Download MetaTrader 5
To add comments, please log in or register
Use MQL5 Cloud Network. Test robots at a super speed!
xptrader
55
xptrader 2015.07.17 21:13 

Dear fellow traders,

I am working on a strategy on EURUSD, that is based on Martingale but with a stop after the 3rd open position, and without any indicators. It should work like this (let's assume there is no spread):

- On EURUSD, at all times, but I expect profits to be better during active trading hours.

- 1. Open LONG trade 'L1' (or SHORT, direction doesn't matter for the first trade) on current price (which I will call "OPEN"): 1 lot (or mini or micro depending on your capital), SL 40 pips, TP 10 pips

AND

- 2. Prepare for a SHORT trade 'S1': open on"OPEN-30pips", 5 lots, TP 10pips, SL40pips

If 1. hits TP without hitting the short order: OK: 10 pips (close pending SHORT order and restart)

If 2. gets ordered and hits TP = 10 pips profit (-40 Long + 50 Short). If not and price reverses to the OPEN a third trade is opened (this one was made pending as soon as the second one was triggered):

Third order: identical to the first: LONG 'L2', same SL/TP, but 20 lots.

Now here it comes:

- if the third hits TP (OPEN+10) rigth away: 10pips profit (+10 -200 +200) BUT

- if it goes down, losses will be big, and a fourth order (SHORT) at the same level as the second one would require too much margin (80 lots (you need 4 times more than the previous to make 10pips)). Besides, and I tested this, price may fluctuate like this up to the fourth and even fifth order.

So after the third order opened (we now have 2 long and 1 short position open) we place a 10 pip trailing stop on the last order. Max loss will be (-10 (L1) -100 (S1) -200 (L2)) -310 pips if price immediately reverses. But it could be better, (-5 +125 -100) -230 if price hits OPEN+5 before reversing, or price hits OPEN+10 = 0 -150 0 = -150. Then you can even let it run and hope for an extra 10pips and break even (0pips). Or more (?+pips).

So now here is my question:

- Who can verify the number of instances (over a considerable time) where price would behave like this i.e. not going up >10pips, followed by going down between 30-40 pips, and back up above "open + <10pips" and back down?

From Jan 2001 until today (>14 years or about 86000 trading hours (1H)): 27240 times price went directly hitting the first long TP (1 out of 3).

But for all the other instances I'm breaking my head to find a way to calculate what the probability is of the scenario described above (theoretically price could move up and down in a 1H bar and there is no way of knowing just looking at the bar, and I don't have the PC-power to work on 1-5 min data). I can do it on hourly data, but there are 86000 rows (!) and it should (as far as I know) be done manually and on smaller time data since price can fluctuate in 1H up and down.

If (worst case of 310 pip loss) this would only happen every 32 trades, this is a sure win game. This is the worst case. I think a 1 in 25 should do it.

But maybe the stats are way better.

Maybe only on active hours.

Who knows?

I don't.

Yet.

Thank you all for your input.

BR.

Keith Watford
Moderator
8007
Keith Watford 2015.07.17 21:56  

I have seen many people try different variations of martingale.

I have not seen a single instance where they haven't resulted in disastrous losses.

You may think that you have conceived an original idea, but believe me, it has been tried many times and failed. 

JD4
1101
JD4 2015.07.17 22:15  
Even if it is a totally new concept, but it being based on martingale as it is, the track record so far on systems based on that strongly suggest that this attempt will wipe out the account it is being traded on.  I have been doing a lot of reading, trying to learn more and more each day about Forex, and I have not ever found someone who tried trading a martingale system that didn't end up either wiping out their account, or quit trading it before it did.
xptrader
55
xptrader 2015.07.18 09:39  
JD4:
Even if it is a totally new concept, but it being based on martingale as it is, the track record so far on systems based on that strongly suggest that this attempt will wipe out the account it is being traded on.  I have been doing a lot of reading, trying to learn more and more each day about Forex, and I have not ever found someone who tried trading a martingale system that didn't end up either wiping out their account, or quit trading it before it did.

Thank you both,

I agree. Martingale leads to total loss. That's why I thought of this system, cutting losses on the third trade.

I agree that this may also be a loosing strategy but I would like to be sure of it, mathematically.

good trading.

xptrader
55
xptrader 2015.07.19 13:14  

Ok, so I gave it a shot.

Unfortunately I can't program an EA to test this so I worked in excel with macro's on hourly data. It took me a lot of time to get the formula's and macro right!

The test uses a fixed SL at 10pips after the third open trade (no way I will try to test a trailing stop in excel! :) )

Results: +6104 pips on 40.000 hours (about 8 months) BUT this assumes a new trade every hour which is not possible in reality (you may then have >10 open positions).

Assuming the system closes (TP or SL reached) within 4 hours (seems reasonable), we need to divide 6104 by 4 = 1526 pips in 8 months if trading 24/5 (which could be done with an EA).

190 pips a month.

Considering we start with 0.01 lots, then  0.05, then 0.2 for the last, we have a profit of approx 19*1.8 € = 34.2 € per month with an EA. Much less if you trade manually (perhaps 10€/month).

WOOW LOL

So, unless someone wants to give it a shot writing and testing the EA in MT4, that's it for me...

Br.

JD4
1101
JD4 2015.07.19 17:56  
I think that given the previous record of martingale style systems, it would be highly unlikely to find someone who would want to spend the effort making an EA on it.  However, there might be someone who would take it on as a practice experiment in learning how to code an EA, maybe you could ask this guy I heard talking about a new martingale system idea, goes by the user name xavier40. ( :p )
whroeder1
13644
whroeder1 2015.07.19 18:42  
xavier40: +6104 pips on 40.000 hours (about 8 months) BUT this assumes a new trade every hour
  1. 40K hours / 24 hours per day / 20 trading days per month = 83 months = 7 years..
  2. 6104 pips/ 40K hours is 1 pip / 6 hours.
xptrader
55
xptrader 2015.07.19 20:07  
WHRoeder:
xavier40: +6104 pips on 40.000 hours (about 8 months) BUT this assumes a new trade every hour
  1. 40K hours / 24 hours per day / 20 trading days per month = 83 months = 7 years..
  2. 6104 pips/ 40K hours is 1 pip / 6 hours.

Sorry, typo, should read 4.000 hours (4000 1H bars).
xptrader
55
xptrader 2015.07.19 20:11  
JD4:
I think that given the previous record of martingale style systems, it would be highly unlikely to find someone who would want to spend the effort making an EA on it.  However, there might be someone who would take it on as a practice experiment in learning how to code an EA, maybe you could ask this guy I heard talking about a new martingale system idea, goes by the user name xavier40. ( :p )

LOL, yeah right, xavier40 is a great coder. :p

Basically I am giving up. Never really believed in Forex trading. Price could go down or up, 50/50 chance, ALWAYS. Never mind if you are in a down- or uptrend, touching RSI2, above EMA 200 or whatever. Price has no memory of the past.

I look for statistical mechanical trading mechanisms and Forex doesn't do it.

Volatility trading does.

Best to all of you.

JD4
1101
JD4 2015.07.19 23:07  

No market traded like this is any different than Forex, other than the item(s) being traded.  Every single one of them always has that possibility.  Like trading the other markets, it is all about probabilities.  There are statistical mechanical trading strategies for all traded markets.  If you do not find one that works for you in Forex, then find one in another market that works how you want it to, and adapt it for Forex, problem solved.

/
To add comments, please log in or register