Random entry EA - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Any other ideas? TP or SL maybe?
i dont think anyone can prove (not even the great igor) this random entries and random exits will lead to total loss
all this out of the box thingy is taking us very far indeed but good old common sense tells us random + random = 2 x random.
just a set percentage of balance for lot size i think. nice and simple
The martingale will bit us in the ass. having said that I have never seen anyone use 2 mm systems one for buy and one for sell and i'm not sure how that would work let alone code it.
Cheers
Beno
OK we are looking for a very creative mm solution for this system any suggestions would be greatly appreciated I have attached the trades for the eurusd 1min.
There is no such thing as a bad idea just un tested.
Cheers
Beno
I don't know how your EA works but take a look on this MM:
MoneyTec Traders Community Forum
Money Management:
Firstly let me comment on the existing MM's, that of fixed % or account size(ie: every trade is x% of total account size). I have tried this method and do not like the results at all becuase there are no built in win goals and when you do start profiting form trades and yoru acct size grows you are now risking larger lots on subsequent trades that may go against you, I've experienced it and again I dont have enough faith in any strats to be a long term winner with that type of MM.
My Light-Bulb MM concept:
1) Plan for disaster, plan that you are going to lose more trades then you are going to win. In the case of my strat using MA's, some trades will be up to a 50 pip loss(I set that as my max SL), while others might be a 1 pip profit or even a 200 pip profit on large trends. (Two optimizations I'm testing in my strat is to add lots on to winning positions and using trailing stops(my current method) vs. following the indicators for exits and taking profits via the scheme: 30(close 1/4),60(close 1/4),90(close 1/4), let the last 1/4 of the position ride till indicator says exit. again it would be really good to have ea's do all this backtesting but the concepts of the MM are such that these are all minor issues)
2) Set a win goal per series of X DOLLARS(not pips)
3) Start a new series once your win goal is reached( a series may take 1 hour to 5 days+ to get resolved into a win goal)
4) Use the "unlimited bankroll concept". Mental Exercise for you all: If you had an unlimited bankroll, say $1M. And you were to take $10 lossesx1000 trades, you would not even make a dent in your bankroll and if the MM System you are using calls for larger risks say $20 etc... then you know you have the bank roll to sustain it. In other words dont rely on backtested strat drawdown numbers to lull you into a false sense of security( "Oh my drawdoan max was 23% so that means I'll never go below 23% inthe future", WRONG! You might and you probably will so forget about drawdown and devise the MM in terms of greater monetary risks you'll need to take to realize your win goal.) When this MM busts its becuase you ran out of money, not because you did'nt have a plan so be sure to start with a strat and a lot amount that can sustain your strats performance AND:
5) PROGRESSIONS! This is the secret folks. Say you have a win goal of $200, if your strat makes more then $200 on your first trade, you simply bank the money and start again with another trade hoping to make $200, the moment you start losing is where it gets interesting. There are 3 types of MM that I want to cover:
Up as You Win
Up as You Lose
Fixed Lot Size
Now obviously we can't use a fixed lot size as our MM becuase then your strat needs to be a big winner. Up as you lose is the dangerous game that all gamblers play and we cant do that either(I've tested this in lenght it just simply does'nt work). Up as you win is the way to go, let me explain:
A typical strat will contain winning trades and losing trades and they will be scattered such as:
WLLLWLLWWWLWLWLWLLLLWLLLWWLLLLWLLLWLWWLWWLWW
What I propse is to increase your lot size(not your pip requirement or take profit levels of your strat simply your lot size) as you win trades and decrease your lot size as you lose trades, that way when your strat starts losing due to "bad" market conditions/false signals etc... you'll always be lowering eventually down to such a small size that you wont even care if you win or lose you'll just be waiting for the win so you can trade a higher lot amount on the next trade. I'll go further on to say that if you lose X trades in a row, simply revert to paper trading untill a win occurs and then enter the market with real money. Now this all comes down to what kind of progression you are using and starting lot size. With my $5,000 account my starting lot size is 0.3 (get a broker that lets you trade fractions). I give myself 2 trades to make my win goal of $100, if I fail to make that goal within 2 trades my progression begins. I have a certain trigger for my progression on when to escalate and how fast to escalate lot size, if I lose 2 trades in a row I begin a decrease of my lot size until I reach my starting point of 0.3 and if I lose 2 more at my base level I simply stop trading and revert to "paper signals". If at any point in my escalation win progression my next "strat" trade yields me a profit of $100 or greater(FOR THE ENTIRE SERIES LOSS TO DATE!), I restart the progression at 0.3. I'm trading this live right now while trying to further optimize the points I brought up above. The beauty of this method is that when the trades are going my way and I'm gaining 80+ pips like the short i'm in now on the GBPUSD, I follow the strat for exits and the measily $100 win target turns into a $500+ profit. Once my account size reaches $10,000 I will simply double my progression and unit values so my starting point is 0.6 lots.
I am still refining my system and must point out that I AM PREPARED to deposit more funds in the event that I am given win signals to escalate my unit size and then losses start happening at my higher lot levels. I trade with confidence knowing that "EVENTUALLY" the trends will start happening and even the small 30 pip profit trades add up with my increased lot sizes to start eating into the loss level of a series and bring me to my $100+win goal.
I will share the details of my progressions with contributors to this concept and again ask that any EA/SCRIPT programmers please contact me so that I can begin testing different progressions along with a simply ma strat or any strat that you think is good.
SO there you have it , the secret truly is MM because with an MM PROGRESSION you DONT need to have a winning strat that is above break even and as long as you have the bankroll and start out with low lot amounts and ONLY ESCALATE as you win trades AND are prepared to increase your bankroll and FOLLOWTHRU on the big trades when your progression calls for it, you will succeed!
Looking forward to feedback and discussion on this.
odirlei Thanks for the reply
nice strategy if you no which trades have a higher probability of being a winner but with our system we don't. having said that I have been playing around with the attached excel sheet using the sysem you posted.
Cheers
Beno
odirlei Thanks for the reply
nice strategy if you no which trades have a higher probability of being a winner but with our system we don't. having said that I have been playing around with the attached excel sheet using the sysem you posted.
Cheers
BenoOK.
Well, there is a lot of Money Management discussions here:
Money & Risk Management - MoneyTec Traders Community Forum
Maybe one of them can help you.
Regards.
To Beno or bobfourie or others,
An interesting feature to add to this EA would be the possibility to set a T/P and a S/L in pips where one can choos the amount wanted.
That would show clearly the correlation between between hitrate and avg. win/loss ratio.
And if I can make an other suggestion is to make an minimum amount of bars for a trade. Because if you test this on a 1min TF then one can notice that in many cases the system already turns position after 2-3 minutes. That is not really helpfull for a test because this type of trading is only going to proof that spread is the killing element in a random set up.
The combination of a minimum amount of bars together with a S/L and T/P will show you more then interesting techinal reports.
regards...iGoR
Hi beno,
here's the new EA I discussed with you. Could help me forward test. If you find any error, just let me know.
Beno and bobfourie,
I did a couple of test with that last EA.
I don't know if bobfourie programmed the randomness element or if that was in the original version.
But I just want to warn for test or conclusion or T/P's or S/L's that you would optimize because this EA is not "really random".
I have a random trade generator in Metastock Professional. And a random trade generator should take in trades randomly meaning that each time you do a test the trades or entrys should be each time completely different.
Every test I do with this EA, the entrys are on the exact same bar. So that is quite dangerous because before you know it you start to optimize and you start to curve fit.
I could be wrong or doing something not correct. But if it is the case as I explained then this is not a random trade generator and as said dangerous to make conclusions on the tests.
Ex. If I do a test on the FDAX with a T/P and an S/L that are equal then it gives me every time a result that the most consecutive losses are ONLY 2 and most consecutive profits are also ONLY 2. I checked the times of entry and on each test I run and the times of entry were exactly the same.
I think you understand that if you run a test where the entrys will be 100% random and on 710 trades and with a T/P that is equal to the S/L that you will have more then only 2 consecutive losses.
So before you know it one would build an MM on those numbers but would not be correct.
To do reliable test and to make reliable conclusions I think it would be better to take it back from the beginning and make sure that the randomness is realy random. But as said I could have overloocked something...
Keep up the good work.
Friendly regards...iGoR