips on this trade.

OK. Here is a Quiz for you on the same chart (attached). Why the EA didn't make a trade on the Big Bear Candle denoted on the chart ??

This is disappointing. almost 2 hours with no answer.

Anyway, since we are trading using ATR, and at the time the big bear candle ended ATR(21) was around 10 pips, and as per the setup we are opening trades only if the candle size is >= 3xATR=30 pips which is greater than the candle size (18 pips). While in the case of the Bull candle where the Long trade opened Candle size was also around 18 pips, but ATR at that time was 5 (so the candle size was > 3xATR=15 pips).

So obviously this is a Volatility Break-out strategy looking to catch long trends.

dont be disappointed,
u know its weekend and almost all the traders are in Miami beach now
good job

OK, that explains it.

Soon I will post release 3 of this EA, that is able to handle multiple pairs for forward testing. The new release will also have some new parameters

1. Break Even Stop (as explained in post # 44)
2. and probably a new parameter I will name it Candle Effective Size. For Bull candle this will mean Close-Low, and for Bear candle it will mean High - Close.

Here is the New Release with 2 added parameters :

1. Break Even Stop (as explained in post # 44) --> to disable this parameter set it higher than Trailing Stop
2. Candle Effective Size (as explained in previous post) : The use of this is that the system will open a trade only if the Close-Low (for Bull candle) or High-Close (for Bear candle) is greater than the Candle Effective Size (in terms of ATR multiples) --> to disable this parameter set it to 0 (zero)

I have run the script on the major pairs on M30 time frame with same parameters and period. Results (Excel) and Parameter file attached.

Please note that the EA file with _x suffix is for MT4 versions prior to Build 600.